Designs for Quantitative Nursing Research: Quick-Access Chart

Descriptive Study Designs

Typical descriptive study designs, p. 216
Comparative descriptive designs, p. 217
Time-dimensional designs:
Longitudinal designs, p. 219
Cross-sectional designs, p. 220
Trend designs, p. 221
Event-partitioning designs, p. 222
Case study designs, p. 223
Correlational Study Designs
Descriptive correlational designs, p. 225
Predictive designs, p. 226
Model-testing designs, p. 227
Quasi-experimental Study Designs
Nonequivalent comparison group designs:
One-group posttest-only design, p. 234
Posttest-only design with comparison group, p. 234
One-group pretest-posttest design, p. 234
Pretest and posttest design with a comparison group, p. 237
Pretest and posttest design with two comparison treatments, p. 237
Pretest and posttest design with two comparison treatments and a standard or routine care group, p. 237
Pretest and posttest design with a removed treatment, p. 238
Pretest and posttest design with a reversed treatment, p. 240
Interrupted time-series designs:
Simple interrupted time-series designs, p. 242
Interrupted time-series design with a no-treatment comparison group, p. 242
Interrupted time-series design with multiple treatment replications, p. 243
Experimental Study Designs
Classic experimental design, p. 245
Experimental posttest-only comparison group design, p. 246
Randomized blocking design, p. 246
Factorial design, p. 247
Nested design, p. 248
Crossover or counterbalanced design, p. 249
Clinical trials, p. 250
Randomized controlled trials, p. 251
YOU’VE JUST PURCHASED

MORE THAN
A TEXTBOOK
ACTIVATE THE COMPLETE LEARNING EXPERIENCE
THAT COMES WITH YOUR BOOK BY REGISTERING AT

http://evolve.elsevier.com/Grove/practice
Once you register, you will have access to your
FREE STUDY TOOLS:

Evolve Student Learning Resources for Grove, Burns, & Gray: The Practice of Nursing Research:
Appraisal, Synthesis, and Generation of Evidence include the following:

•D
 ata Sets and Data Set Activities
Qualitative and quantitative data sets with activities that provide practice in working with actual
research data

• Interactive Review Questions

More than 400 questions, with feedback for both correct and incorrect responses

• Sample Research Proposals

Five sample proposals that serve as excellent models for beginning nurse-researchers

REGISTER TODAY!
 SEVENTH
EDITION

THE

PRACTICE OF
NURSING RESEARCH
Appraisal, Synthesis, and Generation of Evidence
Susan K. Grove, PhD, RN, ANP-BC, GNP-BC
Professor
College of Nursing
The University of Texas at Arlington
Arlington, Texas;
Adult Nurse Practitioner
Family Practice
Grand Prairie, Texas

Nancy Burns, PhD, RN, FCN, FAAN

Professor Emeritus
College of Nursing
The University of Texas at Arlington
Arlington, Texas;
Faith Community Nurse
St. Matthew Cumberland Presbyterian Church
Burleson, Texas

Jennifer Gray, PhD, RN

George W. and Hazel M. Jay Professor
College of Nursing
Associate Dean and Chair
Department of MSN Administration, Education, and PhD Programs
The University of Texas at Arlington
Arlington, Texas
3251 Riverport Lane
St. Louis, Missouri 63043

THE PRACTICE OF NURSING RESEARCH: APPRAISAL, SYNTHESIS, 978-1-455-70736-2

AND GENERATION OF EVIDENCE, SEVENTH EDITION
Copyright © 2013, 2009, 2005, 2001, 1997, 1993, 1987 by Saunders, an imprint of Elsevier Inc.

No part of this publication may be reproduced or transmitted in any form or by any means, electronic or
mechanical, including photocopying, recording, or any information storage and retrieval system, without
permission in writing from the publisher. Details on how to seek permission, further information about the
Publisher’s permissions policies and our arrangements with organizations such as the Copyright Clearance
Center and the Copyright Licensing Agency, can be found at our website: www.elsevier.com/permissions.

This book and the individual contributions contained in it are protected under copyright by the Publisher
(other than as may be noted herein).

Notices

Knowledge and best practice in this field are constantly changing. As new research and experience
broaden our understanding, changes in research methods, professional practices, or medical treatment may
become necessary.
Practitioners and researchers must always rely on their own experience and knowledge in evaluating
and using any information, methods, compounds, or experiments described herein. In using such
information or methods they should be mindful of their own safety and the safety of others, including
parties for whom they have a professional responsibility.
With respect to any drug or pharmaceutical products identified, readers are advised to check the most
current information provided (i) on procedures featured or (ii) by the manufacturer of each product to be
administered, to verify the recommended dose or formula, the method and duration of administration, and
contraindications. It is the responsibility of practitioners, relying on their own experience and knowledge
of their patients, to make diagnoses, to determine dosages and the best treatment for each individual
patient, and to take all appropriate safety precautions.
To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume
any liability for any injury and/or damage to persons or property as a matter of products liability,
negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas
contained in the material herein.

Library of Congress Cataloging-in-Publication Data

Grove, Susan K.
  The practice of nursing research : appraisal, synthesis, and generation of evidence / Susan K. Grove,
Nancy Burns, Jennifer Gray.—7th ed.
   p. ; cm.
  Nancy Burns is first named author on previous edition.
  Includes bibliographical references and index.
  ISBN 978-1-4557-0736-2 (pbk.)
  I.  Burns, Nancy, Ph.D.  II.  Gray, Jennifer, 1955-  III.  Title.
  [DNLM: 1.  Nursing Research—methods.  2.  Evidence-Based Nursing.  WY 20.5]
  610.73072—dc23
    2012019862

Executive Content Strategist: Lee Henderson

Associate Content Development Specialist: Julia Curcio
Publishing Services Manager: Jeff Patterson
Production Manager: Hemamalini Rajendrababu
Senior Project Manager: Antony Prince Working together to grow
Design Direction: Karen Pauls libraries in developing countries
Printed in China www.elsevier.com | www.bookaid.org | www.sabre.org

Last digit is the print number:  9  8  7  6  5  4  3  2

To our readers and researchers, nationally and internationally, who will
provide the science to develop an evidence-based practice for nursing.

To our family members for their constant input, support, and love,
especially our husbands
Jay Suggs,
Jerry Burns,
and
Randy Gray

Susan, Nancy, and Jennifer

 Contributors

Daisha J. Cipher, PhD Christine Miaskowski, RN, PhD, FAAN

Clinical Associate Professor Professor & Associate Dean
College of Nursing Physiological Nursing
University of Texas at Arlington University of California
Arlington, Texas San Francisco, California
Chapters 22, 23, 24 & 25 Chapter 29

Kathryn M. Daniel, PhD, RN Rosemary C. Polomano, PhD, RN, FAAN

Assistant Professor Associate Professor of Pain Practice
College of Nursing Department of Biobehavioral Health Sciences
University of Texas Arlington University of Pennsylvania School of Nursing
Arlington, Texas Philadelphia, Pennsylvania
Clinical Educator Faculty
Diane Doran, RN, PhD, FCAHS Department of Nursing
Professor Hospital of the University of Pennsylvania
Scientific Director, Nursing Health Services Philadelphia, Pennsylvania
Research Unit (University of Toronto site) Associate Professor of Anesthesiology and Critical
Lawrence S. Bloomberg Faculty of Nursing Care
University of Toronto Department of Anesthesiology and Critical Care
Toronto, Ontario University of Pennsylvania School of Medicine
Canada Philadelphia, Pennsylvania
Chapter 13 Chapter 14

Kathryn Aldrich Lee, RN, PhD

Professor and Associate Dean for Research
James and Marjorie Endowed Chair in Nursing
Family Health Care Nursing
University of California
San Francisco
San Francisco, California
Chapters 20 & 27

Judy L. LeFlore PhD, RN, NNP-BC, CPNP-PC &

AC, ANEF
Director
Pediatric, Acute Care Pediatric, Neonatal Nurse
Practitioner Programs
Nursing
University of Texas at Arlington
Arlington, Texas
Nurse Practitioner
Advanced Practice Services
Children’s Medical Center, Dallas
Dallas, Texas
Chapters 10 & 11

vi
 Reviewers

Lisa D. Brodersen ED, RN Ida Slusher, RN, DSN, CNE

Allen College Professor & Nursing Education Coordinator
Waterloo, Iowa Department of Baccalaureate & Graduate Nursing
Eastern Kentucky University
Sara L. Clutter, PhD, RN Richmond, Kentucky
Associate Professor of Nursing
Waynesburg University Jeanne M. Sorrell, PhD, RN, FAAN
Waynesburg, Pennsylvania Cleveland Clinic
Cleveland, Ohio
Josephine DeVito, PhD, RN
Associate Professor Molly J. Walker, PhD, RN. CNS, CNE
College of Nursing Associate Professor, Angelo State University
Seton Hall University San Angelo, Texas
South Orange, New Jersey
Angela F. Wood RN, NNP-BC, PhD
Jacalyn P. Dougherty, PhD, RN Carson-Newman College
Aurora, Colorado Jefferson City, Tennesse

Betsy Frank, RN, PhD Fatma A. Youssef, RN, MPH, DNSc

College of Nursing, Health, and Human Services Professor of Nursing, Marymount University
Indiana State University Arlington, VA
Terre Haute, Indiana
Mary Beth Zeni, MSN, ScD, RN
Sharon Kitchie, RN, PhD, CNS-BC Senior Nurse Researcher, Cleveland Clinic
Patient Education and Interpreter Services Specialist Cleveland, Ohio
Upstate University Hospital
Syracuse, New York

Madelaine Lawrence, PhD, RN

Queens University of Charlotte
Charlotte, NC

vii
 Preface
R
esearch is a major force in the nursing profes-
sion that is used to change practice, education,
and health policy. Our aim in developing the
seventh edition of The Practice of Nursing Research:
Appraisal, Synthesis, and Generation of Evidence is to
increase excitement about research and to facilitate the
development of evidence-based practice for nursing.
It is critically important that all nurses, especially those
in advanced-practice roles (nurse practitioners, clini-
cal nurse specialists, nurse anesthetists, and nurse mid-
wives) and those assuming roles as administrators and
educators, have a strong understanding of the research
methods conducted to generate evidence-based knowl-
edge for nursing practice. Graduate and undergraduate
nursing students and practicing nurses need to be
actively involved in critically appraising and synthe-
sizing research evidence for the delivery of quality,
cost-effective care. This text provides detailed content
and guidelines for implementing critical appraisal and
synthesis processes. The text also contains extensive
coverage of the research methods—quantitative, qual-
itative, outcomes, and intervention—commonly con-
ducted in nursing. Doctoral students might use this text
to facilitate their conduct of quality studies essential
for generating nursing knowledge.
The depth and breadth of content presented in this
edition reflect the increase in research activities and
the growth in research knowledge since the previous
edition. Nursing research is introduced at the bacca-
laureate level and becomes an integral part of graduate
education (master’s and doctoral) and clinical practice.
We hope that this new edition might raise the number
of nurses at all levels involved in research activities
to improve the outcomes for nursing practice.
The seventh edition is written and organized to
facilitate ease in reading, understanding, and imple-
menting the research process. The major strengths of
this text are as follows:
• State-of-the-art coverage of EBP—a topic of vital
and growing importance in a healthcare arena
focused on quality, cost-effective patient care.
• A clear, concise writing style that is consistent
among the chapters to facilitate student learning.
• Comprehensive coverage of quantitative, qualita-
tive, outcomes, and intervention research methods.
viii
• A balanced coverage of qualitative and quantitative
research methodologies.
• Electronic references and websites that direct the
student to an extensive array of information that is
important for conducting studies and using research
findings in practice.
• Rich and frequent illustration of major points and
concepts from the most current nursing research
literature from a variety of clinical practices areas.
• A strong conceptual framework that links nursing
research with EBP, theory, knowledge, and
philosophy.
Our text provides a comprehensive introduction to
nursing research for graduate and practicing nurses.
For use at the master’s and doctoral level, the text
provides not only substantive content related to
research but also practical applications based on the
authors’ experiences in conducting various types of
nursing research, familiarity with the research litera-
ture, and experience in teaching nursing research at
various educational levels.
The seventh edition of this text is now organized
into 5 units and 29 chapters. Unit One introduces the
reader to the world of nursing research. The content
and presentation of this unit have been designed to
introduce EBP, quantitative research, and qualitative
research.
Unit Two provides an in-depth presentation of the
research process for both quantitative and qualitative
research. As with previous editions, this text provides
extensive coverage of the many types of quantitative
and qualitative research.
Unit Three addresses the implications of research
for the discipline and profession of nursing. Content
is provided to direct the student in conducting critical
appraisals of both quantitative and qualitative research.
A detailed discussion of types of research synthesis
and strategies for promoting EBP is provided.
Unit Four gives students and practicing nurses the
content they need for implementing studies. This unit
includes chapters focused on data collection, statisti-
cal analysis, interpretation of research outcomes, and
dissemination of research finding.
Unit Five addresses proposal development and
seeking support for research. Readers are given

direction for developing quantitative and qualitative
research proposals and seeking funding for their
research.
The changes in the seventh edition of this text
reflect the advances in nursing research and also incor-
porate comments from outside reviewers, colleagues,
and students. Our desire to promote the continuing
development of the profession of nursing was the
incentive for investing the time and energy required
to develop this new edition.
New Content
The seventh edition provides current comprehensive
coverage of nursing research and is focused on the
learning needs and styles of today’s nursing students
and practicing nurses. Several exciting new areas of
content based on the changes and expansion in the
field of nursing research are included in this edition.
Some of the major changes from the previous edition
are as follows:
• Chapter 1, “Discovering the World of Nursing
Research,” is a strong introduction to evidence-
based practice (EBP) that is linked to nursing
research using a revised framework model for this
edition of the text.
• Chapter 2, “Evolution of Research in Building Evi-
dence-Based Nursing Practice,” has a new title
and is focused on building an EBP for nursing.
This chapter introduces the most current processes
for synthesizing research knowledge, which are
systematic reviews, meta-analyses, meta-syntheses,
and mixed-method systematic reviews. The chapter
includes a table that presents the purposes of these
syntheses, the types of research they include (the
“sampling frame”), and the analysis for achieving
the different types of syntheses. A model of the
continuum of the levels of research evidence, from
strongest to weakest evidence, is provided.
• Chapter 4, “Introduction to Qualitative Research,”
describes the philosophical perspectives that guide
the following five approaches to qualitative re-
search: (1) phenomenology, (2) grounded theory,
(3) ethnography, (4) exploratory-descriptive quali-
tative research, and (5) historical research. Excerpts
from qualitative studies are provided to emphasize
the contributions researchers using each approach
have made to nursing science.
• Chapter 6, “Review of Relevant Literature,” pro-
vides current, comprehensive strategies for search-
ing the literature to identify relevant sources.
• Chapter 9, “Ethics in Research,” features updated
coverage of (1) the Health Insurance Portability
Preface ix
and Accountability Act (HIPAA), (2) U.S. Depart-
ment of Health and Human Services (DHHS) regu-
lations for protection of human subjects in research,
and (3) U.S. Food and Drug Administration (FDA)
regulations for protection of research subjects. This
chapter also details the escalating problem of
research misconduct in all healthcare disciplines
and the actions that have been taken to manage this
problem.
• Chapter 10, “Understanding Quantitative Research
Designs,” provides new content on mixed-methods
designs that include both quantitative and qual­
itative research methods. Four common mixed-
method research strategies conducted in nursing
are discussed: sequential explanatory strategy,
sequential exploratory strategy, sequential transfor-
mative strategy, and concurrent triangulation strat-
egy. These strategies are presented using models,
narrative descriptions, and examples.
• Chapter 11, “Selecting a Quantitative Research
Design,” describes many currently used designs
that are not covered in other leading texts but that
are important to the generation of nursing knowl-
edge. It contains a detailed discussion of random-
ized controlled trials (RCTs) along with the
Consolidated Standards for Reporting Trials
(CONSORT, 2010) guidelines.
• Chapter 12, “Qualitative Research Methodology,”
is completely reorganized to address each step of
the research process from writing the problem
statement to interpreting the findings for qualitative
studies. The data collection methods of observing,
interviewing, and conducting focus groups are
described in depth. In addition, examples of using
photovoice, videos, and electronic communication
are given. Methods specific to each philosophical
approach are also discussed.
• Chapter 13, “Outcomes Research,” a unique feature
of our text, was significantly rewritten to promote
understanding of the history, significance, and
impact of outcomes research on nursing and health
care, for both students and nurses in clinical prac-
tice. New content is included on nurse-sensitive
patient outcomes, advanced-practice nursing out-
comes, and databases used in conducting outcomes
research. In addition, the methodologies for con-
ducting outcomes research have been updated and
expanded. This chapter was revised by a leading
authority in the conduct of outcomes research,
Dr. Diane Doran.
• Chapter 14, “Intervention-Based Research,” was
extensively rewritten to focus on the conduct of
intervention-based research. It offers students and
x Preface
practicing nurses detailed, current content and
guidelines for critically appraising and conducting
intervention studies. The chapter was revised by
Dr. Rosemary Polomano, an authority in the
conduct of intervention research.
• Chapter 15, “Sampling,” contains extensive cover-
age of current sampling methods and the processes
for determining sample size for quantitative and
qualitative studies. This chapter includes formulas
for calculating the acceptance and refusal rates for
potential study participants and the retention and
attrition rates for subjects participating in a study.
Additional current content is provided to assist
researchers in determining sample size for quanti-
tative and qualitative research and for recruiting
and retaining subjects for their studies.
• Chapter 16, “Measurement Concepts,” features
detailed information for examining the reliability
and validity of measurement methods and the pre-
cision and accuracy of physiological measures used
in nursing studies. Students are provided a back-
ground for understanding sensitivity, specificity,
and likelihood ratios used to determine the quality
of diagnostic tests.
• Chapter 17, “Measurement Methods Used in
Developing Evidence for Practice,” provides more
detail on the use of physiological measurement
methods in research. A growing number of nursing
studies are focused on the measurement of the
outcomes from interventions using physiological
measurement methods, and this chapter equips
the reader to understand and participate in these
studies.
• Chapter 18, “Critical Appraisal of Nursing Studies,”
has a more refined process for critically appraising
quantitative studies that consists of the following
steps: (1) identifying the steps of the research
process, (2) determining the study strengths and
weaknesses, and (3) evaluating the credibility and
meaning of a study for nursing knowledge and prac-
tice. The process of critically appraising qualitative
studies was revised to evaluate studies using the
standards of philosophical congruence, method-
ological coherence, intuitive comprehension, and
intellectual contribution.
• Chapter 19, “Evidence Synthesis with Strategies
for Promoting Evidence-Based Practice,” has
undergone extensive revision to achieve a com-
pletely new focus on how to conduct research
syntheses and use the best research evidence
in practice. The chapter now contains extensive
details for conducting systematic reviews, meta-
analyses, meta-syntheses, and mixed-method
systematic reviews. Guidelines are also provided
to direct students in evaluating these research syn-
theses, which are appearing more frequently in the
nursing and healthcare literature. Current informa-
tion is given on the activities of Evidence-Based
Practice Centers and the new initiative for funding
translation research through the National Institutes
of Health to increase the implementation of evi-
dence-based interventions in practice.
• Chapter 20, “Collecting and Managing Data,” now
covers practical aspects of developing a data collec-
tion plan, including formatting instruments, creat-
ing a data flow chart, and training data collectors. In
addition, common problems that occur during data
collection are described, with possible solutions.
• Major revisions have been made in the chapters
focused on statistical concepts and analysis tech-
niques (Chapters 21 through 25). The content is
presented in a clear, concise manner and supported
with examples of analyses conducted on actual
clinical data. Dr. Daisha Cipher, a noted statistician
and healthcare researcher, assisted with the revi-
sion of these chapters.
• Chapter 27, “Disseminating Research Findings,”
features expanded and updated content on com-
municating study findings through oral and poster
presentations and publications.
• Chapter 29, “Seeking Funding for Research,” pro-
vides current strategies to assist students and prac-
ticing nurses in obtaining funding for their studies.
Student Ancillaries
An Evolve Resources website, which is available at
http://evolve.elsevier.com/Grove/practice/, features a
wealth of assets, including the following:
• Interactive Review Questions
• Data Sets and Data Set Activities
• Sample Research Proposals
An electronic Study Guide accompanies this
edition of The Practice of Nursing Research. This
study guide is keyed chapter-by-chapter to the text. It
includes the following:
• Relevant Terms activities that help students under-
stand and apply the language of nursing research
• Key Ideas exercises that reinforce essential
concepts
• Making Connections activities that give students
practice in the higher-level skills of comprehension
and content synthesis
• Crossword Puzzles that serve not only as a clever
learning activity but also as a welcome “fun” activ-
ity for busy adult learners

• Exercises in Critical Appraisal that provide expe­
riences for students and practicing nurses to cri­
tically evaluate both quantitative and qualitative
studies
• Going Beyond activities that provide suggestions
for further study
• An Answer Key is provided at the end of each
chapter that offers immediate feedback to reinforce
learning
• A Published Studies appendix is provided for the
critical appraisal exercises in the study guide, and
other current studies are included on the Evolve
website for faculty to use in providing learning
experiences for their students.
Preface xi
Instructor Ancillaries
The Instructor Resources are available on Evolve, at
http://evolve.elsevier.com/Grove/practice/. Instructors
also have access to the online student resources. The
Instructor Resources are an Instructor’s Manual, an
expanded Test Bank including 600 questions, Power-
Point Presentations totaling more than 700 slides, and
an Image Collection consisting of most images from
the text.
 Acknowledgments
W
riting the seventh edition of this textbook
has allowed us the opportunity to examine
and revise the content of the previous
edition based on input from a number of scholarly
colleagues, the literature, and our graduate and under-
graduate students. A textbook such as this requires
synthesizing the ideas of many people and resources.
For the first time, expert contributors have revised key
chapters of this textbook. These experts have added
invaluable content in critical areas of outcomes
research, intervention research, design, data collec-
tion, and statistics. We thank these scholars for sharing
their expertise.
We have also attempted to extract from the nursing
and healthcare literature the essence of knowledge
related to the conduct of nursing research. Thus we
would like to thank those scholars who shared their
knowledge with the rest of us in nursing and who have
made this knowledge accessible for inclusion in this
textbook. The ideas from the literature were synthe-
sized and discussed with our colleagues and students
to determine the revisions needed for the seventh
edition.
We would also like to express our appreciation to
Dean Elizabeth Poster and faculty members of the
Susan K. Grove,
PhD, RN, ANP-BC, GNP-BC
xii
College of Nursing at The University of Texas at
Arlington, for their support during the long and some-
times arduous experiences that are inevitable in devel-
oping a book of this magnitude. We would also like
to thank Dr. Julie Barroso for her suggestions regard-
ing the qualitative research content in this text. We
particularly value the questions raised by our students
regarding the content of this text, which allow us a
unique view of our learners’ perceptions.
We would also like to recognize the excellent
reviews of the colleagues who helped us make impor-
tant revisions in this text. These reviewers are located
in large and small universities across the United States
and provided a broad range of research expertise.
Finally, we thank the people at Elsevier, who have
been extremely helpful to us in producing a scholarly,
attractive, appealing text. We extend a special thank
you to the people most instrumental in the develop-
ment and production of this book: Lee Henderson,
Executive Content Strategist; and Julia Curcio, Asso-
ciate Content Development Specialist. We also want
to thank others involved with the production and mar-
keting of this book—Antony Prince, Project Manager;
Karen Pauls, Designer; and Pat Crowe, Marketing
Manager.
Nancy Burns,
PhD, RN, FAAN
Jennifer Gray,
PhD, RN

Unit One
Introduction to Nursing
Research, 1
  1 Discovering the World of Nursing
Research, 1  
Definition of Nursing Research, 1
Framework Linking Nursing Research to the World of
Nursing, 2
Significance of Research in Building an Evidence-
Based Practice for Nursing, 11
  2 Evolution of Research in Building Evidence-
Based Nursing Practice, 17
Historical Development of Research in Nursing, 17
 
Methodologies for Developing Research Evidence in
Nursing, 23
Classification of Research Methodologies Presented
in this Text, 25
Introduction to Best Research Evidence for Practice, 28
  3 Introduction to Quantitative Research, 34
Concepts Relevant to Quantitative Research, 34
Steps of the Quantitative Research Process, 38
 
Selecting a Research Design, 43
Types of Quantitative Research, 49
  4 Introduction to Qualitative Research, 57
Perspective of the Qualitative Researcher, 57
Approaches to Qualitative Research, 59
Unit Two
The Research Process, 73
  5 Research Problem and Purpose, 73  
What Is a Research Problem and Purpose?, 73
Sources of Research Problems, 75
Formulating a Research Problem and Purpose, 81
Contents
Example of Problem and Purpose Development, 83
Feasibility of a Study, 84
Example Research Topics, Problems, and Purposes
for Different Types of Research, 87
6 Review of Relevant Literature, 97
What Is “The Literature”?, 97
What Is a Literature Review?, 97
Purposes of Reviewing the Literature, 98
Practical Considerations, 100
Stages of a Literature Review, 102
Processing the Literature, 108
Writing the Review of Literature, 110
Example of a Literature Review, 112
7 Frameworks, 116
Definition of Terms, 116
Understanding Concepts, 117
Examining Relational Statements, 120
Grand Theories, 125
Application of Middle-Range Theories, 125
Appraising Theories and Research Frameworks, 127
Developing a Research Framework for Study, 130
8 Objectives, Questions, Hypotheses,
and Study Variables, 138
Formulating Research Objectives or Aims, 138
Formulating Research Questions, 140
Formulating Hypotheses, 142
Selecting Objectives, Questions, or Hypotheses for
Quantitative or Qualitative Research, 150
Identifying and Defining Study Variables, 150
Operationalizing Variables or Concepts for a
Study, 155
9 Ethics in Research, 159
Historical Events Affecting the Development of
Ethical Codes and Regulations, 159
xiii
xiv Contents
Protection of Human Rights, 163
Balancing Benefits and Risks for a Study, 175
Obtaining Informed Consent, 176
Institutional Review, 183
Research Misconduct, 187
Animals as Research Subjects, 190
14 Intervention-Based Research, 323
10 Understanding Quantitative Research
Design, 195
Concepts Important to Design, 195
Study Validity, 197
Elements of a Good Design, 202
Questions to Direct Design Development and
Implementation in a Study, 207
Mixed Methods, 208
11 Selecting a Quantitative Research
Design, 214
Descriptive Study Designs, 215
Surveys, 224
Correlational Study Designs, 224
15 Sampling, 351
Defining Therapeutic Nursing Interventions, 228
Quasi-experimental Study Designs, 231
Experimental Study Designs, 244
Studies That Do Not Use Traditional Research
Designs, 254
Algorithms for Selecting Research Designs, 256
12 Qualitative Research Methodology, 264
Clinical Context and Research Problems, 264
Literature Review for Qualitative Studies, 265
Theoretical Frameworks, 265
Purposes, 266
Research Objectives or Questions, 266
16 Measurement Concepts, 382
Obtaining Research Participants, 267
Data Collection Methods, 268
Electronically Mediated Data, 276
Transcribing Recorded Data, 278
Data Management, 279
Data Analysis, 279
Methods Specific to Qualitative Approaches, 284
13 Outcomes Research, 294
Theoretical Basis of Outcomes Research, 294
Evaluating Structure, 298
Federal Government Involvement in Outcomes
Research, 300
Outcomes Research and Nursing Practice, 302
Methodologies for Outcomes Studies, 306
Disseminating Outcomes Research Findings, 318
Intervention-Based Research Conducted by
Nurses, 323
Nursing Interventions, 324
Programs of Nursing Intervention Research, 326
Terminology for Intervention-Based Research, 329
Types of Research Designs, 335
Planning Intervention Research, 338
Design and Testing of Interventions, 339
Process of Testing the Intervention, 342
Data Collection, 345
Threats to Study Validity, 345
Critical Appraisal of Intervention-Based
Research, 345
Sampling Theory, 351
Probability (Random) Sampling Methods, 357
Nonprobability (Nonrandom) Sampling Methods
Commonly Applied in Quantitative Research, 362
Nonprobability Sampling Methods Commonly
Applied in Qualitative Research, 364
Sample Size in Quantitative Research, 367
Sample Size in Qualitative Research, 371
Research Settings, 373
Recruiting and Retaining Research
Participants, 374
Directness of Measurement, 382
Measurement Error, 383
Levels of Measurement, 385
Reference Testing of Measurement, 388
Reliability, 389
Validity, 393
Accuracy, Precision, and Error of Physiological
Measures, 402
Sensitivity, Specificity, and Likelihood Ratios, 406
 Contents xv

17 Measurement Methods Used in Developing

Evidence-Based Practice, 411
Unit Four
Analyzing Data, Determining
Physiological Measurement, 411
Observational Measurement, 421
Outcomes, and Disseminating
Interviews, 422 Research, 507
Questionnaires, 425 20 Collecting and Managing Data, 507
Scales, 429
Data Collection Modes, 507
Q-Sort Methodology, 434
Factors Influencing Data Collection, 515
Delphi Technique, 435
Data Collection and Coding Plan, 517
Diaries, 437
Pilot Study, 523
Measurement Using Existing Databases, 439
Collecting Data, 523
Selection of an Existing Instrument, 440
Serendipity, 530
Constructing Scales, 442
Having Access to Support Systems, 530
Translating a Scale to Another Language, 445
Managing Data, 531

Unit Three 21 Introduction to Statistical Analysis, 534

Putting It All Together
for Evidence-Based Health
Care, 451
18 Critical Appraisal of Nursing
Studies, 451
Evolution of Critical Appraisal of Research in
Nursing, 451
Nurses’ Expertise in Critical Appraisal of
Research, 454
Critical Appraisal Process for Quantitative
Research, 454
Critical Appraisal Process for Qualitative
Studies, 462
19 Evidence Synthesis and Strategies
for Evidence-Based Practice, 468
Benefits and Barriers Related to Evidence-Based
Nursing Practice, 468
Guidelines for Synthesizing Research Evidence, 471
Models to Promote Evidence-Based Practice in
Nursing, 493
Implementing Evidence-Based Guidelines in
Practice, 496
Evidence-Based Practice Centers, 502
Introduction to Translational Research, 503
Concepts of Statistical Theory, 534
Practical Aspects of Data Analysis, 542
Choosing Appropriate Statistical Procedures for a
Study, 546
22 Using Statistics to Describe Variables, 550
Using Statistics to Summarize Data, 550
Using Statistics to Explore Deviations in the
Data, 554
23 Using Statistics to Examine
Relationships, 560
Scatter Diagrams, 560
Bivariate Correlational Analysis, 560
Bland and Altman Plots, 565
Factor Analysis and Principal Components
Analysis, 566
24 Using Statistics to Predict, 570
Simple Linear Regression, 570
Calculation, 571
Multiple Regression, 573
Odds Ratio, 575
Logistic Regression, 576
Cox Proportional Hazards Regression, 577
xvi Contents

25 Using Statistics to Determine Seeking Approval for a Study, 644

Differences, 580
Choosing Parametric versus Nonparametric Statistics
to Determine Differences, 580
t-tests, 580
One-Way Analysis of Variance, 584
Chi-Square Test of Independence, 587
26 Interpreting Research Outcomes, 590
Examining Evidence, 590
Determining Findings, 595
Forming Conclusions, 596
Identifying Limitations, 598
Generalizing the Findings, 598
Considering Implications, 599
Recommending Further Research, 599
27 Disseminating Research Findings, 602
Content of a Research Report, 602
Audiences for Communication of Research
Findings, 619
Presenting Research Findings, 621
Publishing Research Findings, 627
Example of a Quantitative Research
Proposal, 647
29 Seeking Funding for Research, 663
Building a Program of Research, 663
Getting Started, 665
Identifying Funding Sources, 667
Submitting a Proposal for a Federal Grant, 670
Grant Management, 671
Planning Your Next Grant, 672
Appendices
Appendix A. Z Values Table, 674
Appendix B. Critical Values for Student’s t
Distribution, 677
Appendix C. Critical Values of r for Pearson Product
Moment Correlation Coefficient, 679
Appendix D. Critical Values of F for α = 0.05 and
α = 0.01, 680
Appendix E. Critical Values of the χ2 Distribution, 683
Appendix F. Statistical Power Tables
(Δ = Effect Size), 684

Glossary, 686
Unit Five
Proposing and Seeking Funding
for Research, 635 Index, 714

28 Writing Research Proposals, 635

Writing a Research Proposal, 635
Content of a Research Proposal, 636
UNIT ONE
Introduction to Nursing Research
Discovering the World
of Nursing Research
  http://evolve.elsevier.com/Grove/practice/
W
elcome to the world of nursing research. You
might think it is strange to consider research
a “world,” but research is truly a new way of
experiencing reality. Entering a new world requires
learning a unique language, incorporating new rules,
and using new experiences to learn how to interact
effectively within that world. As you become a part of
this new world, your perceptions and methods of rea-
soning will be modified and expanded. Understanding
the world of nursing research is critical to providing
evidence-based care to your patients. Since the 1990s,
there has been a growing emphasis for nurses—
especially advanced practice nurses (nurse practi­
tioners, clinical nurse specialists, nurse anesthetists,
and nurse midwives), administrators, educators, and
nurse researchers—to promote an evidence-based
practice in nursing (Brown, 2009; Craig & Smyth,
2012; Melnyk & Fineout-Overholt, 2011). Evidence-
based practice in nursing requires a strong body of
research knowledge that nurses must synthesize and
use to promote quality care for their patients, families,
and communities. We developed this text to facilitate
your understanding of nursing research and its con­
tribution to the implementation of evidenced-based
nursing practice.
This chapter broadly explains the world of research.
A definition of nursing research is provided followed
by the framework for this textbook that connects
nursing research to the world of nursing. The chapter
concludes with a discussion of the significance of
research in developing an evidence-based practice for
nursing.
Definition of Nursing Research
The root meaning of the word research is “search
again” or “examine carefully.” More specifically,
1
CHAPTER
research is the diligent, systematic inquiry or investi-
gation to validate and refine existing knowledge and
generate new knowledge. The concepts systematic and
diligent are critical to the meaning of research because
they imply planning, organization, and persistence.
Many disciplines conduct research, so what distin-
guishes nursing research from research in other disci-
plines? In some ways, there are no differences, because
the knowledge and skills required to conduct research
are similar from one discipline to another. However,
when one looks at other dimensions of research within
a discipline, it is clear that research in nursing must
be unique to address the questions relevant to the
profession. Nurse researchers need to implement the
most effective research to develop a unique body of
knowledge for nursing.
The American Nurses Association (ANA, 2012)
developed the following definition of nursing that
identifies the unique body of knowledge needed by the
profession: “Nursing is the protection, promotion, and
optimization of health and abilities, prevention of
illness and injury, alleviation of suffering through the
diagnosis and treatment of human response, and advo-
cacy in the care of individuals, families, communities,
and populations.” On the basis of this definition,
nursing research is needed to generate knowledge
about human responses and the best interventions to
promote health, prevent illness, and manage illness
(ANA, 2010b).
Many nurses hold the view that nursing research
should focus on acquiring knowledge that can be
directly implemented in clinical practice, which is
sometimes referred to as applied research or practical
research (Brown, 2009; Mackay, 2009). However,
another view is that nursing research should include
studies of nursing education, nursing administration,
health services, and nurses’ characteristics and roles
1
2 UNIT ONE  Introduction to Nursing Research
as well as clinical situations. Riley, Beal, Levi, and
McCausland (2002) support this second view and
believe nursing scholarship should include education,
practice, and service. Research is needed to identify
teaching-learning strategies to promote nurses’ man-
agement of practice. Thus, nurse researchers are
involved in building a science for nursing education
so the teaching-learning strategies used are evidence-
based (National League for Nursing [NLN], 2009).
Nurse administrators are involved in research to
enhance nursing leadership and the delivery of quality,
cost-effective patient care. Studies of health services
and nursing roles are important to promote quality
outcomes in the nursing profession and the healthcare
system (Doran, 2011).
Thus, the knowledge generated through nursing
research provides the scientific foundation essential
for all areas of nursing. In this text, nursing research
is defined as a scientific process that validates and
refines existing knowledge and generates new knowl-
edge that directly and indirectly influences the deliv-
ery of evidence-based nursing.
Framework Linking
Nursing Research to
the World of Nursing
To best explore nursing research, we have developed
a framework to help establish connections between
research and the various elements of nursing. The
framework presented in the following pages links
nursing research to the world of nursing and is used
as an organizing model for this textbook. In the
framework model (see Figure 1-1), nursing research
is not an entity disconnected from the rest of nursing
but rather is influenced by and influences all other
nursing elements. The concepts in this model are pic-
tured on a continuum from concrete to abstract. The
discussion introduces this continuum and progresses
from the concrete concept of the empirical world of
nursing practice to the most abstract concept of
nursing philosophy. The use of two-way arrows in the
model indicates the dynamic interaction among the
concepts.
Concrete-Abstract Continuum
As previously mentioned, Figure 1-1 presents the
components of nursing on a concrete-abstract contin-
uum. This continuum demonstrates that nursing
thought flows both from concrete to abstract thinking
and from abstract to concrete. Concrete thinking is
oriented toward and limited by tangible things or by
Abstract
Philosophy
Knowledge
Science
Theory Research
Abstract
Thought
Processes
Empirical
World
(Evidence-
based nursing
practice)
Concrete
Figure 1-1  Framework linking nursing research to the world of
nursing.
events that we observe and experience in reality. Thus,
the focus of concrete thinking is immediate events that
are limited by time and space. Most nurses believe
they are concrete thinkers because they focus on the
specific actions in nursing practice. Abstract think-
ing is oriented toward the development of an idea
without application to, or association with, a particular
instance. Abstract thinkers tend to look for meaning,
patterns, relationships, and philosophical implica-
tions. This type of thinking is independent of time and
space. Currently, graduate nursing education fosters
abstract thinking, because it is an essential skill for
developing theory and creating an idea for study.
Nurses assuming advanced roles and registered nurses
(RNs) need to use both abstract and concrete thinking.
For example, a nurse practitioner must explore the
best research evidence about a practice problem
(abstract thinking) before using his or her clinical
expertise to diagnose and manage an individual
patient’s health problem (concrete thinking). RNs also
use abstract and concrete thinking to develop and
refine protocols and policies based on current research
to direct patient care.
Nursing research requires skills in both concrete
and abstract thinking. Abstract thought is required to
 CHAPTER 1  Discovering the World of Nursing Research 3
identify researchable problems, design studies, and
interpret findings. Concrete thought is necessary in
both planning and implementing the detailed steps of
data collection and analysis. This back-and-forth flow
between abstract and concrete thought may be one
reason why nursing research seems complex and
challenging.
Empirical World
The empirical world is what we experience through
our senses and is the concrete portion of our existence.
It is what we often call reality, and “doing” kinds of
activities are part of this world. There is a sense of
certainty about the empirical or real world; it seems
understandable, predictable, controllable. Concrete
thinking focuses on the empirical world; words associ-
ated with this thinking include “practical,” “down-to-
earth,” “solid,” and “factual.” Concrete thinkers want
facts. They want to be able to apply whatever they
know to the current situation.
The practice of nursing takes place in the empirical
world, as demonstrated in Figure 1-1. The scope of
nursing practice varies for the RN and the advanced
practice nurse (APN). RNs provide care to and coor-
dinate care for patients, families, and communities
in a variety of settings. They initiate interventions
as well as carry out treatments authorized by other
healthcare providers (ANA, 2010a). APNs, such
as nurse practitioners, nurse anesthetists, nurse mid-
wives, and clinical nurse specialists, have an expanded
practice. Their knowledge, skills, and expertise pro­
mote role autonomy and overlap with medical prac-
tice. APNs usually concentrate their clinical practice
in a specialty area, such as acute care, pediatrics, ger-
ontology, adult or family primary care, psychiatric-
mental health, women’s health, maternal child, or
anesthesia (ANA, 2010b). You can access the most
current nursing scope and standards for practice from
the ANA (2010a). Within the empirical world of
nursing, the goal is to provide evidence-based practice
to improve the health outcomes of individuals, fami-
lies, and communities (see Figure 1-1). The aspects
of evidence-based practice and the significance of
research in developing evidence-based practice are
covered later in this chapter.
Reality Testing Using Research
People tend to validate or test the reality of their
existence through their senses. In everyday activities,
they constantly check out the messages received from
their senses. For example, they might ask, “Am I
really seeing what I think I am seeing?” Sometimes
their senses can play tricks on them. This is why
instruments have been developed to record sensory
experiences more accurately. For example, does the
patient just feel hot or actually have a fever? Ther-
mometers were developed to test this sensory percep-
tion accurately. Through research, the most accurate
and precise measurement devices have been devel-
oped to assess the temperature of patients on the
basis of age and health status (Waltz, Strickland, &
Lenz, 2010). Thus, research is a way to test reality
and generate the best evidence to guide nursing
practice.
Nurses use a variety of research methods to
test their reality and generate nursing knowledge,
including quantitative research, qualitative research,
outcomes research, and intervention research. Quan-
titative research, the most frequently conducted
method, is a formal, objective, systematic meth­
odology to describe variables, test relationships, and
examine cause-and-effect interactions (Kerlinger &
Lee, 2000; Shadish, Cook, & Campbell, 2002). Since
the 1980s, nurses have been conducting qualitative
research to generate essential theories and knowledge
for nursing. Qualitative research is a rigorous, inter-
active, holistic, subjective research approach used to
describe life experiences and give them meaning
(Marshall & Rossman, 2011; Munhall, 2012). Both
quantitative and qualitative research methods are
important to the development of nursing knowledge
(Fawcett & Garity, 2009; Munhall, 2012; Shadish et
al., 2002). Some researchers effectively combine these
two methods in implementing mixed method research
to address selected nursing research problems (Cre-
swell, 2009).
Medicine, healthcare agencies, and now nursing
are focusing on the outcomes of patient care. Out-
comes research is an important scientific methodol-
ogy that has evolved to examine the end results of
patient care and the outcomes for healthcare providers,
such as RNs, APNs, and physicians, and for healthcare
agencies (Doran, 2011). Nurses are also engaged in
intervention research, a methodology for investigat-
ing the effectiveness of nursing interventions in
achieving the desired outcomes in natural settings
(Forbes, 2009). These different types of research are
all essential to the development of nursing science,
theory, and knowledge (see Figure 1-1). Nurses have
varying roles related to research that include conduct-
ing research, critically appraising research, and using
research evidence in practice.
Roles of Nurses in Research
Generating a scientific knowledge base with imple-
mentation in practice requires the participation of all
4 UNIT ONE  Introduction to Nursing Research

TABLE 1-1  Nurses’ Participation in Research at Various Levels of Education

Educational Preparation Research Functions
BSN Read and critically appraise studies. Use best research evidence in practice with
guidance. Assist with problem identification and data collection.
MSN Critically appraise and synthesize studies to develop and revise protocols, algorithms,
and policies for practice. Implement best research evidence in practice.
Collaborate in research projects and provide clinical expertise for research.
DNP Participate in evidence-based guideline development. Develop, implement, evaluate,
and revise as needed protocols, policies, and evidence-based guidelines in practice.
Conduct clinical studies, usually in collaboration with other nurse researchers.
PhD Major role in conducting independent research and contributing to the empirical
knowledge generated in a selected area of study. Obtain initial funding for
research. Coordinate research teams of BSN, MSN, and DNP nurses.
Post-doctorate Assume a full researcher role with a funded program of research. Lead and/or
participate in nursing and interdisciplinary research teams. Identified as experts in
their areas of research. Mentor PhD-prepared researchers.

nurses in a variety of research activities. Some nurses
are developers of research and conduct studies to gen-
erate and refine the knowledge needed for nursing
practice. Others are consumers of research and use
research evidence to improve their nursing practice.
The American Association of Colleges of Nursing
(AACN, 2006) and ANA (2010a, 2010b) have pub-
lished statements about the roles of nurses in research.
No matter their education or position, all nurses have
roles in research and some ideas about those roles are
presented in Table 1-1. The research role a nurse
assumes usually expands with his or her advanced
education, expertise, and career path. Nurses with a
Bachelor of Science in Nursing (BSN) degree have
knowledge of the research process and skills in reading
and critically appraising studies. They assist with the
implementation of evidence-based guidelines, proto-
cols, algorithms, and policies in practice. In addition,
these nurses might provide valuable assistance in
identifying research problems and collecting data for
studies.
Nurses with a Master of Science in Nursing (MSN)
have undergone the educational preparation to criti-
cally appraise and synthesize findings from studies to
revise or develop protocols, algorithms, or policies for
use in practice. They also have the ability to identify
and critically appraise the quality of evidence-based
guidelines developed by national organizations. APNs
and nurse administrators have the ability to lead
healthcare teams in making essential changes in
nursing practice and in the healthcare system on the
basis of current research evidence. Some MSN-
prepared nurses conduct studies but usually do so in
collaboration with other nurse scientists (AACN,
2006; ANA 2010a).
The doctorate in nursing can be practice focused
(doctorate of nursing practice [DNP]) or research
focused (doctorate of philosophy [PhD]). Nurses with
DNPs are educated to have the highest level of clinical
expertise, with the ability to translate scientific knowl-
edge for use in practice. These doctorally prepared
nurses have advanced research and leadership knowl-
edge to develop, implement, evaluate, and revise
evidence-based guidelines, protocols, algorithms, and
policies for practice (Clinton & Sperhac, 2006). In
addition, DNP-prepared nurses have the expertise to
conduct and/or collaborate with clinical studies.
PhD-prepared nurses assume a major role in the
conduct of research and the generation of nursing
knowledge in a selected area of interest (Brar, Boschma,
& McCuaig, 2010). These nurse scientists often coor-
dinate research teams that include DNP-, MSN-, and
BSN-prepared nurses to facilitate the conduct of
high-quality studies in a variety of healthcare agencies.
Postdoctorate nurses usually assume full-time re-
searcher roles and have funded programs of research.
They lead interdisciplinary teams of researchers and
sometimes conduct studies in multiple settings. These
scientists often are identified as experts in selected
areas of research and provide mentoring of new PhD-
prepared researchers (AACN, 2006) (see Table 1-1).
Abstract Thought Processes
Abstract thought processes influence every element
of the nursing world. In a sense, they link all the ele-
ments together. Without skills in abstract thought, we
are trapped in a flat existence; we can experience the
empirical world, we cannot explain or understand it
(Abbott, 1952). Through abstract thinking, however,
we can test our theories (which explain the nursing
 CHAPTER 1  Discovering the World of Nursing Research 5
world) and then include them in the body of scientific
knowledge (Smith & Liehr, 2008). Abstract thinking
also allows scientific findings to be developed into
theories (Munhall, 2012). Abstract thought enables
both science and theories to be blended into a cohesive
body of knowledge, guided by a philosophical frame-
work, and applied in clinical practice (see Figure 1-1).
Thus, abstract thought processes are essential for syn-
thesizing research evidence and knowing when and
how to use this knowledge in practice.
Three major abstract thought processes—
introspection, intuition, and reasoning—are important
in nursing (Silva, 1977). These thought processes are
used in critically appraising and applying best research
evidence in practice, planning and implementing
research, and developing and evaluating theory.
Introspection
Introspection is the process of turning your attention
inward toward your own thoughts. It occurs at two
levels. At the more superficial level, you are aware of
the thoughts you are experiencing. You have a greater
awareness of the flow and interplay of feelings and
ideas that occur in constantly changing patterns. These
thoughts or ideas can rapidly fade from view and
disappear if you do not quickly write them down.
When you allow introspection to occur in more depth,
you examine your thoughts more critically and in
detail. Patterns or links between thoughts and ideas
emerge, and you may recognize fallacies or weak-
nesses in your thinking. You may question what
brought you to this point and find yourself really
enjoying the experience.
Imagine the following clinical situation. You have
just left John Brown’s home. John has a colostomy
and has been receiving home health care for several
weeks. Although John is caring for his colostomy, he
is still reluctant to leave home for any length of time.
You are irritated and frustrated with this situation. You
begin to review your nursing actions and to recall
other patients who reacted in similar ways. What were
the patterns of their behavior?
You have an idea: Perhaps the patient’s behavior is
linked to the level of family support. You feel unsure
about your ability to help the patient and family deal
with this situation effectively. You recall other nurses
describing similar reactions in their patients, and you
wonder how many patients with colostomies have this
problem. Your thoughts jump to reviewing the charts
of other patients with colostomies and reading rele-
vant ideas discussed in the literature. Some research
has been conducted on this topic recently, and you
could critically appraise these findings to determine
the level of evidence for possible use of the ideas in
practice. If the findings are inadequate, perhaps other
nurses would be interested in studying this situation
with you.
Intuition
Intuition is an insight into or understanding of a situ-
ation or event as a whole that usually cannot be logi-
cally explained (Smith, 2009). Because intuition is a
type of knowing that seems to come unbidden, it may
also be described as a “gut feeling” or a “hunch.”
Because intuition cannot be explained with ease sci-
entifically, many people are uncomfortable with it.
Some even say that it does not exist. Sometimes,
therefore, the feeling or sense is suppressed, ignored,
or dismissed as silly. However, intuition is not the lack
of knowing; rather, it is a result of deep knowledge—
tacit knowing or personal knowledge (Benner, 1984;
Polanyi, 1962, 1966). The knowledge is incorporated
so deeply within that it is difficult to bring it con-
sciously to the surface and express it in a logical
manner. One of the most commonly cited examples of
nurses’ intuition is their recognition of a patient’s
physically deteriorating condition. Odell, Victor, and
Oliver (2009) conducted a review of the research lit-
erature and described nurses’ use of intuition in clini-
cal practice. They noted that nurses have an intuition
or a knowing that something is not right with their
patients by recognizing changes in behavior and phys-
ical signs. Through clinical experience and the use of
intuition, nurses are able to recognize patterns of devi-
ations from the normal clinical course and to know
when to take action.
Intuition is generally considered unscientific and
unacceptable for use in research. In some instances,
that consideration is valid. For example, a hunch about
significant differences between one set of scores and
another set of scores is not particularly useful as an
analysis technique. However, even though intuition is
often unexplainable, it has some important scientific
uses. Researchers do not always need to be able to
explain something in order to use it. A burst of intu-
ition may identify a problem for study, indicate impor-
tant variables to measure, or link two ideas together
in interpreting the findings. The trick is to recognize
the feeling, value it, and hang on to the idea long
enough to consider it. Some researchers keep a journal
to capture elusive thoughts and hunches as they think
about their phenomenon of interest. These intuitive
hunches often become important later as they conduct
their studies.
Imagine the following situation. You have been
working in an oncology center for the past 3 years.
6 UNIT ONE  Introduction to Nursing Research
You and two other nurses working in the center have
been meeting with the acute care nurse practitioner to
plan a study to determine which factors are important
for promoting positive patient outcomes in the center.
The group has met several times with a nursing profes-
sor at the university, who is collaborating with the
group to develop the study. At present, the group is
concerned with identifying the outcomes that need to
be measured and how to measure them.
You have had a busy morning. Mr. Green, a patient,
stops by to chat on his way out of the clinic. You listen,
but not attentively at first. You then become more
acutely aware of what he is saying and begin to have
a feeling about one variable that should be studied.
Although he didn’t specifically mention fear of break-
ing the news about having cancer to his children, you
sense that he is anxious about conveying bad news to
his loved ones. You cannot really explain the origin of
this feeling, something in the flow of Mr. Green’s
words has stimulated a burst of intuition. You suspect
other patients diagnosed with cancer face similar fear
and hesitation about informing their family members
of bad news, that they have cancer or that their cancer
has spread. You believe the variable “fear of breaking
bad news to loved ones” needs to be studied. You feel
both excited and uncertain. What will the other nurses
think? If the variable has not been studied, is it really
significant? Somehow, you feel that it is important to
consider.
Reasoning
Reasoning is the processing and organizing of ideas
in order to reach conclusions. Through reasoning,
people are able to make sense of their thoughts and
experiences. This type of thinking is often evident in
the verbal presentation of a logical argument in which
each part is linked together to reach a logical conclu-
sion. Patterns of reasoning are used to develop theo-
ries and to plan and implement research. Barnum
(1998) identified four patterns of reasoning as being
essential to nursing: (1) problematic, (2) operational,
(3) dialectic, and (4) logistic. An individual uses all
four types of reasoning, but frequently one type of
reasoning is more dominant than the others. Reason-
ing is also classified by the discipline of logic into
inductive and deductive modes (Chinn & Kramer,
2008).
Problematic Reasoning
Problematic reasoning involves (1) identifying a
problem and the factors influencing it, (2) selecting
solutions to the problem, and (3) resolving the problem.
For example, nurses use problematic reasoning in the
nursing process to identify diagnoses and to imple-
ment nursing interventions to resolve these problems.
Problematic reasoning is also evident when one identi-
fies a research problem and successfully develops a
methodology to examine it.
Operational Reasoning
Operational reasoning involves the identification
of and discrimination among many alternatives and
viewpoints. It focuses on the process (debating alter-
natives) rather than on the resolution (Barnum,
1998). Nurses use operational reasoning to develop
realis­tic, measurable health goals with patients and
families. Nurse practitioners use operational reason-
ing to debate which pharmacological and nonphar-
macological treatments to use in managing patient
illnesses. In research, operationalizing a treatment
for implementation and debating which measurement
methods or data analysis techniques to use in a study
require operational thought (Kerlinger & Lee, 2000;
Waltz et al., 2010).
Dialectic Reasoning
Dialectic reasoning involves looking at situations in
a holistic way. A dialectic thinker believes that the
whole is greater than the sum of the parts and that the
whole organizes the parts (Barnum, 1998). For
example, a nurse using dialectic reasoning would view
a patient as a person with strengths and weaknesses
who is experiencing an illness, and not just as the
“stroke in room 219.” Dialectic reasoning also involves
examining factors that are opposites and making sense
of them by merging them into a single unit or idea that
is greater than either alone. For example, analyzing
studies with conflicting findings and summarizing
these findings to determine the current knowledge
base for a research problem require dialectic reason-
ing. Analysis of data collected in qualitative research
requires dialectic reasoning to gain an understanding
of the phenomenon being investigated (Munhall,
2012).
Logistic Reasoning
Logic is a science that involves valid ways of relating
ideas to promote understanding. The aim of logic is to
determine truth or to explain and predict phenomena.
The science of logic deals with thought processes, such
as concrete and abstract thinking, and methods of rea-
soning, such as logistic, inductive, and deductive.
Logistic reasoning is used to break the whole into
parts that can be carefully examined, as can the rela-
tionships among the parts. In some ways, logistic rea-
soning is the opposite of dialectic reasoning. A logistic
 CHAPTER 1  Discovering the World of Nursing Research 7
reasoner assumes that the whole is the sum of the
parts and that the parts organize the whole. For
example, a patient states that she is cold. You logically
examine the following parts of the situation and their
relationships: (1) room temperature, (2) patient’s tem-
perature, (3) patient’s clothing, and (4) patient’s activ-
ity. The room temperature is 65° F, the patient’s
temperature is 98.6° F, and the patient is wearing
lightweight pajamas and drinking ice water. You con-
clude that the patient is cold because of external envi-
ronmental factors (room temperature, lightweight
pajamas, and drinking ice water). Logistic reasoning
is used frequently in quantitative, outcomes, and
intervention research to develop a study design, plan
and implement data collection, and conduct statistical
analyses.
Inductive and Deductive Reasoning
The science of logic also includes inductive and
deductive reasoning. People use these modes of rea-
soning constantly, although the choice of types of rea-
soning may not always be conscious (Kaplan, 1964).
Inductive reasoning moves from the specific to the
general, whereby particular instances are observed
and then combined into a larger whole or general
statement (Chinn & Kramer, 2008). An example of
inductive reasoning follows:
A headache is an altered level of health that is
stressful.
A fractured bone is an altered level of health that
is stressful.
A terminal illness is an altered level of health that
is stressful.
Therefore, all altered levels of health are
stressful.
In this example, inductive reasoning is used to
move from the specific instances of altered levels of
health that are stressful to the general belief that all
altered levels of health are stressful. By testing many
different altered levels of health through research to
determine whether they are stressful, one can confirm
the general statement that all types of altered health
are stressful.
Deductive reasoning moves from the general to
the specific or from a general premise to a particular
situation or conclusion. A premise or hypothesis is a
statement of the proposed relationship between two or
more variables. An example of deductive reasoning
follows:
PREMISES:
All human beings experience loss.
All adolescents are human beings.
CONCLUSION:
All adolescents experience loss.
In this example, deductive reasoning is used to move
from the two general premises about human beings
experiencing loss and adolescents being human beings
to the specific conclusion, “All adolescents experience
loss.” However, the conclusions generated from deduc-
tive reasoning are valid only if they are based on valid
premises. Consider the following example:
PREMISES:
All health professionals are caring.
All nurses are health professionals.
CONCLUSION:
All nurses are caring.
The premise that all health professionals are caring
is not necessarily valid or an accurate reflection of
reality. Research is a means to test and confirm or
refute a premise so that valid premises can be used as
a basis for reasoning in nursing practice.
Science
Science is a coherent body of knowledge composed
of research findings and tested theories for a specific
discipline (see Figure 1-1). Science is both a product
(end point) and a process (mechanism to reach an end
point) (Silva & Rothbart, 1984). An example from the
discipline of physics is Newton’s law of gravity, which
was developed through extensive research. The knowl-
edge of gravity (product) is a part of the science of
physics that evolved through formulating and testing
theoretical ideas (process). The ultimate goal of
science is to explain the empirical world and thus to
have greater control over it. To accomplish this goal,
scientists must discover new knowledge, expand
existing knowledge, and reaffirm previously held
knowledge in a discipline (Greene, 1979). Health pro-
fessionals integrate this evidence-based knowledge to
control the delivery of care and thereby improve
patient outcomes (evidence-based practice).
The science of a field determines the accepted
process for obtaining knowledge within that field.
Research is an important process for obtaining scien-
tific knowledge in nursing. Some sciences rigidly limit
the types of research that can be used to obtain knowl-
edge. A valued method for developing a science is the
traditional research process, or quantitative research.
According to this process, the information gained
from one study is not sufficient for its inclusion in the
body of science. A study must be replicated several
times and must yield similar results each time before
that information can be considered to be sound empiri-
cal evidence (Fahs, Morgan, & Kalman, 2003).
Consider the research on the relationships between
smoking, lung damage, and cancer. Numerous studies
conducted on animals and humans over the past
8 UNIT ONE  Introduction to Nursing Research
decades indicate causative relationships between
smoking and lung damage and between smoking and
lung cancer. Everyone who smokes experiences lung
damage; and although not everyone who smokes gets
lung cancer, smokers are at a much higher risk for
cancer. Extensive, quality research has been con-
ducted to generate empirical evidence about the health
hazards of smoking, and this evidence guides the
actions of nurses in practice. We provide smoking
cessation education, emotional support, and drugs
like nicotine patches and Chantix (Varenicline) to
assist individuals to stop smoking. On the basis of this
scientific evidence about the hazards of smoking,
society has moved toward providing many smoke-
free environments.
Findings from studies are systematically related to
one another in a way that seems to best explain the
empirical world. Abstract thought processes are used
to make these linkages. The linkages are called laws
or principles, depending on the certainty of the facts
and relationships within the linkage. Laws express the
most certain relationships and provide the best research
evidence for use in practice. The certainty depends on
the amount of research conducted to test a relationship
and, to some extent, on the skills in abstract thought
processes to link the research findings to form mean-
ingful evidence. The truths or explanations of the
empirical world reflected by these laws and principles
are never absolutely certain and may be disproved by
further research.
Nursing is in the beginning stages of developing a
science for the profession, and additional original and
replication studies are needed to develop the knowl-
edge necessary for practice (Fahs et al., 2003; Melnyk
& Fineout-Overholt, 2011). As discussed earlier,
nursing science is being developed with the use of a
variety of research methodologies, including quantita-
tive, qualitative, outcomes, and intervention. The
focus of this textbook is to increase your understand-
ing of these different types of research used in the
development and testing of nursing theory.
Theory
A theory is a creative and rigorous structuring of ideas
used to describe, explain, predict, or control a particu-
lar phenomenon or segment of the empirical world
(Chinn & Kramer, 2008; Smith & Liehr, 2008). A
theory consists of a set of concepts that are defined
and interrelated to present a systematic view of a phe-
nomenon. A classic example is the theory of stress
developed by Selye (1976) to explain the physical and
emotional affects of illness on peoples’ lives. This
theory of stress continues to be important in under-
standing the affects of health changes on patients and
families. Extensive research has been conducted to
detail the types, number, and severity of stressors
experienced in life and the effective interventions for
managing these stressful situations.
A theory is developed from a combination of per-
sonal experiences, research findings, and abstract
thought processes. The theorist may use findings from
research as a starting point and then organize the find-
ings to best explain the empirical world. This is the
process Selye used to develop his theory of stress.
Alternatively, the theorist may use abstract thought
processes, personal knowledge, and intuition to
develop a theory of a phenomenon. This theory then
requires testing through research to determine whether
it is an accurate reflection of reality. Thus, research
has a major role in theory development, testing, and
refinement. Some forms of qualitative research focus
on developing new theories or extending existing
theories. Quantitative, outcomes, and intervention
methods of research are often implemented to test the
accuracy of theory. The study findings either support
or fail to support the theory, providing a basis for
refining the theory (Chinn & Kramer, 2008; Fawcett
& Garity, 2009).
Knowledge
Knowledge is a complex, multifaceted concept. For
example, you may say that you know your friend John,
know that the earth rotates around the sun, know how
to give an injection, and know pharmacology. These
are examples of knowing—being familiar with a
person, comprehending facts, acquiring a psychomo-
tor skill, and mastering a subject. There are differences
in types of knowing, yet there are also similarities.
Knowing presupposes order or imposes order on
thoughts and ideas (Engelhardt, 1980). People have a
desire to know what to expect. There is a need for
certainty in the world, and individuals seek it by trying
to decrease uncertainty through knowledge. Think of
the questions you ask a person who has presented
some bit of knowledge: “Is it true?” “Are you sure?”
“How do you know?” Thus, the knowledge that we
acquire is expected to be an accurate reflection of
reality.
Ways of Acquiring Nursing Knowledge
We acquire knowledge in a variety of ways and expect
it to be an accurate reflection of the real world (White,
1982). Nurses have historically acquired knowledge
through (1) traditions, (2) authority, (3) borrowing,
 CHAPTER 1  Discovering the World of Nursing Research 9
(4) trial and error, (5) personal experience, (6) role-
modeling and mentorship, (7) intuition, (8) reasoning,
and (9) research. Intuition, reasoning, and research
were discussed earlier in this chapter; the other ways
of acquiring knowledge are briefly described in this
section.
Traditions
Traditions consist of “truths” or beliefs that are based
on customs and past trends. Nursing traditions from
the past have been transferred to the present by written
and verbal communication and role-modeling and
continue to influence the present practice of nursing.
For example, some of the policies and procedures in
hospitals and other healthcare facilities contain tradi-
tional ideas. In addition, some nursing interventions
are transmitted verbally from one nurse to another
over the years or by the observation of experienced
nurses. For example, the idea of providing a patient
with a clean, safe, well-ventilated environment origi-
nated with Florence Nightingale (1859).
However, traditions can also narrow and limit the
knowledge sought for nursing practice. For example,
tradition has established the time and pattern for pro-
viding baths, evaluating vital signs, and allowing
patient visitation on many hospital units. The nurses
on these units quickly inform new staff members
about the accepted or traditional behaviors for the unit.
Traditions are difficult to change because people with
power and authority have accepted and supported
them for a long time. Many traditions have not been
tested for accuracy or efficiency and require research
for continued use in practice.
Authority
An authority is a person with expertise and power
who is able to influence opinion and behavior. A
person is thought of as an authority because she or he
knows more in a given area than others do. Knowl-
edge acquired from authority is illustrated when one
person credits another person as the source of informa-
tion. Nurses who publish articles and books or develop
theories are frequently considered authorities. Stu-
dents usually view their instructors as authorities, and
clinical nursing experts are considered authorities
within their clinical settings. However, persons viewed
as authorities in one field are not necessarily authori-
ties in other fields. An expert is an authority only when
addressing his or her area of expertise. Like tradition,
the knowledge acquired from authorities sometimes
has not been validated through research and is not
considered the best evidence for practice.
Borrowing
As some nursing leaders have noted, knowledge in
nursing practice is partly made up of information that
has been borrowed from disciplines such as medicine,
psychology, physiology, and education (McMurrey,
1982; Walker & Avant, 2011). Borrowing in nursing
involves the appropriation and use of knowledge from
other fields or disciplines to guide nursing practice.
Nursing practice has borrowed knowledge in two
ways. For years, some nurses have taken information
from other disciplines and applied it directly to
nursing practice. This information was not integrated
within the unique focus of nursing. For example,
some nurses have used the medical model to guide
their nursing practice, thus focusing on the diagnosis
and treatment of physiological diseases with limited
attention to the patient’s holistic nature. This type of
borrowing continues today as nurses use technologi-
cal advances to focus on the detection and treatment
of disease, to the exclusion of health promotion and
illness prevention.
Another way of borrowing, which is more useful
in nursing, is the integration of information from other
disciplines within the focus of nursing. Because disci-
plines share knowledge, it is sometimes difficult to
know where the boundaries exist between nursing’s
knowledge base and the knowledge bases of other
disciplines. Boundaries blur as the knowledge bases
of disciplines evolve (McMurrey, 1982). For example,
information about self-esteem as a characteristic of the
human personality is associated with psychology, but
this knowledge also directs the nurse in assessing the
psychological needs of patients and families. However,
borrowed knowledge has not been adequate to answer
many questions generated in nursing practice.
Trial and Error
Trial and error is an approach with unknown out-
comes that is used in a situation of uncertainty when
other sources of knowledge are unavailable. The
nursing profession evolved through a great deal of trial
and error before knowledge of effective practices was
codified in textbooks and journals. The trial-and-error
way of acquiring knowledge can be time-consuming,
because multiple interventions might be implemented
before one is found to be effective. There is also a risk
of implementing nursing actions that are detrimental
to a patient’s health. Because each patient responds
uniquely to a situation, however, uncertainty in nursing
practice continues. Because of the uniqueness of
patient response and the resulting uncertainty, nurses
must use trial and error in providing care. The
10 UNIT ONE  Introduction to Nursing Research
trial-and-error approach to developing knowledge
would be more efficient if nurses documented the
patient and situational characteristics that provided the
context for the patient’s unique response.
Personal Experience
Personal experience is the knowledge that comes
from being personally involved in an event, situation,
or circumstance. In nursing, personal experience
enables one to gain skills and expertise by providing
care to patients and families in clinical settings. The
nurse not only learns but is able to cluster ideas into
a meaningful whole. For example, students may be
told how to give an injection in a classroom setting,
but they do not know how to give an injection until
they observe other nurses giving injections to patients
and actually give several injections themselves.
The amount of personal experience you have will
affect the complexity of your knowledge base as a
nurse. Benner (1984) described five levels of experi-
ence in the development of clinical knowledge and
expertise that are important today. These levels of
experience are (1) novice, (2) advanced beginner,
(3) competent, (4) proficient, and (5) expert. Novice
nurses have no personal experience in the work that
they are to perform, but they have preconceived
notions and expectations about clinical practice that
are challenged, refined, confirmed, or contradicted by
personal experience in a clinical setting. The advanced
beginner has just enough experience to recognize and
intervene in recurrent situations. For example, the
advanced beginner nurse is able to recognize and inter-
vene to meet patients’ needs for pain management.
Competent nurses frequently have been on the job
for 2 or 3 years, and their personal experiences enable
them to generate and achieve long-range goals and
plans (Benner, 1984). Through experience, the com-
petent nurse is able to use personal knowledge to take
conscious, deliberate actions that are efficient and
organized. From a more complex knowledge base, the
proficient nurse views the patient as a whole and as a
member of a family and community. The proficient
nurse recognizes that each patient and family have
specific values and needs that lead them to respond
differently to illness and health.
The expert nurse has had extensive experience and
is able to identify accurately and intervene skillfully
in a situation (Benner, 1984). Personal experience
increases an expert nurse’s ability to grasp a situation
intuitively with accuracy and speed. Lyneham, Parkin-
son, and Denholm (2009) studied Benner’s fifth stage
of practice development and noted the links of intu-
ition, science, knowledge, and theory to expert clinical
practice. The clinical expertise of the nurse is a critical
component of evidence-based practice. It is the expert
nurse who has the greatest skill and ability to imple-
ment the best research evidence in practice to meet the
unique values and needs of patients and families.
Role-Modeling and Mentorship
Role-modeling is learning by imitating the behaviors
of an exemplar. An exemplar or role model knows the
appropriate and rewarded roles for a profession, and
these roles reflect the attitudes and include the stan-
dards and norms of behavior for that profession (ANA,
2010a). In nursing, role-modeling enables the novice
nurse to learn from interacting with expert nurses or
following their examples. Examples of role models are
admired teachers, expert practitioners, researchers,
and illustrious individuals who inspire students, prac-
ticing nurses, educators, and researchers through their
examples.
An intense form of role-modeling is mentorship.
In a mentorship, the expert nurse, or mentor, serves
as a teacher, sponsor, guide, exemplar, and counselor
for the novice nurse (or mentee). Both the mentor and
the mentee or protégé invest time and effort, which
often result in a close, personal mentor-mentee rela-
tionship. This relationship promotes a mutual exchange
of ideas and aspirations relative to the mentee’s career
plans. The mentee assumes the values, attitudes, and
behaviors of the mentor while gaining intuitive knowl-
edge and personal experience. Mentorship is essential
for building research competence in nursing (Byrne &
Keefe, 2002).
To summarize, in nursing, a body of knowledge
must be acquired (learned), incorporated, and assimi-
lated by each member of the profession and collec-
tively by the profession as a whole. This body of
knowledge guides the thinking and behavior of the
profession and of individual practitioners. It also
directs further development and influences how
science and theory are interpreted within the discipline
(see Figure 1-1). This knowledge base is necessary in
order for health professionals, consumers, and society
to recognize nursing as a science.
Philosophy
Philosophy provides a broad, global explanation of
the world. It is the most abstract and most all-
encompassing concept in the model (see Figure 1-1).
Philosophy gives unity and meaning to the world of
nursing and provides a framework within which think-
ing, knowing, and doing occur (Kikuchi & Simmons,
1994). Nursing’s philosophical position influences its
knowledge base. How nurses use science and theories
 CHAPTER 1  Discovering the World of Nursing Research 11
to explain the empirical world depends on their phi-
losophy. Ideas about truth and reality, as well as
beliefs, values, and attitudes, are part of philosophy.
Philosophy asks questions such as, “Is there an abso-
lute truth, or is truth relative?” and “Is there one
reality, or is reality different for each individual?”
Everyone’s world is modified by her or his philoso-
phy, as a pair of eyeglasses would modify vision.
Perceptions are influenced first by philosophy and
then by knowledge. For example, if what you see is
not within your ideas of truth or reality, if it does not
fit your belief system, you may not see it. Your mind
may reject it altogether or may modify it to fit your
philosophy (Scheffler, 1967). For example, you might
believe that education is not effective in promoting
smoking cessation, so you do not provide your patients
this education. As you start to discover the world of
nursing research, it is important for you to keep an
open mind to the value of research and your future
role in the development or use of research evidence in
practice.
Philosophical positions commonly held within the
nursing profession include the view that human beings
are holistic, rational, and responsible. Nurses believe
that people desire health, and health is considered to
be better than illness. Quality of life is as important as
quantity of life. Good nursing care facilitates improved
patterns of health and quality of life (ANA, 2010a,
2010b). In nursing, truth is relative, and reality tends
to vary with perception (Kikuchi, Simmons, & Romyn,
1996; Silva, 1977). For example, because nurses
believe that reality varies with perception and that
truth is relative, they would not try to impose their
views of truth and reality on patients. Rather, they
would accept patients’ views of the world and help
Clinical
Expertise
Best
Research Evidence-Based
Practice
Evidence
Patient
Needs and
Values
them seek health from within those worldviews, an
approach that is a critical component of evidence-
based practice.
Significance of Research in
Building an Evidence-Based
Practice for Nursing
The ultimate goal of nursing is to provide evidence-
based care that promotes quality outcomes for patients,
families, healthcare providers, and the healthcare
system (Craig & Smyth, 2012; Melnyk & Fineout-
Overholt, 2011). Evidence-based practice (EBP)
evolves from the integration of the best research evi-
dence with clinical expertise and patient needs and
values (Sackett, Straus, Richardson, Rosenberg, &
Haynes, 2000). Figure 1-2 demonstrates the major
contribution of the best research evidence to the deliv-
ery of EBP. Best research evidence is the empirical
knowledge generated from the synthesis of quality
study findings to address a practice problem. A discus-
sion of the levels of best research evidence and the
sources for this evidence is presented in Chapter 2. A
team of expert researchers, healthcare professionals,
policy makers, and consumers often synthesizes the
best research evidence for developing standardized
guidelines for clinical practice. For example, research
related to the chronic health problem of hypertension
(HTN) has been conducted, critically appraised, and
synthesized by experts to develop a practice guideline
for implementation by APNs, such as nurse practi­
tioners, and physicians to ensure that patients with
HTN receive quality, cost-effective care (Chobanian
Figure 1-2  Model of evidence-based practice.
12 UNIT ONE  Introduction to Nursing Research
JNC 7
Figure 1-3  Evidence-based practice for elderly
Hypertension
African-American women with hypertension
Guideline
(HTN). JNC, Joint National Committee on
(Best research
Prevention, Detection, Evaluation, and Treatment
evidence)
of High Blood Pressure.
et al., 2003). The most current guidelines for the diag-
nosis and management of HTN, “The Seventh Report
of the Joint National Committee on Prevention, Detec-
tion, Evaluation, and Treatment of High Blood Pres-
sure: The JNC 7 Report,” were published in 2003
(Chobanian et al., 2003) and are available online at
www.nhlbi.nih.gov/guidelines/hypertension. The JNC
8 Report is currently under development, with pro-
jected publication of the revised HTN guidelines
in 2012 or 2013 (see http://www.nhlbi.nih.gov/guide
lines/hypertension/jnc8). Many national standardized
guidelines are available through the Agency for
Healthcare Research and Quality (AHRQ) and profes-
sional organizations, which are discussed in more
detail in Chapters 2 and 19.
Clinical expertise is the knowledge and skills of
the healthcare professional providing care. A nurse’s
clinical expertise is determined by his or her years of
practice, current knowledge of the research and clini-
cal literature, and educational preparation. The stron-
ger the nurse’s clinical expertise, the better his or her
clinical judgment is in the delivery of quality care
(Craig & Smyth, 2012; Eizenberg, 2010). The patient’s
need(s) might focus on health promotion, illness pre-
vention, acute or chronic illness management, or reha-
bilitation (see Figure 1-2). In addition, patients bring
values or unique preferences, expectations, concerns,
and cultural beliefs to the clinical encounter. With
EBP, patients and their families are encouraged to take
an active role in managing their health care. In
summary, expert clinicians use the best research evi-
dence available to deliver quality, cost-effective care
to a patients and families with specific health needs
and values to achieve EBP (Brown, 2009; Craig &
Smyth, 2012; Sackett et al., 2000).
Nurse
Practitioner
(Clinical expert)
Women with Normal
Blood Pressure of
Less Than 120/80
(Evidence-Based Practice)
Female, elderly,
African-American
with HTN
(Patient needs
and values)
Figure 1-3 provides an example of the delivery of
evidence-based care to women with HTN. In this
example, the best research evidence on HTN is the
JNC 7 National Standardized Guideline (Chobanian et
al., 2003). An expert nurse practitioner translates this
guideline to meet the needs (chronic illness manage-
ment) and values of elderly African-American women
with HTN. In this case, the outcome of EBP is women
with a normal blood pressure, less than 120 mm Hg
systolic/80 mm Hg diastolic (see Figure 1-3). A
detailed discussion of how to locate, critically appraise,
and use national standardized guidelines in practice is
presented in Chapter 19.
In nursing, the research evidence must focus on the
description, explanation, prediction, and control of
phenomena important to practice. The following sec-
tions address the types of knowledge that need to be
generated in these four areas as nursing moves toward
EBP.
Description
Description involves identifying and understanding
the nature of nursing phenomena and, sometimes, the
relationships among them (Chinn & Kramer, 2008;
Munhall, 2012). Through research, nurses are able to
(1) explore and describe what exists in nursing prac-
tice, (2) discover new information, (3) promote under-
standing of situations, and (4) classify information for
use in the discipline. Some examples of clinically
important research evidence developed from research
focused on description include the following:
• Identification of the responses of individuals to a
variety of health conditions
• Description of the health promotion and illness pre-
vention strategies used by various populations
 CHAPTER 1  Discovering the World of Nursing Research 13
• Determination of the incidence of a disease locally,
nationally, and internationally
• Identification of the cluster of symptoms for a par-
ticular disease
• Description of the effects and side effects of
selected pharmacological agents in a variety of
populations
For example, Imes, Daugherty, Pyper, and Sullivan
(2011, p. 208) conducted “a qualitative study to
describe the experience of living with heart failure
(HF) from the perspective of the partner.” These
researchers synthesized their findings as follows: “The
severity of the patient’s disease limited the partner’s
lifestyle, resulting in social isolation and difficulties in
planning for the future for both the patient and the
partner. The partners were unprepared to manage the
disease burden at home without consistent information
and assistance by healthcare providers. Moreover,
end-of-life planning was neither encouraged by health-
care providers nor embraced by patients or partners”
(Imes et al., 2011, p. 208).
The findings from this study provide insights into
the experience of HF by a loved one and their experi-
ences with healthcare providers. This type of research,
focused on description, is essential groundwork for
studies that will help to explain, predict, and control
nursing phenomena.
Explanation
Explanation clarifies the relationships among phe-
nomena and clarifies why certain events occur.
Research focused on explanation provides the follow-
ing types of evidence essential for practice:
• Determination of the assessment data (both subjec-
tive data from the health history and objective data
from physical examination) needed to address a
patient’s health need
• Link of assessment data to determine a diagnosis
(both nursing and medical)
• Link of causative risk factors or etiologies to
illness, morbidity, and mortality
• Determine the relationships among health risks,
health status, and healthcare costs
For example, Manojlovich, Sidani, Covell, and
Antonakos (2011) conducted an outcomes study to
examine the links between a “nurse dose” (nurse char-
acteristics and staffing) and adverse patient outcomes.
The nurse characteristics examined were education,
experience, and skill mix. The staffing variables
included full-time employees, RN:patient ratio, and
RN hours per patient day. The adverse outcomes
examined were methicillin-resistant Staphylococcus
aureus (MRSA) infections and reported patient falls
for a sample of inpatient adults in acute care units. The
researchers found that the nurse characteristics and
staffing variables were significantly correlated with
MRSA infections and reported patient falls. Thus, the
nursing characteristics and staffing were potential pre-
dictors of MRSA infections and patient falls. This
study illustrates how explanatory research can identify
relationships among nursing phenomena that are the
basis for future research focused on prediction and
control.
Prediction
Through prediction, one can estimate the probability
of a specific outcome in a given situation (Chinn &
Kramer, 2008). However, predicting an outcome does
not necessarily enable one to modify or control the
outcome. It is through prediction that the risk of illness
is identified and linked to possible screening methods
that will identify the illness. Knowledge generated
from research focused on prediction is critical for EBP
and includes the following:
• Prediction of the risk for a disease in different
populations
• Prediction of the accuracy and precision of a
screening instrument, such as mammogram, to
detect a disease
• Prediction of the prognosis once an illness is identi-
fied in a variety of populations
• Prediction of the impact of nursing actions on
selected outcomes
• Prediction of behaviors that promote health and
prevent illness
• Prediction of the health care required based on a
patient’s need and values
For example, Scheetz and Kolassa (2007, p. 399)
examined “crash scene variables to predict the need
for trauma center care in older persons.” The research-
ers analyzed 26 crash scene variables and developed
triage decision rules for managing persons with severe
and moderate injuries. Further research is needed to
determine whether the triage decision rules improve
the health outcomes of the elderly following trauma.
Predictive studies isolate independent variables that
require additional research to ensure that their manip-
ulation or control results in successful outcomes for
patients, healthcare professionals, and healthcare
agencies.
Control
If one can predict the outcome of a situation, the next
step is to control or manipulate the situation to produce
the desired outcome. Dickoff, James, and Wiedenbach
(1968) described control as the ability to write a
14 UNIT ONE  Introduction to Nursing Research
prescription to produce the desired results. Using the
best research evidence, nurses could prescribe specific
interventions to meet the needs of patients. Nurses
need this type of research evidence to provide EBP
(see Figure 1-2). Research in the following areas is
important for generating EBP in nursing:
• Testing interventions to improve the health status
of individuals, families, and communities
• Testing interventions to improve healthcare
delivery
• Determining the quality and cost-effectiveness of
interventions
• Implementing an evidence-based intervention to
determine whether it is effective in managing a
patient’s health need (health promotion, illness pre-
vention, acute and chronic illness management, and
rehabilitation) and producing quality outcomes
Yoo, Kim, Hur, and Kim (2011) conducted a study
that examined the effect of a prescribed animation
distraction intervention on the pain response of pre-
school children during venipuncture. The intervention
or independent variable was a 3-minute animation
video that could be downloaded from the Internet and
shown to the child using a laptop computer. The pain
response was measured by the following dependent
variables: “self-reported pain response, behavioral
pain response, blood cortisol, and blood glucose” (Yoo
et al., 2011, p. 94). The researchers found a significant
difference between the experimental and control
groups for all four dependent variables of pain
response. Thus, the animation distraction intervention
was determined to be an effective method of managing
children’s pain during venipuncture. The researchers
concluded that this intervention required minimal
effort and time and might be a convenient and cost-
effective intervention to be used in clinical settings to
reduce children’s pain.
Many more studies need to be conducted to gener-
ate the research evidence in the areas of prediction and
control (Brown, 2009; Craig & Smyth, 2012; Melnyk
& Fineout-Overholt, 2011). This need for additional
nursing research provides you with many opportuni-
ties to be involved in the world of nursing research.
This chapter introduced you to the world of nursing
research and the significance of research in developing
an EBP for nursing. The following chapters will
expand your understanding of different research meth-
odologies so you can critically appraise studies, syn-
thesize research findings, and use the best research
evidence available in clinical practice. This text also
gives you a background for conducting research in
collaboration with expert nurse researchers. We think
you will find that nursing research is an exciting
adventure that holds much promise for the future prac-
tice of nursing.
KEY POINTS
• This chapter introduces you to the world of nursing
research.
• Nursing research is defined as a scientific process
that validates and refines existing knowledge and
generates new knowledge that directly and indi-
rectly influences the delivery of evidence-based
nursing practice (EBP).
• This chapter presents a framework that links nursing
research to the world of nursing and organizes the
content presented in this textbook (see Figure 1-1).
The concepts in this framework range from concrete
to abstract and include concrete and abstract think-
ing, the empirical world (evidence-based nursing
practice), research, abstract thought processes,
science, theory, knowledge, and philosophy.
• The empirical world is what we experience through
our senses and is the concrete portion of our exis-
tence where nursing practice occurs.
• Research is a way to test reality, and nurses use a
variety of research methods to test their reality and
generate nursing knowledge, such as quantitative,
qualitative, outcomes, and intervention.
• All nurses have a role in research—some are
developers of research and conduct studies to gen-
erate and refine the knowledge needed for nursing
practice, and others are consumers of research and
use research evidence to improve their nursing
practice.
• Three major abstract thought processes—
introspection, intuition, and reasoning—are impor-
tant in nursing.
• A theory is a creative and rigorous structuring of
ideas used to describe, explain, predict, or control
a particular phenomenon or segment of the empiri-
cal world.
• Reliance on tradition, authority, trial and error, and
personal experience is no longer an adequate basis
for sound nursing practice.
• The goal of nurses and other healthcare profession-
als is to deliver evidence-based health care to
patients and their families.
• EBP evolves from the integration of best research
evidence with clinical expertise and patient needs
and values (see Figure 1-2).
• The best research evidence is the empirical knowl-
edge generated from the synthesis of quality studies
to address a practice problem.
 CHAPTER 1  Discovering the World of Nursing Research 15
• The clinical expertise of a nurse is determined by
his or her years of clinical experience, current
knowledge of the research and clinical literature,
and educational preparation.
• The patient brings values—such as unique prefer-
ences, expectations, concerns, and cultural beliefs,
and health needs—to the clinical encounter, which
are important to consider in providing evidence-
based care.
• The knowledge generated through research is
essential for describing, explaining, predicting, and
controlling nursing phenomena.
• Nursing practice based on synthesized research
findings can have a powerful, positive impact on
patient outcomes and the healthcare system.
REFERENCES
Abbott, E. A. (1952). Flatland. New York, NY: Dover.
American Association of Colleges of Nursing (AACN, 2006).
AACN Position statement on nursing research. Washington, DC:
AACN. Retrieved from http://www.aacn.nche.edu/Publications/
positions/NsgRes.htm.
American Nurses Association (ANA, 2012). What is nursing?
Retrieved from http://www.nursingworld.org/EspeciallyForYou/
StudentNurses/WhatisNursing.aspx.
American Nurses Association. (2010a). Nursing: Scope and stan-
dards of practice (2nd ed.). Silver Spring, MD: Author.
American Nurses Association. (2010b). Nursing’s social policy
statement: The essence of the profession. Silver Spring, MD:
Author.
Barnum, B. S. (1998). Nursing theory: Analysis, application, evalu-
ation (5th ed.). Philadelphia, PA: Lippincott Williams & Wilkins.
Benner, P. (1984). From novice to expert: Excellence and power in
clinical nursing practice. Menlo Park, CA: Addison-Wesley.
Brar, K., Boschma, G., & McCuaig, F. (2010). The development of
nurse practitioner preparation beyond the master’s level: What is
the debate about? International Journal of Nursing Education
Scholarship, 7(1), Article 9.
Brown, S. J. (2009). Evidence-based nursing: The research-practice
connection. Sudbury, MA: Jones and Bartlett.
Byrne, M. W., & Keefe, M. R. (2002). Building research compe-
tence in nursing through mentoring. Journal of Nursing Scholar-
ship, 34(4), 391–396.
Chinn, P. L., & Kramer, M. K. (2008). Theory and nursing: Inte-
grated knowledge development (6th ed.). St. Louis, MO: Mosby.
Chobanian, A.V., Bakris, G. L., Black, H. R., Cushman, W. C.,
Green, L. A., Izzo, J. L., et al. (2003). The seventh report of the
Joint National Committee on prevention, detection, evaluation,
and treatment of high blood pressure: The JNC 7 report. Journal
of the American Medical Association, 289(19), 2560–2572.
Clinton, P., & Sperhac, A. M. (2006). National agenda for advanced
practice nursing: The practice doctorate. Journal of Professional
Nursing, 22(1), 7–14.
Craig, J. V., & Smyth, R. L. (2012). The evidence-based practice
manual for nurses (3rd ed.). Edinburgh, Scotland: Churchill
Livingstone.
Creswell, J. W. (2009). Research design: Qualitative, quantitative,
and mixed methods approaches (3rd ed.). Los Angeles, CA: Sage.
Dickoff, J., James, P., & Wiedenbach, E. (1968). Theory in a prac-
tice discipline: Practice oriented theory (Part I). Nursing Research,
17(5), 415–435.
Doran, D. M. (2011). Nursing-sensitive outcomes: State of the
science. Sudbury, MA: Jones & Bartlett.
Eizenberg, M. M. (2010). Implementation of evidence-based
nursing practice: Nurses’ personal and professional factors?
Journal of Advanced Nursing, 67(1), 33–42.
Engelhardt, H. T., Jr. (1980). Knowing and valuing: Looking for
common roots. In H. T. Engelhardt & D. Callahan (Eds.), Knowing
and valuing: The search for common roots (Vol. 4, pp. 1–17). New
York: Hastings Center.
Fahs, P. S., Morgan, L. L., & Kalman, M. (2003). A call for replica-
tion. Journal of Nursing Scholarship, 35(1), 67–71.
Fawcett, J., & Garity, J. (2009). Evaluating research for evidence-
based nursing practice. Philadelphia, PA: F. A. Davis.
Forbes, A. (2009). Clinical intervention research in nursing. Inter-
national Journal of Nursing Studies, 46(4), 557–568.
Greene, J. A. (1979). Science, nursing and nursing science: A con-
ceptual analysis. Advances in Nursing Science, 2(1), 57–64.
Imes, C. C., Dougherty, C. M., Pyper, G., & Sullivan, M. D. (2011).
Care of patients with heart failure: Descriptive study of partners’
experiences of living with severe heart failure. Heart & Lung,
40(3), 208–216.
Kaplan, A. (1964). The conduct of inquiry. New York, NY: Harper
& Row.
Kerlinger, F. N., & Lee, H. B. (2000). Foundations of behavioral
research (4th ed.). Fort Worth, TX: Harcourt College
Publishers.
Kikuchi, J. F., & Simmons, H. (1994). Developing a philosophy of
nursing. Thousand Oaks, CA: Sage.
Kikuchi, J. F., Simmons, H., & Romyn, D. (1996). Truth in nursing
inquiry. Thousand Oaks, CA: Sage.
Lyneham, J., Parkinson, C., & Denholm, C. (2009). Expert nursing
practice: A mathematical explanation of Benner’s 5th stage of
practice development. Journal of Advanced Nursing, 65(11),
2477–2484.
Mackay, M. (2009). Why nursing has not embraced the clinician-
scientist role. Nursing Philosophy, 10(4), 287–296.
Manojlovich, M., Sidani, S., Covell, C. L., & Antonakos, C. L.
(2011). Nurse dose: Linking staffing variables to adverse patient
outcomes. Nursing Research, 60(4), 214–220.
Marshall, C., & Rossman, G. B. (2011). Designing qualitative
research (5th ed.). Los Angeles, CA: Sage.
McMurrey, P. H. (1982). Toward a unique knowledge base in
nursing. Image-Journal of Nursing Scholarship, 14(1), 12–15.
Melnyk, B. M., & Fineout-Overholt, E. (2011). Evidence-based
practice in nursing & healthcare: A guide to best practice (2nd
ed.). Philadelphia, PA: Lippincott Williams & Wilkins.
Munhall, P. L. (2012). Nursing research: A qualitative perspective
(5th ed.). Sudbury, MA: Jones & Bartlett Learning.
National League for Nursing (NLN, 2009). Building a science of
nursing education: Foundation for evidence-based teaching and
learning. New York, NY: Author.
Nightingale, F. (1859). Notes on nursing: What it is, and what it is
not. Philadelphia, PA: Lippincott.
16 UNIT ONE  Introduction to Nursing Research
Odell, M., Victor, C., & Oliver, D. (2009). Nurses’ role in detecting
deterioration in ward patients: Systematic literature review.
Journal of Advanced Nursing, 65(10), 1992–2006.
Polanyi, M. (1962). Personal knowledge. Chicago, IL: University
of Chicago Press.
Polanyi, M. (1966). The tacit dimension. New York, NY:
Doubleday.
Riley, J. M., Beal, J., Levi, P., & McCausland, M. P. (2002). Revi-
sioning nursing scholarship. Journal of Nursing Scholarship,
34(4), 383–389.
Sackett, D. L., Straus, S. E., Richardson, W. S., Rosenberg, W.,
& Haynes, R. B. (2000). Evidence-based medicine: How to prac-
tice & teach EBM (2nd ed.). London, England: Churchill
Livingstone.
Scheetz, L. J., & Kolassa, J. E. (2007). Using crash scene variables
to predict the need for trauma center care in older persons.
Research in Nursing & Health, 30(4), 399–412.
Scheffler, I. (1967). Science and subjectivity. Indianapolis, IN:
Bobbs-Merrill.
Selye, H. (1976). The stress of life. New York, NY: McGraw-Hill.
Shadish, W. R., Cook, T. D., & Campbell, D. T. (2002). Experimen-
tal and quasi-experimental designs for generalized causal infer-
ence. Chicago, IL: Rand McNally.
Silva, M.C. (1977). Philosophy, science, theory: Interrelationships
and implications for nursing research. Image-Journal of Nursing
Scholarship, 9(3), 59–63.
Silva, M. C., & Rothbart, D. (1984). An analysis of changing trends
in philosophies of science on nursing theory development and
testing. Advances in Nursing Science, 6(2), 1–13.
Smith, A. (2009). Exploring the legitimacy of intuition as a form of
nursing knowledge. Nursing Standard, 23 (40), 35–40.
Smith, M. J., & Liehr, P. R. (2008). Middle range theory for nursing
(2nd ed.). New York, NY: Springer Publishing Company.
Walker, L. O., & Avant, K. C. (2011). Strategies for theory construc-
tion in nursing (5th ed.). Norwalk, CT: Appleton & Lange.
Waltz, C. F., Strickland, O. L., & Lenz, E. R. (2010). Measurement
in nursing and health research (4th ed.). New York, NY: Springer
Publishing Company.
White, A. R. (1982). The nature of knowledge. Totowa, NJ: Rowman
& Littlefield.
Yoo, H., Kim, S., Hur, H., & Kim, H. (2011). The effects of an
animation distraction intervention on pain response of preschool
children during venipuncture. Applied Nursing Research, 24(2),
94–100.
   http://evolve.elsevier.com/Grove/practice/
2
Evolution of Research in Building
Evidence-Based Nursing Practice
I
nitially, nursing research evolved slowly, from
Florence Nightingale’s investigations of patient
mortality in the nineteenth century to the studies
of nursing education in the 1930s and 1940s. Nurses
and nursing roles were the focus of research in the
1950s and 1960s. However, in the late 1970s and
1980s, many researchers designed studies aimed at
improving nursing practice. This emphasis continued
in the 1990s with research focused on testing the
effectiveness of nursing interventions and examining
patient outcomes. The goal in this millennium is the
development of an evidence-based practice for
nursing, with the current best research evidence being
used to deliver quality health care.
Evidence-based practice (EBP) is the conscien-
tious integration of best research evidence with
clinical expertise and patient values and needs in
the delivery of quality, cost-effective health care.
Chapter 1 presents a model depicting the elements
of EBP and provides an example (see Figures 1-2 and
1-3). You probably have many questions about
EBP because it is an evolving concept in nursing and
health care. What does “best research evidence”
mean? How is research evidence developed? Are there
levels of quality in the types of research evidence?
This chapter will increase your understanding of how
nursing research has evolved over the past 150 years
and of the current movement of the profession toward
EBP. The chapter describes the historical events rel-
evant to nursing research in building an EBP, identifies
the methodologies used in nursing to develop research
evidence, and concludes with a discussion of the best
research evidence needed to build an EBP.
Historical Development
of Research in Nursing
Some people think that research is relatively new to
nursing, but Florence Nightingale initiated nursing
 
CHAPTER
research more than 150 years ago (Nightingale, 1859).
Following Nightingale’s work (1840-1910), nursing
research received minimal attention until the mid-
1900s. In the 1960s, nurses gradually recognized the
value of research, but few had the educational back-
ground to conduct studies until the 1970s. However,
in the 1980s and 1990s, research became a major force
in developing a scientific knowledge base for nursing
practice. Today, nurses obtain federal, corporate, and
foundational funding for their research, conduct
complex studies in multiple settings, and generate
sound research evidence for practice. Table 2-1 identi-
fies key historical events that have influenced the
development of nursing research and the movement
toward EBP. These events are discussed in the follow-
ing section.
Florence Nightingale
Nightingale has been described as a reformer, reac-
tionary, and researcher who influenced nursing spe-
cifically and health care in general. Nightingale’s
book, Notes on Nursing (1859), described her initial
research activities, which focused on the importance
of a healthy environment in promoting the patient’s
physical and mental well-being. She identified the
need to gather data on the environment, such as ven-
tilation, cleanliness, temperature, purity of water, and
diet, to determine their influence on the patient’s
health (Herbert, 1981).
Nightingale is also noted for her data collection and
statistical analyses during the Crimean War. She gath-
ered data on soldier morbidity and mortality rates and
the factors influencing them and presented her results
in tables and pie charts, a sophisticated type of data
presentation for the period (Cohen, 1984; Palmer,
1977). Nightingale was the first woman elected to the
Royal Statistical Society (Oakley, 2010), and her
research was highlighted in the periodical Scientific
American in 1984 (Cohen, 1984).
17
18 UNIT ONE  Introduction to Nursing Research

TABLE 2-1  Historical Events Influencing Research in Nursing

Year Event
1850 Florence Nightingale is recognized as the first nurse researcher.
1900 American Journal of Nursing is published.
1923 Teachers College at Columbia University offers the first educational doctoral program for nurses.
1929 First Master’s in Nursing Degree is offered at Yale University.
1932 Association of Collegiate Schools of Nursing is organized to promote conduct of research.
1950 American Nurses Association (ANA) publishes study of nursing functions and activities.
1952 First research journal in nursing, Nursing Research, is published.
1953 Institute of Research and Service in Nursing Education is established.
1955 American Nurses Foundation is established to fund nursing research.
1957 Southern Regional Educational Board (SREB), Western Interstate Commission on Higher Education (WICHE),
Midwestern Nursing Research Society (MNRS), and New England Board of Higher Education (NEBHE) are
developed to support and disseminate nursing research.
1963 International Journal of Nursing Studies is published.
1965 ANA sponsors the first nursing research conferences.
1967 Sigma Theta Tau International Honor Society of Nursing publishes Image, emphasizing nursing scholarship; now
entitled Journal of Nursing Scholarship.
1970 ANA Commission on Nursing Research is established.
1972 Cochrane published Effectiveness and Efficiency, introducing concepts relevant to evidence-based practice (EBP).
ANA Council of Nurse Researchers is established.
1973 First Nursing Diagnosis Conference is held, which evolved into North American Nursing Diagnosis Association
(NANDA).
1976 Stetler/Marram Model for Application of Research Findings to Practice is published.
1978 Research in Nursing & Health and Advances in Nursing Science are published.
1979 Western Journal of Nursing Research is published.
1980s-1990s Sackett and colleagues developed methodologies to determine “best evidence” for practice.
1982-1983 Conduct and Utilization of Research in Nursing (CURN) Project is published.
1983 Annual Review of Nursing Research is published.
1985 National Center for Nursing Research (NCNR) is established to support and fund nursing research.
1987 Scholarly Inquiry for Nursing Practice is published.
1988 Applied Nursing Research and Nursing Science Quarterly are published.
1989 Agency for Health Care Policy and Research (AHCPR) is established and publishes EBP guidelines.
1990 Nursing Diagnosis, official journal of NANDA, is published; now entitled International Journal of Nursing
Terminologies and Classifications.
ANA established the American Nurses Credentialing Center (ANCC), which implemented the Magnet Hospital
Designation Program for Excellence in Nursing Services.
1992 Healthy People 2000 is published by the U.S. Department of Health and Human Services (DHHS).
Clinical Nursing Research is published.
1993 NCNR is renamed the National Institute of Nursing Research (NINR) to expand funding for nursing research.
Journal of Nursing Measurement is published.
Cochrane Collaboration is initiated providing systematic reviews and EBP guidelines (http://www.cochrane.org/).
1994 Qualitative Health Research is published.
1999 AHCPR is renamed Agency for Healthcare Research and Quality (AHRQ).
2000 Healthy People 2010 is published by DHHS.
Biological Research for Nursing is published.
2001 Stetler publishes her model Steps of Research Utilization to Facilitate Evidence-Based Practice.
2002 Joint Commission revises accreditation policies for hospitals supporting evidence-based health care.
NANDA becomes international—NANDA-I.
2004 Worldviews on Evidence-Based Nursing is published.
2011 NINR identifies mission and funding priorities (http://www.ninr.nih.gov/).
Healthy People 2020 is published; available at DHHS website http://www.healthypeople.gov/2020/
topicsobjectives2020/default.aspx.
2012 AHRQ identifies mission and funding priorities (http://www.ahrq.gov/).
American Nurses Association (ANA) Research Agenda is published.
 CHAPTER 2  Evolution of Research in Building Evidence-Based Nursing Practice 19
Nightingale’s research enabled her to instigate
attitudinal, organizational, and social changes. She
changed the attitudes of the military and society
toward the care of the sick. The military began to view
the sick as having the right to adequate food, suitable
quarters, and appropriate medical treatment, a change
that greatly reduced the mortality rate (Cook, 1913).
Nightingale improved the organization of army admin-
istration, hospital management, and hospital construc-
tion. Because of Nightingale’s research evidence and
influence, society began to accept responsibility for
testing public water, improving sanitation, preventing
starvation, and decreasing morbidity and mortality
rates (Palmer, 1977).
Early 1900s
From 1900 to 1950, research activities in nursing
were limited, but a few studies advanced nursing edu-
cation. These studies included the Nutting Report,
1912; Goldmark Report, 1923; and Burgess Report,
1926 (Abdellah, 1972; Johnson, 1977). On the basis
of recommendations of the Goldmark Report, more
schools of nursing were established in university set-
tings. The baccalaureate degree in nursing provided a
basis for graduate nursing education, with the first
master of nursing degree offered by Yale University
in 1929. Teachers College at Columbia University
offered the first doctoral program for nurses in 1923
and granted a degree in education (Ed.D.) to prepare
teachers for the profession. The Association of Col-
legiate Schools of Nursing, organized in 1932, pro-
moted the conduct of research to improve education
and practice. This organization also sponsored the
publication of the first research journal in nursing,
Nursing Research, in 1952 (Fitzpatrick, 1978).
A research trend that started in the 1940s and con-
tinued in the 1950s focused on the organization and
delivery of nursing services. Studies were conducted
on the numbers and kinds of nursing personnel, staff-
ing patterns, patient classification systems, patient and
nurse satisfaction, and unit arrangement. Types of care
such as comprehensive care, home care, and progres-
sive patient care were evaluated. These evaluations of
care laid the foundation for the development of self-
study manuals, which are similar to the quality assur-
ance manuals of today (Gortner & Nahm, 1977).
Nursing Research in the 1950s and 1960s
In 1950, the American Nurses Association (ANA) ini-
tiated a 5-year study on nursing functions and activi-
ties. The findings were reported in Twenty Thousand
Nurses Tell Their Story, and this study enabled the
ANA to develop statements on functions, standards,
and qualifications for professional nurses. Also during
this time, clinical research began expanding as spe-
cialty groups, such as community health, psychiatric,
medical-surgical, pediatric, and obstetrical nurses,
developed standards of care. The research conducted
by ANA and the specialty groups provided the basis
for the nursing practice standards that currently guide
professional nursing practice (Gortner & Nahm,
1977).
Educational studies were conducted in the 1950s
and 1960s to determine the most effective educational
preparation for the registered nurse. A nurse educator,
Mildred Montag, developed and evaluated the 2-year
nursing preparation (associate degree) in junior col-
leges. Student characteristics, such as admission and
retention patterns and the elements that promoted
success in nursing education and practice, were studied
for both associate and baccalaureate degree–prepared
nurses (Downs & Fleming, 1979).
In 1953, an Institute for Research and Service in
Nursing Education was established at Teachers Col­
lege, Columbia University, which provided research-
learning experiences for doctoral students (Werley,
1977). The American Nurse’s Foundation, chartered
in 1955, was responsible for receiving and administer-
ing research funds, conducting research programs,
consulting with nursing students, and engaging in
research. In 1956, a Committee on Research and
Studies was established to guide ANA research
(See, 1977).
A Department of Nursing Research was established
in the Walter Reed Army Institute of Research in 1957.
This was the first nursing unit in a research institution
that emphasized clinical nursing research (Werley,
1977). Also in 1957, the Southern Regional Educa-
tional Board (SREB), the Western Interstate Commis-
sion on Higher Education (WICHE), Midwest Nursing
Research Society (MNRS), and the New England
Board of Higher Education (NEBHE) were devel-
oped. These organizations are actively involved in
promoting research and disseminating the findings
today. ANA sponsored the first of a series of research
conferences in 1965, and the conference sponsors
required that the studies presented be relevant to
nursing and conducted by a nurse researcher (See,
1977). During the 1960s, a growing number of clinical
studies focused on quality care and the development
of criteria to measure patient outcomes. Intensive care
units were being developed, promoting the investiga-
tion of nursing interventions, staffing patterns, and
cost-effectiveness of care (Gortner & Nahm, 1977).
20 UNIT ONE  Introduction to Nursing Research
Nursing Research in the 1970s
In the 1970s, the nursing process became the focus of
many studies, with the investigations of assessment
techniques, nursing diagnoses classification, goal-
setting methods, and specific nursing interventions.
The first Nursing Diagnosis Conference, held in 1973,
evolved into the North American Nursing Diagnosis
Association (NANDA). In 2002, NANDA became
international and is now known as NANDA-I.
NANDA-I supports research activities focused on
identifying appropriate diagnoses for nursing and gen-
erating an effective diagnostic process. NANDA’s
journal, Nursing Diagnosis, was published in 1990
and was later renamed International Journal of
Nursing Terminology and Classifications. Details on
NANDA-I can be found on their website at http://
www.nanda.org/.
The educational studies of the 1970s evaluated
teaching methods and student learning experiences.
The National League for Nursing (NLN), founded in
1893, has had a major role in the conduct of research
to shape nursing education. Currently, NLN provides
programs, grants, and resources to advance nursing
education research in “pursuit of quality nursing edu-
cation for all types of nursing education programs”
(NLN, 2011; http://www.nln.org/aboutnln/index.
htm/). A number of studies were conducted to differ-
entiate the practices of nurses with baccalaureate and
associate degrees. These studies, which primarily
measured abilities to perform technical skills, were
ineffective in clearly differentiating between the two
levels of education.
Primary nursing care, which involves the delivery
of patient care predominantly by registered nurses
(RNs), was the trend for the 1970s. Studies were con-
ducted to examine the implementation and outcomes
of primary nursing care delivery models. The number
of nurse practitioners (NPs) and clinical nurse special-
ists (CNSs) with master’s degrees increased rapidly
during the 1970s. Limited research has been con-
ducted on the CNS role; however, the NP and nurse
midwifery roles have been researched extensively to
determine their positive impact on productivity,
quality, and cost of health care. In addition, those clini-
cians with master’s degrees acquired the background
to conduct research and to use research evidence in
practice.
In the 1970s, nursing scholars began developing
models, conceptual frameworks, and theories to guide
nursing practice. The works of these nursing theorists
also directed future nursing research. In 1978, a new
journal, Advances in Nursing Science, began publish-
ing the works of nursing theorists and the research
related to their theories. The number of doctoral pro-
grams in nursing and the number of nurses prepared
at the doctoral level greatly expanded in the 1970s
(Jacox, 1980). Some of the nurses with doctoral
degrees increased the conduct and complexity of
nursing research; however, many doctorally prepared
nurses did not become actively involved in research.
In 1970, the ANA Commission on Nursing Research
was established; in turn, this commission established
the Council of Nurse Researchers in 1972 to advance
research activities, provide an exchange of ideas, and
recognize excellence in research. The commission
also prepared position papers on subjects’ rights in
research and on federal guidelines concerning research
and human subjects, and it sponsored research pro-
grams nationally and internationally (See, 1977).
Federal funds for nursing research increased sig-
nificantly, with a total of slightly more than $39 million
awarded for research in nursing from 1955 to 1976.
Even though federal funding for nursing studies rose,
the funding was not comparable to the $493 million
in federal research funds received by those doing
medical research in 1974 alone (de Tornyay, 1977).
Sigma Theta Tau, the International Honor Society
for Nursing, sponsored national and international
research conferences, and the chapters of this organi-
zation sponsored many local conferences to promote
the dissemination of research findings. Image was a
journal initially published in 1967 by Sigma Theta Tau
now titled Journal of Nursing Scholarship, the journal
publishes many nursing studies and articles about
research methodology. A major goal of Sigma Theta
Tau is to advance scholarship in nursing by promoting
the conduct, communication, and use of research evi-
dence in nursing. The addition of two new research
journals in the 1970s, Research in Nursing & Health
in 1978 and Western Journal of Nursing Research in
1979, also increased the communication of nursing
research findings. However, the findings of many
studies conducted and published in the 1970s were not
being used in practice, so Stetler and Marram (1976)
developed a model to promote the communication and
use of research findings in practice.
Professor Archie Cochrane originated the concepts
of evidence-based practice with a book he published
in 1972 titled Effectiveness and Efficiency: Random
Reflections on Health Services. Cochrane advocated
the provision of health care based on research to
improve the quality of care and patient outcomes. To
facilitate the use of research evidence in practice, the
Cochrane Center was established in 1992, and the
Cochrane Collaboration in 1993. The Cochrane Col-
laboration and Library house numerous resources to
 CHAPTER 2  Evolution of Research in Building Evidence-Based Nursing Practice 21
promote EBP, such as systematic reviews of research
and evidence-based guidelines for practice (discussed
later in this chapter) (see the Cochrane Collaboration
at www.cochrane.org/).
Nursing Research in the 1980s and 1990s
The conduct of clinical nursing research was the focus
in the 1980s and 1990s. A variety of clinical journals
(Achieves of Psychiatric Nursing; Cancer Nursing;
Cardiovascular Nursing; Dimensions of Critical Care
Nursing; Heart & Lung; Journal of Neurosurgical
Nursing; Journal of Obstetric, Gynecologic, and Neo-
natal Nursing; Journal of Pediatric Nursing; Oncol-
ogy Nursing Forum; and Rehabilitation Nursing)
published a growing number of studies. One new
research journal was started in 1987, Scholarly Inquiry
for Nursing Practice, and two in 1988, Applied Nurs­
ing Research and Nursing Science Quarterly.
Even though the body of empirical knowledge gen-
erated through clinical research grew rapidly in the
1970s and 1980s, little of this knowledge was used
in practice. Two major projects were launched to
promote the use of research-based nursing interven-
tions in practice: the Western Interstate Commission
for Higher Education (WICHE) Regional Nursing
Research Development Project and the Conduct and
Utilization of Research in Nursing (CURN) Project.
In these projects, nurse researchers, with the assis-
tance of federal funding, designed and implemented
strategies for using research findings in practice. The
WICHE Project participants selected research-based
interventions for use in practice and then functioned
as change agents to implement the selected interven-
tion in a clinical agency. Because of the limited amount
of research that had been conducted, the project staff
and participants had difficulty identifying adequate
clinical studies with findings ready for use in practice
(Krueger, Nelson, & Wolanin, 1978).
The CURN Project was a 5-year venture (1975-
1980) directed by Horsley, Crane, Crabtree, and Wood
(1983) to increase the utilization of research findings
by (1) disseminating findings, (2) facilitating organi-
zational modifications necessary for implementation,
and (3) encouraging collaborative research that was
directly transferable to clinical practice. Research uti-
lization was seen as a process to be implemented by
an organization rather than by an individual nurse. The
Project team identified the activities of research utili-
zation to involve identification and synthesis of mul-
tiple studies in a common conceptual area (research
base) as well as transformation of the knowledge
derived from a research base into a solution or clinical
protocol. The clinical protocol was then transformed
into specific nursing actions (innovations) that were
administered to patients. The implementation of the
innovation was to be followed by clinical evaluation
of the new practice to ascertain whether it produced
the predicted result (Horsley et al., 1983). The clinical
protocols developed during the project were published
to encourage nurses in other healthcare agencies to use
these research-based intervention protocols in their
practice (CURN Project, 1981-1982).
To ensure that the studies were incorporated into
nursing practice, the findings needed to be synthesized
for different topics. In 1983, the first volume of the
Annual Review of Nursing Research was published
(Werley & Fitzpatrick, 1983). This annual publication
contains experts’ reviews of research in selected areas
of nursing practice, nursing care delivery, nursing edu-
cation, and the profession of nursing. The Annual
Review of Nursing Research continues to be published
each year to (1) expand the synthesis and dissemina-
tion of research findings, (2) promote the use of
research findings in practice, and (3) identify direc-
tions for future research.
Many nurses obtained master’s and doctoral
degrees during the 1980s and 1990s, and postdoctoral
education was encouraged for nurse researchers. The
ANA (1989) stated that nurses at all levels of educa-
tion have a role in research, which extends from
reading research to conducting complex, funded pro-
grams of research (see Chapter 1). Another priority of
the 1980s and 1990s was to obtain greater funding for
nursing research. Most of the federal funds in the
1980s were designated for studies involving the diag-
nosis and cure of diseases. Therefore, nursing received
a small percentage of the federal research and devel-
opment (R&D) funds (approximately 2% to 3%) com-
pared with medicine (approximately 90%), even
though nursing personnel greatly outnumbered medi­
cal personnel (Larson, 1984). However, in 1985, the
ANA achieved a major political victory for nursing
research with the creation of the National Center for
Nursing Research (NCNR) within the National Insti-
tutes of Health (NIH). This center was created after
years of work and two presidential vetoes (Bauknecht,
1986). The purpose of the National Center was to
support the conduct of basic and clinical nursing
research and the dissemination of findings. With its
creation, nursing research had visibility at the federal
level for the first time. In 1993, during the tenure of
its first director, Dr. Ada Sue Hinshaw, the NCNR
became the National Institute of Nursing Research
(NINR). This change in title enhanced the recognition
of nursing as a research discipline and expanded the
funding for nursing research.
22 UNIT ONE  Introduction to Nursing Research
Outcomes research emerged as an important meth-
odology for documenting the effectiveness of health-
care services in the 1980s and 1990s. This type of
research evolved from the quality assessment and
quality assurance functions that originated with the
professional standards review organizations (PSROs)
in 1972. During the 1980s, William Roper, the director
of the Health Care Finance Administration (HCFA),
promoted outcomes research for determining the
quality and cost-effectiveness of patient care (Johnson,
1993).
In 1989, the Agency for Health Care Policy and
Research (AHCPR) was established to facilitate the
conduct of outcomes research (Rettig, 1991). This
Agency also had an active role in communicating
research findings to healthcare practitioners and was
responsible for publishing the first evidence-based
national clinical practice guidelines in 1989. Several
of these guidelines, including the latest research find-
ings with directives for practice, were published in the
1990s. The Healthcare Research and Quality Act of
1999 reauthorized the AHCPR, changing its name to
the Agency for Healthcare Research and Quality
(AHRQ, 2012). This significant change positioned the
AHRQ as a scientific partner with the public and
private sectors to improve the quality and safety of
patient care by promoting the use of the best research
evidence available in practice.
Building on the process of research utilization,
physicians, nurses, and other healthcare professions
focused on the development of EBP during the 1990s.
A research group led by Dr. David Sackett at McMas-
ter University in Canada developed explicit research
methodologies to determine the “best evidence” for
practice. The term evidence based was first used by
David Eddy in 1990, with the focus on providing EBP
for medicine (Craig & Smyth, 2012; Sackett, Straus,
Richardson, Rosenberg, & Haynes, 2000).
In 1990, the ANA leaders established the American
Nursing Credentialing Center (ANCC) and approved
a recognition program for hospitals called the Magnet
Hospital Designation Program for Excellence in
Nursing Services (ANCC, 2012). This program has
evolved over the last 20 years but has remained true
to its commitment to promote research conducted by
nurses in clinical settings and to support implementa-
tion of care based on the best current research
evidence.
Nursing Research in the 21st Century
The vision for nursing research in the 21st century
includes conducting quality studies through the use of
a variety of methodologies, synthesizing the study
findings into the best research evidence, using this
research evidence to guide practice, and examining the
outcomes of EBP (Brown, 2009; Craig & Smyth, 2012;
Doran, 2011; Melnyk & Fineout-Overholt, 2011). The
focus on EBP has become stronger over the last decade.
In 2002, the Joint Commission on Accreditation of
Healthcare Organizations (JCAHO) revised the accred­
itation policies for hospitals to support the implemen-
tation of evidence-based health care. To facilitate the
movement of nursing toward EBP in clinical agencies,
Stetler (2001) developed her Research Utilization to
Facilitate EBP Model (see Chapter 19 for a description
of this model). The focus on EBP in nursing was sup-
ported with the initiation of the Worldviews on
Evidence-Based Nursing journal in 2004.
The focus of healthcare research and funding has
expanded from the treatment of illness to include
health promotion and illness prevention. Healthy
People 2000 and Healthy People 2010, documents
published by the U.S. Department of Health and
Human Services (U.S. DHHS 1992, 2000), have
increased the visibility of health promotion goals and
research. Health People 2020 (U.S. DHHS, 2012)
information is now available at the department’s
website, http://www.healthypeople.gov/2020/. Some
of the new topics covered by Healthy People 2020
include: adolescent health; blood disorders and blood
safety; dementias; early and middle childhood; genom-
ics; global health; healthcare-associated infections;
lesbian, gay, bisexual, and transgender health; older
adults; preparedness; sleep health; and social determi-
nants of health. In the next decade, nurse researchers
will have a major role in the development of interven-
tions to promote health and prevent illness in individu-
als, families, and communities.
The AHRQ has been designated the lead agency
supporting research designed to improve the quality
of health care, reduce its cost, improve patient safety,
decrease medical errors, and broaden access to essen-
tial services. The AHRQ sponsors and conducts
research that provides evidence-based information on
healthcare outcomes, quality, cost, use, and access.
This research information promotes effective health-
care decision making by patients, clinicians, health
system executives, and policy makers. The three future
goals of the AHRQ are focused on the following:
Safety and quality: Reduce the risk of harm by
promoting delivery of the best possible health
care.
Effectiveness: Improve healthcare outcomes
by encouraging the use of evidence to make
informed healthcare decisions.
 CHAPTER 2  Evolution of Research in Building Evidence-Based Nursing Practice 23
Efficiency: Transform research into practice to
facilitate wider access to effective healthcare
services and reduce unnecessary costs. (AHRQ,
2012)
AHRQ identifies funding priorities and research
findings on their website at http://www.ahrq.gov/.
Currently, the AHRQ and NINR work collaboratively
to promote funding for nursing studies. These agen-
cies often jointly call for proposals for studies of high
priority to both agencies.
NINR is one of the most influential organizations
committed to providing funding, support, and educa-
tion to advance research in nursing. The current
mission, goals, research priorities, and strategies of
NINR are as follows:
The mission of the NINR is to promote and
improve the health of individuals, families, com-
munities, and populations. NINR supports and
conducts clinical and basic research and
research training on health and illness across
the lifespan. The research focus encompasses
health promotion and disease prevention, quality
of life, health disparities, and end-of-life. NINR
seeks to extend nursing science by integrating
the biological and behavioral sciences, employ-
ing new technologies to research questions,
improving research methods, and developing the
scientists of the future. (NINR, 2011; http://
www.ninr.nih.gov/AboutNINR/NINRMissionand
StrategicPlan/)
The NINR has supported the development of nurse
scientists in genetics and genomics and sponsored the
Summer Genetics Institute to expand nurses’ contribu-
tions to genetic research. The funding priorities,
funding process, and current research findings are
available on the NINR website at http://www.ninr
.nih.gov/.
The mission of ANA is to ensure the advancement
of nurses in their profession to improve health for
all. Central to this mission is the promotion of quality
outcomes that require the use of research to provide
EBP. ANA’s (2012) research agenda can be viewed
online. To accomplish this agenda, we need to ensure
an effective research enterprise in nursing by (1)
creating a research culture; (2) providing quality edu-
cational (baccalaureate, master’s, doctoral, and post-
doctoral) programs to prepare a workforce of nurse
scientists; (3) developing a sound research infrastruc-
ture; and (4) obtaining sufficient funding for essential
research (ANA, 2012; AACN, 1999, 2012). With this
professional support, nurses can conduct studies
using a variety of research methodologies to generate
the essential knowledge needed to promote EBP and
quality health outcomes for all.
Methodologies for Developing
Research Evidence in Nursing
Scientific method incorporates all procedures that sci-
entists have used, currently use, or may use in the
future to pursue knowledge (Kaplan, 1964). This
broad definition dispels the belief that there is one way
to conduct research and embraces the use of both
quantitative and qualitative research methodologies in
developing research evidence for practice.
Since the 1930s, many researchers have narrowly
defined scientific method to include only quantitative
research. This research method is based in the philoso-
phy of logical empiricism or positivism (Norbeck,
1987; Scheffler, 1967). Therefore, scientific knowl-
edge is generated through an application of logical
principles and reasoning whereby the researcher
adopts a distant and noninteractive posture with the
research subject to prevent bias (Silva & Rothbart,
1984). Thus, quantitative research is best defined as
a formal, objective, systematic process implemented
to obtain numerical data for understanding aspects of
the world. This research method is used to describe
variables, examine relationships among variables, and
determine cause-and-effect interactions between vari-
ables (Kerlinger & Lee, 2000; Shadish, Cook, &
Campbell, 2002). Currently, the predominantly used
method of scientific investigation in nursing is quan-
titative research.
Qualitative research is a systematic, interactive,
subjective, holistic approach used to describe life
experiences and give them meaning (Marshall &
Rossman, 2011; Munhall, 2012). Qualitative research
is not a new idea in the social and behavioral sciences
(Baumrind, 1980; Glaser & Strauss, 1967). This type
of research is conducted to explore, describe, and
promote understanding of human experiences, events,
and cultures over time.
Comparison of Quantitative Research
and Qualitative Research
The quantitative and qualitative types of research
complement each other because they generate differ-
ent kinds of knowledge that are useful in nursing prac-
tice. The problem and purpose to be studied determine
the type of research to be conducted, and the research-
er’s knowledge of both types of research promotes
24 UNIT ONE  Introduction to Nursing Research
TABLE 2-2  Characteristics of Quantitative and Qualitative Research Methods
Characteristic Quantitative Research
Philosophical origin Logical positivism, post positivism
Focus Concise, objective, reductionistic
Reasoning Logistic, deductive
Basis of knowing Cause-and-effect relationships
Theoretical focus Tests theory
Researcher involvement Control
Methods of measurement Structured interviews, questionnaires,
observations, scales, physiological measures
Data Numbers
Analysis Statistical analysis
Findings Acceptance or rejection of theoretical propositions
Generalization
accurate selection of the methodology for the problem
identified (Creswell, 2009). Quantitative and qualita-
tive research methodologies have some similarities,
because both require researcher expertise, involve
rigor in implementation, and result in the generation
of scientific knowledge for nursing practice. Some of
the differences between the two methodologies are
presented in Table 2-2. Some researchers include both
quantitative and qualitative research methodologies in
their studies, an approach referred to as mixed methods
research (see Chapter 10).
Philosophical Origins of Quantitative
and Qualitative Research Methods
The quantitative approach to scientific inquiry emerged
from a branch of philosophy called logical positivism,
which operates on strict rules of logic, truth, laws,
axioms, and predictions. Quantitative researchers hold
the position that truth is absolute and that there is a
single reality that one could define by careful measure-
ment. To find truth as a quantitative researcher, you
need to be completely objective, meaning that your
values, feelings, and personal perceptions cannot enter
into the measurement of reality. Quantitative research-
ers believe that all human behavior is objective, pur-
poseful, and measurable. The researcher needs only
to find or develop the “right” instrument or tool to
measure the behavior.
Today, however, many nurse researchers base their
quantitative studies on more of a postpositivist phi-
losophy (Clark, 1998). This philosophy evolved from
positivism but focuses on the discovery of reality that
is characterized by patterns and trends that can be used
to describe, explain, and predict phenomena. With
Qualitative Research
Naturalistic, interpretive, humanistic
Broad, subjective, holistic
Dialectic, inductive
Meaning, discovery, understanding
Develops theory and frameworks
Shared interpretation
Unstructured interviews, observations, focus groups
Words
Text-based analysis
Uniqueness, dynamic, understanding of phenomena,
new theory, models, and/or frameworks
postpositivism, “truth can be discovered only imper-
fectly and in a probabilistic sense, in contrast to the
positivist ideal of establishing cause-and-effect expla-
nations of immutable facts” (Ford-Gilboe, Campbell,
& Berman, 1995, p. 16). The postpositivist approach
also rejects the idea that the researcher is completely
objective about what is to be discovered but continues
to emphasize the need to control environmental influ-
ences (Newman, 1992; Shadish et al., 2002).
Qualitative research is an interpretive method­
ological approach that values more of a subjective
science than quantitative research. Qualitative research
evolved from the behavioral and social sciences as a
method of understanding the unique, dynamic, holistic
nature of human beings. The philosophical base of
qualitative research is interpretive, humanistic, and
naturalistic and is concerned with helping those
involved to understand the meaning of their social
interactions. Qualitative researchers believe that truth
is both complex and dynamic and can be found only
by studying persons as they interact with and within
their sociohistorical settings (Marshall & Rossman,
2011; Munhall, 2012).
Focuses of Quantitative and
Qualitative Research Methods
The focus or perspective for quantitative research is
usually concise and reductionistic. Reductionism
involves breaking the whole into parts so that the parts
can be examined. Quantitative researchers remain
detached from the study and try not to influence it with
their values (objectivity). Researcher involvement in
the study is thought to bias or sway the study toward
the perceptions and values of the researcher, and
 CHAPTER 2  Evolution of Research in Building Evidence-Based Nursing Practice 25
biasing a study is considered poor scientific technique
(Creswell, 2009; Kerlinger & Lee, 2000; Shadish et
al., 2002).
The focus of qualitative research is usually broad,
and the intent is to give meaning to the whole (holis-
tic). The qualitative researcher has an active part in
the study and acknowledges that personal values and
perceptions may influence the findings. Thus, this
research approach is subjective, because the approach
assumes that subjectivity is essential for understand-
ing human experiences (Marshall & Rossman, 2011;
Munhall, 2012).
Uniqueness of Conducting Quantitative
Research and Qualitative Research
Quantitative research describes and examines rela-
tionships and determines causality among variables.
Thus, this method is useful for testing a theory by
testing the validity of the relationships that compose
the theory (Creswell, 2009). Quantitative research
incorporates logistic, deductive reasoning as the
researcher examines particulars to make generaliza-
tions about the universe.
Qualitative research generates knowledge about
meaning through discovery. Inductive reasoning and
dialectic reasoning are predominant in these studies.
For example, the qualitative researcher studies the
whole person’s response to pain by examining prem-
ises about human pain and determining the meaning
that pain has for a particular person. Because qualita-
tive research is concerned with meaning and under-
standing, researchers using qualitative approaches
may identify relationships among the variables, and
these relational statements may be used to develop and
extend theories.
Quantitative research requires control (see Table
2-2). The investigator uses control to identify and limit
the problem to be researched and attempts to limit the
effects of extraneous or other variables that are not the
focus of the study. For example, as a quantitative
researcher, you might study the effects of nutritional
education on serum lipid levels (total serum cho­
lesterol, low-density lipoprotein [LDL] cholesterol,
high-density lipoprotein [HDL] cholesterol, and tri-
glycerides). You would control the educational
program by manipulating the type of education pro-
vided, the teaching methods, the length of the program,
the setting for the program, and the instructor. The
nutritional program might be consistently imple-
mented with the use of DVDs shown to subjects in a
structured setting. You could also control other extra-
neous variables, such as participant’s age, history of
cardiovascular disease, and exercise level, because
these extraneous variables might affect the serum lipid
levels. The intent of this control is to more precisely
examine the effects of nutritional education on serum
lipid levels.
Quantitative research also requires the use of (1)
structured interviews, questionnaires, or observations,
(2) scales, and (3) physiological measures that gener-
ate numerical data. Statistical analyses are conducted
to reduce and organize data, describe variables,
examine relationships, and determine differences
among groups. Control, precise measurement methods,
and statistical analyses are used to ensure that the
research findings accurately reflect reality so that the
study findings can be generalized. Generalization
involves the application of trends or general tenden-
cies (which are identified by studying a sample) to the
population from which the research sample was
drawn. Researchers must be cautious in making gen-
eralizations, because a sound generalization requires
the support of many studies with a variety of samples
(Shadish et al., 2002).
Qualitative researchers use observations, inter-
views, and focus groups to gather data. The interac-
tions are guided but not controlled in the way that
quantitative data collection is controlled. For example,
the researcher may ask subjects to share their experi-
ences of powerlessness in the healthcare system. Qual-
itative researchers would begin interpreting the
subjective data during data collection, recognizing
that their interpretation is influenced by their own
perceptions and beliefs (Munhall, 2012).
Qualitative data take the form of words and are
analyzed according to the qualitative approach that is
being used. The intent of the analysis is to organize
the data into a meaningful, individualized interpreta-
tion, framework, or theory that describes the phenom-
enon studied. The findings from a qualitative study are
unique to that study, and it is not the researcher’s
intent to generalize the findings to a larger population.
Qualitative researchers are encouraged to question
generalizations and to interpret meaning based on
individual study participants’ perceptions and realities
(Munhall, 2012).
Classification of Research
Methodologies Presented
in this Text
Research methods used frequently in nursing can be
classified in different ways, so a classification system
26 UNIT ONE  Introduction to Nursing Research
Box 2-1 Classification of Research
Methods for this Textbook
Types of quantitative research
Descriptive research
Correlational research
Quasi-experimental research
Experimental research
Types of qualitative research
Phenomenological research
Grounded theory research
Ethnographic research
Exploratory-descriptive qualitative research
Historical research
Outcomes research
Intervention research
was developed for this textbook and is presented in
Box 2-1. This textbook includes quantitative, qualita-
tive, outcomes, and intervention methods of research.
The quantitative research methods are classified into
four categories: (1) descriptive, (2) correlational, (3)
quasi-experimental, and (4) experimental. Types of
quantitative research are used to test theories and gen-
erate and refine knowledge for nursing practice. Quan-
titative research methods are introduced in this section
and described in more detail in Chapter 3.
The qualitative research methods included in this
textbook are (1) phenomenological research, (2)
grounded theory research, (3) ethnographic research,
(4) exploratory-descriptive qualitative research, and
(5) historical research. These approaches, all method-
ologies for discovering knowledge, are introduced in
this section and described in depth in Chapters 4 and
12. Unit Two of this textbook focuses on understand-
ing the research process and includes discussions of
both quantitative and qualitative research.
Quantitative Research Methods
Descriptive Research
Descriptive research provides an accurate portrayal
or account of characteristics of a particular individual,
situation, or group (Kerlinger & Lee, 2000). Descrip-
tive studies offer researchers a way to (1) discover
new meaning, (2) describe what exists, (3) determine
the frequency with which something occurs, and (4)
categorize information. Descriptive studies are usually
conducted when little is known about a phenomenon
and provide the basis for the conduct of correlational,
quasi-experimental, and experimental studies (Cres­
well, 2009).
Correlational Research
Correlational research involves the systematic
investigation of relationships between or among two
or more variables that have been identified in theories,
observed in practice, or both. If the relationships exist,
the researcher determines the type (positive or nega-
tive) and the degree or strength of the relationships.
The primary intent of correlational studies is to explain
the nature of relationships, not to determine cause and
effect. However, correlational studies are the means
for generating hypotheses to guide quasi-experimental
and experimental studies that focus on examining
cause-and-effect interactions.
Quasi-Experimental Research
The purposes of quasi-experimental studies are (1)
to identify causal relationships, (2) to examine the
significance of causal relationships, (3) to clarify why
certain events happened, or (4) a combination of these
objectives (Shadish et al., 2002). These studies test the
effectiveness of nursing interventions that can then be
implemented to improve patient and family outcomes
in nursing practice.
Quasi-experimental studies are less powerful than
experimental studies because they involve a lower
level of control in at least one of three areas: (1)
manipulation of the treatment or independent variable,
(2) manipulation of the setting, and (3) selection of
subjects. When studying human behavior, especially
in clinical areas, researchers are commonly unable to
manipulate or control certain variables. Also, subjects
are usually not randomly selected but are selected on
the basis of convenience. Thus, as a nurse researcher
you will probably conduct more quasi-experimental
than experimental studies.
Experimental Research
Experimental research is an objective, systematic,
controlled investigation conducted for the purpose of
predicting and controlling phenomena. This type of
research examines causality (Shadish et al., 2002).
Experimental research is considered the most power-
ful quantitative method because of the rigorous control
of variables. Experimental studies have three main
characteristics: (1) a controlled manipulation of at
least one treatment variable (independent variable),
(2) administration of the treatment to some of the
subjects in the study (experimental group) and not to
others (control group), and (3) random selection of
subjects or random assignment of subjects to groups,
 CHAPTER 2  Evolution of Research in Building Evidence-Based Nursing Practice 27
or both. Experimental studies usually are conducted in
highly controlled settings, such as laboratories or
research units in clinical agencies. A randomized con-
trolled trial (RCT) is a type of experimental research
that produces the strongest research evidence for
practice.
Qualitative Research Methods
Phenomenological Research
Phenomenological research is a humanistic study of
phenomena. The aim of phenomenology is to explore
an experience as it is lived by the study participants
and interpreted by the researcher. During the study, the
researcher’s experiences, reflections, and interpreta-
tions influence the data collected from the study par-
ticipants (Munhall, 2012). Thus, the participants’ lived
experiences are expressed through the researcher’s
interpretations that are obtained from immersion in the
study data and the underlying philosophy of the phe-
nomenological study. Phenomenological research is
an effective methodology for discovering the meaning
of a complex experience as it is lived by a person, such
as the lived experience of chronic illness.
Grounded Theory Research
Grounded theory research is an inductive research
method initially described by Glaser and Strauss
(1967). This research approach is useful for discover-
ing what problems exist in a social setting and the
processes people use to handle them. Grounded theory
is particularly useful when little is known about the
area to be studied or when what is known does not
provide a satisfactory explanation. Grounded theory
methodology emphasizes interaction, observation, and
development of relationships among concepts.
Throughout the study, the researcher explores, pro-
poses, formulates, and validates relationships among
the concepts until a theory evolves. The theory devel-
oped is “grounded,” in or has its roots in, the data from
which it was derived (Wuest, 2012).
Ethnographic Research
Ethnographic research was developed by anthro-
pologists to investigate cultures through in-depth
study of the members of the cultures. This type of
research attempts to tell the story of people’s daily
lives while describing the culture in which they live.
The ethnographic research process is the systematic
collection, description, and analysis of data to develop
a description of cultural behavior. The researcher (eth-
nographer) actually lives in or becomes a part of the
cultural setting to gather the data. Through the use of
ethnographic research, different cultures are described,
compared, and contrasted to add to our understanding
of the impact of culture on human behavior and health
(Wolf, 2012).
Exploratory-Descriptive Qualitative Research
Exploratory-descriptive qualitative research is
conducted to address an issue or problem in need of a
solution and/or understanding. Qualitative nurse
researchers explore an issue or problem area using
varied qualitative techniques with the intent of describ-
ing the topic of interest and promoting understanding.
Although the studies result in descriptions and could
be labeled as descriptive qualitative studies, most of
the researchers are in the exploratory stage of studying
the area of interest. This type of qualitative research
usually lacks a clearly identified qualitative methodol-
ogy, such as phenomenology, grounded theory, or eth-
nography. In this text, studies that the researchers
identified as being qualitative without indicating a
specific approach like phenomenology or grounded
theory will be labeled as being exploratory-descriptive
qualitative studies.
Historical Research
Historical research is a narrative description or
analysis of events that occurred in the remote or
recent past. Data are obtained from records, artifacts,
or verbal reports. Through historical research,
nursing has a way of understanding the discipline
and interpreting its contributions to health care and
society. Initial historical research focused on nursing
leaders, such as Nightingale, and her contributions to
nursing research and practice. In addition, the mis-
takes of the past can be examined to help nurses
understand and respond to present situations affect-
ing nurses and nursing practice. Thus, historical
research has the potential to provide a foundation for
and to direct the future movements of the profession
(Lundy, 2012).
Outcomes Research
The spiraling cost of health care has generated many
questions about the quality and effectiveness of health-
care services and the patient outcomes. Consumers
want to know what services they are buying and
whether these services will improve their health.
Healthcare policy makers want to know whether the
care is cost-effective and high quality. These concerns
have promoted the development of outcomes
research, which examines the results of care and mea-
sures the changes in health status of patients (AHRQ,
2012; Doran, 2011). Key ideas related to outcomes
28 UNIT ONE  Introduction to Nursing Research
research are addressed throughout the text, and
Chapter 13 contains a detailed discussion of this
methodology.
Intervention Research
Intervention research investigates the effectiveness
of a nursing intervention in achieving the desired
outcome or outcomes in a natural setting. “Interven-
tions are defined as treatments, therapies, procedures,
or actions implemented by health professionals to and
with clients, in a particular situation, to move the
clients’ condition toward desired health outcomes that
are beneficial to the clients” (Sidani & Braden, 1998,
p. 8). An intervention can be a specific treatment
implemented to manage a well-defined patient problem
or a program. A program intervention, such as a
cardiac rehabilitation program, consists of multiple
nursing actions that are implemented as a package to
improve the health conditions of the participants
(Brown, 2002; Forbes, 2009). The goal of intervention
research is to generate sound scientific knowledge for
actions or interventions that nurses can use to provide
evidence-based nursing care. The details of interven-
tion research are presented in Chapter 14. In summary,
nurse researchers conduct a variety of research meth-
odologies (quantitative, qualitative, outcomes, and
intervention research) to develop the best research
evidence for practice.
Introduction to Best Research
Evidence for Practice
EBP involves the use of best research evidence to
support clinical decisions in practice. As a nurse, you
make numerous clinical decisions each day that affect
the health outcomes of your patients and their fami-
lies. By using the best research evidence available,
you can make quality clinical decisions that will
improve the health outcomes for patients, families,
and communities. This section introduces you to the
concept of best research evidence for practice by pro-
viding (1) a definition of the term best research evi-
dence, (2) a model of the levels of research evidence
available, and (3) a link of the best research evidence
to evidence-based guidelines for practice.
Definition of Best Research Evidence
Best research evidence is a summary of the highest-
quality, current empirical knowledge in a specific area
of health care that is developed from a synthesis of
quality studies (quantitative, qualitative, outcomes,
and intervention) in that area. The synthesis of study
findings is a complex, highly structured process that
is conducted most effectively by at least two research-
ers or even a team of expert researchers and healthcare
providers. There are various types of research synthe-
ses, and the type of synthesis conducted varies accord-
ing to the quality and types of research evidence
available.
The quality of the research evidence available in an
area depends on the number and strength of the studies.
Replicating or repeating of studies with similar meth-
odology adds to the quality of the research evidence.
The strengths and weaknesses of the studies are deter-
mined by critically appraising the validity or credibil-
ity of the study outcomes (see Chapter 18). The types
of research commonly conducted in nursing were
identified earlier in this chapter as quantitative, quali-
tative, outcomes, and intervention (see Box 2-1). The
research synthesis process used to summarize knowl-
edge varies for quantitative and qualitative research
methods. In building the best research evidence for
practice, the quantitative experimental study, such as
an RCT, has been identified as producing the strongest
research evidence for practice (Craig & Smyth, 2012;
Institute of Medicine, 2001; Melnyk & Fineout-
Overholt, 2011; Sackett et al., 2000).
Research evidence in nursing and health care is
synthesized by using the following processes: (1) sys-
tematic review, (2) meta-analysis, (3) meta-synthesis,
and (4) mixed methods systematic review. Depending
on the quantity and strength of the research findings
available, nurses and healthcare professionals use one
or more of these four synthesis processes to determine
the current best research evidence in an area. Table 2-3
identifies the processes used in research synthesis, the
purpose of each synthesis process, the types of research
included in the synthesis (sampling frame), and the
analysis techniques used to achieve the synthesis of
research evidence (Craig & Smyth, 2012; Sandelowski
& Barroso, 2007; Whittemore, 2005).
A systematic review is a structured, comprehen-
sive synthesis of the research literature to determine
the best research evidence available to address a
healthcare question. A systematic review involves
identifying, locating, appraising, and synthesizing
quality research evidence for expert clinicians to use
to promote an EBP (Craig & Smyth, 2012; Higgins &
Green, 2008). Teams of expert researchers, clinicians,
and sometimes students conduct these reviews to
determine the current best knowledge for use in
practice. Systematic reviews are also used in the
development of national and international standard-
ized guidelines for managing health problems such as
depression, hypertension, and type 2 diabetes. The
 CHAPTER 2  Evolution of Research in Building Evidence-Based Nursing Practice 29
TABLE 2-3  Processes Used to Synthesize Research Evidence
Synthesis
Process Purpose of Synthesis
Systematic Use of specific, systematic methods to identify,
review select, critically appraise, and synthesize
research evidence to address a particular
problem in practice (Craig & Smyth, 2012;
Higgins & Green, 2008).
Meta-analysis Synthesis or pooling of the results from several
previous studies using statistical analysis to
determine the effect of an intervention or the
strength of relationships (Higgins & Green,
2008).
Meta-synthesis Systematic compiling and integration of
qualitative studies to expand understanding
and develop a unique interpretation of the
studies’ findings in a selected area (Barnett-
Page & Thomas, 2009; Finfgeld-Connett,
2010; Sandelowski & Barroso, 2007).
Mixed methods Synthesis of the findings from independent
systematic studies conducted with a variety of methods
review (both quantitative and qualitative) to
determine the current knowledge in an area
(Higgins & Green, 2008).
processes for critically appraising and conducting sys-
tematic reviews are detailed in Chapter 19.
A meta-analysis is conducted to statistically pool
the results from previous studies into a single quantita-
tive analysis that provides one of the highest levels of
evidence about an intervention’s effectiveness (Andrel,
Keith, & Leiby, 2009; Craig & Smyth, 2012; Higgins
& Green, 2008). The studies synthesized are usually
quasi-experimental or experimental types of studies.
In addition, a meta-analysis can be performed on cor-
relational studies to determine the type (positive or
negative) or strength of relationships among selected
variables (see Table 2-3). Because meta-analyses
involve statistical analysis to combine study findings,
it is possible to be objective rather than subjective in
synthesizing research evidence. Some of the strongest
evidence for using an intervention in practice is gener-
ated from a meta-analysis of multiple, controlled
quasi-experimental and experimental studies. Thus,
many systematic reviews conducted to generate
evidence-based guidelines include meta-analyses. The
process for conducting a meta-analysis is presented in
Chapter 19.
Qualitative research synthesis is the process and
product of systematically reviewing and formally
integrating the findings from qualitative studies
Analysis for
Types of Research Included in the Achieving
Synthesis (Sampling Frame) Synthesis
Usually includes quantitative studies with Narrative and
similar methodology, such as statistical
randomized controlled trials (RCTs), and
can also include meta-analyses focused
on an area of the practice problem.
Includes quantitative studies with similar Statistical
methodology, such as quasi-experimental
and experimental studies focused on the
effect of an intervention or correlational
studies focused on relationships.
Uses original qualitative studies and Narrative
summaries of qualitative studies to
produce the synthesis.
Synthesis of a variety of quantitative, Narrative
qualitative, and mixed methods studies.
(Sandelowski & Barroso, 2007). The process for con-
ducting a synthesis of qualitative research is still in
the developmental phase, and a variety of synthesis
methods have appeared in the literature (Barnett-Page
& Thomas, 2009; Finfgeld-Connett, 2010; Higgins &
Green, 2008). In this text, the concept meta-synthesis
is used to describe the process for synthesizing quali-
tative research. Meta-synthesis is defined as the sys-
tematic compiling and integration of qualitative study
results to expand understanding and develop a unique
interpretation of study findings in a selected area. The
focus is on interpretation rather than the combining of
study results as with quantitative research synthesis
(see Table 2-3). The process for conducting a meta-
synthesis is presented in Chapter 19.
Over the past 10 to 15 years, nurse researchers
have conducted mixed methods studies (previously
referred to as triangulation studies) that include
both quantitative and qualitative research methods
(Creswell, 2009). In addition, determining the current
research evidence in an area might require synthesiz-
ing both quantitative and qualitative studies. Higgins
and Green (2008) refer to this synthesis of quantita-
tive, qualitative, and mixed methods studies as a
mixed methods systematic review (see Table 2-3).
Mixed methods systematic reviews might include a
30 UNIT ONE  Introduction to Nursing Research

variety of study designs, such as qualitative research
and quasi-experimental, correlational, and/or descrip-
tive studies (Higgins & Green, 2008). Some research-
ers have conducted syntheses of quantitative and/or
qualitative studies and called them “integrative
reviews of research.” In this text, the synthesis of a
variety of quantitative and qualitative study findings
is referred to as mixed methods systematic reviews.
The value of these reviews depends on the standards
used to the conduct them. The process for conducting
a mixed method systematic review is discussed in
Chapter 19.
Levels of Research Evidence
The strength or validity of the best research evidence
in an area depends on the quality and quantity of the
studies conducted in the area. Quantitative studies,
especially experimental studies like RCTSs, are
thought to provide the strongest research evidence. In
addition, the replication of studies with similar meth-
odology increases the strength of the research
evidence generated. The levels of the research evi-
dence can be visualized as a continuum with the
highest quality of research evidence at one end and
weakest research evidence at the other (see Figure 2-1)
(Craig & Smyth, 2012; Higgins & Green, 2008;
Melnyk & Fineout-Overholt, 2011). The systematic
research reviews and meta-analyses of high-quality
experimental studies provide the strongest or best
research evidence for use by expert clinicians in prac-
tice. Meta-analyses and integrative reviews of quasi-
experimental and experimental studies also provide

Strongest or Best Research Evidence

Systematic Review of experimental studies (well-designed randomized

controlled trials [RCTs]) and meta-analyses

Meta-analyses of experimental (RCT) and quasi-experimental studies

Integrative Reviews of experimental (RCT) and quasi-experimental

Single Experimental study (RCT)

Single Quasi-experimental study

Meta-analysis of correlational studies

Integrative Reviews of correlational and descriptive studies

Mixed methods systematic review of quantitatives, qualitatives, and mixed methods studies

Qualitative Research Meta-synthesis

Single Correlational study

Single Qualitative or Descriptive study

Opinions of respected authorities based upon clinical evidence, reports

of expert committees

Weakest Research Evidence

Figure 2-1  Levels of research evidence.

 CHAPTER 2  Evolution of Research in Building Evidence-Based Nursing Practice 31
strong research evidence for managing practice prob-
lems. Correlational, descriptive, and qualitative studies
direct further research and provide some useful find-
ings for practice (see Figure 2-1). The weakest evi-
dence comes from expert opinions, which can include
expert clinicians’ opinions or the opinions expressed
in committee reports. When making a decision in your
clinical practice, be sure to base your decision on the
best research evidence available.
The levels of research evidence identified in Figure
2-1 help nurses determine the quality and validity of
the evidence that is available for them to use in prac-
tice. Advance practice nurses must seek out the best
research knowledge available in an area to ensure that
they manage patients’ acute and chronic illnesses with
quality care (Craig & Smyth, 2012; Higgins & Green,
2008; Melnyk & Fineout-Overholt, 2011). This best
research evidence generated from systematic reviews,
meta-analyses, and mixed methods systematic reviews
is used most often to develop standardized or evidence-
based guidelines for practice.
Introduction to Evidence-Based
Practice Guidelines
Evidence-based practice guidelines are rigorous,
explicit clinical guidelines that are based on the best
research evidence available in that area. These guide-
lines are usually developed by a team or panel of
expert researchers; expert clinicians (physicians,
nurses, pharmacists, and other health professionals);
and sometimes consumers, policy makers, and econo-
mists. The expert panel seeks consensus on the content
of the guideline to provide clinicians with the best
information for making clinical decisions in practice.
There has been a dramatic growth in the production
of EBP guidelines to assist healthcare providers in
building an EBP and in improving healthcare out-
comes for patients, families, providers, and healthcare
agencies.
Every year, new guidelines are developed, and
some of the existing guidelines are revised on the basis
of new research evidence. These guidelines have
become the gold standard (or standard of excellence)
for patient care, and nurses and other healthcare pro-
viders are encouraged to incorporate these standard-
ized guidelines into their practice. Expert national and
international government agencies, professional orga-
nizations, and centers of excellence have made many
of these evidence-based guidelines available online.
When selecting a guideline for practice, be sure that a
credible agency or organization developed the guide-
line and that the reference list reflects the synthesis of
extensive research evidence.
An extremely important source for evidence-based
guidelines in the United States is the National Guide-
line Clearinghouse (NGC), which was initiated in
1998 by the AHRQ. The Clearinghouse started with
200 guidelines and has expanded to contain more than
1500 EBP guidelines (see http://www.guideline.gov/).
Another excellent source of systematic reviews and
EBP guidelines is the Cochrane Collaboration and
Library in the United Kingdom, which can be accessed
at http://www.cochrane.org/. The Joanna Briggs Insti-
tute has also been a leader in developing evidence-
based guidelines for nursing practice (http://www
.joannabriggs.edu.au/). In addition, professional nur­
sing organizations, such as the Oncology Nursing
Society (http://www.ons.org/) and the National Asso-
ciation of Neonatal Nurses (http://www.nann.org/),
have developed EBP guidelines for their specialties.
These websites will introduce you to some of guide-
lines that exist nationally and internationally. Chapter
19 will help you to critically appraise the quality of an
EBP guideline and implement that guideline in your
practice.
KEY POINTS
• Florence Nightingale initiated nursing research
more than 150 years ago; this start was followed
by decades of limited research. During the 1950s
and 1960s, research became a higher priority, with
the development of graduate programs in nursing
that increased the number of nurses with doctorates
and master’s degrees. In the 1970s and 1980s, the
major focus was on the conduct of clinical research
to improve nursing practice.
• Outcomes research emerged as an important meth-
odology for documenting the effectiveness of
healthcare service in the 1980s and 1990s. In 1989,
the Agency for Health Care Policy and Research
(later renamed the Agency for Healthcare Research
and Quality [AHRQ]) was established to facilitate
the conduct of outcomes research.
• The vision for nursing in the 21st century is the
development of a scientific knowledge base that
enables nurses to implement an EBP.
• Nursing research incorporates quantitative, qualita-
tive, outcomes, and intervention research method-
ologies.
• Quantitative research is classified into four types
for this textbook: descriptive, correlational, quasi-
experimental, and experimental.
• Qualitative research is classified into five types
for this textbook: phenomenological research,
32 UNIT ONE  Introduction to Nursing Research
grounded theory research, ethnographic research,
exploratory-descriptive qualitative research, and
historical research.
• Outcomes research focuses on determining the end
results of care or a measure of the change in health
status of the patient and family.
• Intervention research involves the investigation
of the effectiveness of a nursing intervention in
achieving the desired outcomes in a natural setting.
• Best research evidence is a summary of the highest-
quality, current empirical knowledge in a specific
area of health care that is developed from a synthe-
sis of high-quality studies (quantitative, qualitative,
outcomes, and intervention) in that area.
• Research evidence in nursing and health care
is synthesized using the following processes: (1)
systematic review, (2) meta-analysis, (3) meta-
synthesis, and (4) mixed methods systematic
review.
• The levels of the research evidence can be thought
of as a continuum with the highest quality of
research evidence at one end and the weakest at the
other. The best research evidence is synthesized by
a team or panel of experts to develop evidence-
based guidelines for clinicians in practice.
• EBP guidelines are rigorous, explicit clinical guide-
lines that are based on the best research evidence
available in that area.
• EBP guidelines have become the gold standard (or
standard of excellence) for patient care, and nurses
and other healthcare providers are encouraged to
incorporate them into their practice.
REFERENCES
Abdellah, F. G. (1972). Evolution of nursing as a profession. Inter-
national Nursing Review, 19(3), 219–235.
Agency for Healthcare Research and Quality (AHRQ, 2012). AHRQ
at a glance. Retrieved http://www.ahrq.gov/about/ataglance
.htm/.
American Association of Colleges of Nursing (AACN, 1999). Posi-
tion statement on nursing research. Journal of Professional
Nursing, 15(4), 253–257.
American Association of Colleges of Nursing (AACN, 2012). Mis-
sions and values. Retrieved from http://www.aacn.nche.edu/
about-aacn/mission-values/.
American Nurses Association (ANA, 1950). Twenty thousand
nurses tell their story. Kansas City, MO: Author.
American Nurses Association (ANA, 1989). Education for partici-
pation in nursing research. Kansas City, MO: Author.
American Nurses Association (ANA, 2012). American
Nurses Association Research Agenda. Retrieved from
http://www.nursingworld.org/MainMenuCategories/ThePractice
ofProfessionalNursing/PatientSafetyQuality/Research-Measure
ment/Research-Agenda-.asp/.
American Nurses Credentialing Center (ANCC, 2012). Magnet
Program Overview. Retrieved from http://www.nurse
credentialing.org/Magnet/ProgramOverview.aspx.
Andrel, J. A., Keith, S. W., & Leiby, B. E. (2009). Meta-analysis:
A brief introduction. Clinical & Translational Science, 2(5),
374–378.
Barnett-Page, E., & Thomas, J. (2009). Methods for the synthesis
of qualitative research: A critical review. BMC Medical Research
Methodology, 9, 59. DOI: 10.1186/147–2288–9-59.
Bauknecht, V. L. (1986). Congress overrides veto, nursing gets
center for research. American Nurse, 18(1), 24.
Baumrind, D. (1980). New directions in socialization research.
American Psychologist, 35(7), 639–652.
Brown, S. J. (2002). Focus on research methods. Nursing interven-
tion studies: A descriptive analysis of issues important to clini-
cians. Research in Nursing & Health, 25(4), 317–327.
Brown, S. J. (2009). Evidence-based nursing: The research-practice
connection. Sudbury, MA: Jones and Bartlett Publishers.
Clark, A. M. (1998). The qualitative-quantitative debate: Moving
from positivism and confrontation to post-positivism and recon-
ciliation. Journal of Advanced Nursing, 27I(6), 1242–1249.
Cohen, B. (1984). Florence Nightingale. Scientific American,
250(3), 128–137.
Conduct and Utilization of Research in Nursing (CURN) Project.
(1981-1982). Using research to improve nursing practice. New
York, NY: Grune & Stratton.
Cook, Sir E. (1913). The life of Florence Nightingale (Vol. 1).
London, England: Macmillan.
Craig, J. V., & Smyth, R. L. (2012). The evidence-based practice
manual for nurses (3rd ed.). Edinburgh, Scotland: Churchill
Livingstone.
Creswell, J. W. (2009). Research design: Qualitative, quantitative,
and mixed methods approaches (3rd ed.). Los Angeles, CA: Sage.
de Tornyay, R. (1977). Nursing research—the road ahead. Nursing
Research, 26(6), 404–407.
Doran, D. M. (2011). Nursing-sensitive outcomes: State of the
science. Sudbury, MA: Jones & Bartlett.
Downs, F. S., & Fleming, W. J. (1979). Issues in nursing research.
New York, NY: Appleton-Century-Crofts.
Finfgeld-Connett, D. (2010). Generalizability and transferability of
meta-synthesis research findings. Journal of Advanced Nursing,
66(2), 246–254.
Fitzpatrick, M. L. (1978). Historical studies in nursing. New York,
NY: Teachers College Press.
Forbes, A. (2009). Clinical intervention research in nursing. Inter-
national Journal of Nursing Studies, 46(4), 557–568.
Ford-Gilboe, M., Campbell, J., & Berman, H. (1995). Stories and
numbers: Coexistence without compromise. Advances in Nursing
Science, 18(1), 14–26.
Glaser, B. G., & Strauss, A. L. (1967). The discovery of grounded
theory: Strategies for qualitative research. Chicago, IL: Aldine.
Gortner, S. R., & Nahm, H. (1977). An overview of nursing research
in the United States. Nursing Research, 26(1), 10–33.
Herbert, R. G. (1981). Florence Nightingale: Saint, reformer or
rebel? Malabar, FL: Robert E. Krieger.
Higgins, J. P. T., & Green, S. (2008). Cochrane handbook for sys-
tematic reviews of interventions. West Sussex, England: Wiley-
Blackwell and The Cochrane Collaboration.
 CHAPTER 2  Evolution of Research in Building Evidence-Based Nursing Practice 33
Horsley, J. A., Crane, J., Crabtree, M. K., & Wood, D. J. (1983).
Using research to improve nursing practice: A guide; CURN
Project. New York, NY: Grune & Stratton.
Institute of Medicine. (2001). Crossing the quality chasm: A new
health system for the 21st century. Washington, DC: National
Academy Press.
Jacox, A. (1980). Strategies to promote nursing research. Nursing
Research, 29(4), 213–218.
Johnson, J. E. (1993). Outcomes research and health care reform:
Opportunities for nurses. Nursing Connections, 6(4), 1–3.
Johnson, W. L. (1977). Research programs of the National League
for Nursing. Nursing Research, 26(3), 172–176.
Kaplan, A. (1964). The conduct of inquiry: Methodology for behav-
ioral science. New York, NY: Chandler.
Kerlinger, F. N., & Lee, H. B. (2000). Foundations of behavioral
research (4th ed.). Fort Worth, TX: Harcourt.
Krueger, J. C., Nelson, A. H., & Wolanin, M. A. (1978). Nursing
research: Development, collaboration, and utilization. German-
town, MD: Aspen.
Larson, E. (1984). Health policy and NIH: Implications for nursing
research. Nursing Research, 33(6), 352–356.
Lundy, K. S. (2012). Historical research. In P. L. Munhall (Ed.),
Nursing research: A qualitative perspective (5th ed., pp. 381–
397). Sudbury, MA: Jones & Bartlett Learning.
Marshall, C., & Rossman, G. B. (2011). Designing qualitative
research (5th ed.). Los Angeles, CA: Sage.
Melnyk, B. M., & Fineout-Overholt, E. (2011). Evidence-based
practice in nursing & healthcare: A guide to best practice (2nd
ed.). Philadelphia, PA: Lippincott Williams & Wilkins.
Munhall, P. L. (2012). Nursing research: A qualitative perspective
(5th ed.). Sudbury, MA: Jones & Bartlett Learning.
National Institute of Nursing Research (NINR, 2011). NINR mission
and strategic plan. Retrieved from http://www.ninr.nih.gov/
AboutNINR/NINRMissionandStrategicPlan/.
National League for Nursing (NLN; 2011). About the NLN.
Retrieved from http://www.nln.org/aboutnln/index.htm.
Newman, M. A. (1992). Prevailing paradigms in nursing. Nursing
Outlook, 40(1), 10–13, 32.
Nightingale, F. (1859). Notes on nursing: What it is, and what it is
not. Philadelphia, PA: Lippincott.
Norbeck, J. S. (1987). In defense of empiricism. Image—Journal of
Nursing Scholarship, 19(1), 28–30.
Oakley, K. (2010). Nursing by the numbers. Occupational Health,
62(4), 28–29.
Palmer, I. S. (1977). Florence Nightingale: Reformer, reactionary,
researcher. Nursing Research, 26(2), 84–89.
Rettig, R. (1991). History, development, and importance to nursing
of outcomes research. Journal of Nursing Quality Assurance,
5(2), 13–17.
Sackett, D. L., Straus, S. E., Richardson, W. S., Rosenberg, W., &
Haynes, R. B. (2000). Evidence-based medicine: How to practice
& teach EBM (2nd ed.). London, England: Churchill
Livingstone.
Sandelowski, M., & Barroso, J. (2007). Handbook for synthesizing
qualitative research. New York, NY: Springer.
Scheffler, I. (1967). Science and subjectivity. Indianapolis, IN:
Bobbs-Merrill.
Shadish, S. R., Cook, T. D., & Campbell, D. T. (2002). Experimental
and quasi-experimental designs for generalized causal inference.
Boston, MA: Houghton Mifflin Company.
See, E. M. (1977). The ANA and research in nursing. Nursing
Research, 26(3), 165–171.
Sidani, S., & Braden, C. P. (1998). Evaluating nursing interven-
tions: A theory-driven approach. Thousand Oaks, CA: Sage.
Silva, M. C., & Rothbart, D. (1984). An analysis of changing trends
in philosophies of science on nursing theory development and
testing. Advances in Nursing Science, 6(2), 1–13.
Stetler, C. B. (2001). Updating the Stetler Model of research utiliza-
tion to facilitate evidence-based practice. Nursing Outlook, 49(6),
272–279.
Stetler, C. B., & Marram, G. (1976). Evaluating research findings
for applicability in practice. Nursing Outlook, 24(9), 559–563.
U.S. Department of Health and Human Services (DHHS, 1992).
Healthy People 2000. Washington, D.C.: Author.
U.S. Department of Health and Human Services (DHHS, 2000).
Healthy People 2010. Washington, D.C.: Author.
U.S. Department of Health and Human Services (DHHS, 2012).
Health People 2020 topics and objectives. Retrieved from http://
www.healthypeople.gov/2020/topicsobjectives2020/default.aspx.
Werley, H. H. (1977). Nursing research in perspective. International
Nursing Review, 24(3), 75–83.
Werley, H. H., & Fitzpatrick, J. J. (Eds.) (1983). Annual review of
nursing research (Vol. 1). New York, NY: Springer.
Wolf, Z. (2012). Ethnography: The method. In P. L. Munhall (Ed.),
Nursing research: A qualitative perspective (5th ed., pp. 285–
338). Sudbury, MA: Jones & Bartlett Learning.
Wuest, J. (2012). Grounded theory: The method. In P. L. Munhall
(Ed.), Nursing research: A qualitative perspective (5th ed., pp.
225–256). Sudbury, MA: Jones & Bartlett Learning.
Whittemore, R. (2005). Combining evidence in nursing research:
Methods and implications. Nursing Research, 54(1), 56–62.

3  
CHAPTER
Introduction to Quantitative Research
W
hat do you think of when you hear the
word research? Frequently, the word exper-
iment comes to mind. One might equate
experiments with randomizing subjects into groups,
collecting data, and conducting statistical analyses.
Many people believe that an experiment is conducted
to answer a clinical question, such as “Is one pain
medicine more effective than another?” These ideas
are associated with the classic experimental design
originated by Sir Ronald Fisher (1935). Fisher is
noted for adding structure and control to the steps of
the quantitative research process to decrease the
potential for error and improve the accuracy of study
findings.
Four types of quantitative research are included in
this text: descriptive, correlational, quasi-experimental,
and experimental. Fisher’s experimentation provided
the groundwork for what is now known as experi­
mental research. Throughout the years, other quantita-
tive approaches have been developed. Campbell and
Stanley (1963) are noted for developing quasi-
experimental designs for conducting quantitative
research. Karl Pearson (Kerlinger & Lee, 2000) devel-
oped statistical approaches for examining relation-
ships among variables, which increased the conduct
of correlational research. The fields of sociology, edu-
cation, and psychology are noted for their develop-
ment and expansion of strategies for conducting
descriptive research. The steps of the research process
used in these different types of quantitative study are
the same, but the philosophy and strategies for imple-
menting these steps vary with the approach.
Many quantitative research approaches and designs
are essential to develop the body of knowledge needed
for evidence-based practice. Thus, quantitative re-
search is a major focus throughout this textbook. This
chapter provides an overview of quantitative research
by (1) discussing concepts relevant to quantitative re-
search, (2) identifying the steps of the quantitative
34
  http://evolve.elsevier.com/Grove/practice/
research process, and (3) providing examples of dif-
ferent types of quantitative studies.
Concepts Relevant to
Quantitative Research
Some concepts relevant to quantitative research are
basic research, applied research, rigor, and control.
These concepts are defined here, and the major points
are reinforced with examples from quantitative
studies.
Basic Research
Basic, or pure, research is a scientific investigation
that involves the pursuit of “knowledge for knowl-
edge’s sake,” or for the pleasure of learning and
finding truth (Nagel, 1961). The purpose of basic
research is to generate and refine theory and build
constructs; thus, the findings are frequently not directly
useful in practice. However, because the findings are
more theoretical in nature, they can be generalized to
various settings (Wysocki, 1983).
Basic research also examines pathological and
physiological responses as well as underlying mecha-
nisms of actions of an intervention or outcome
(Fawcett & Garity, 2009). Often basic research is con-
ducted in a laboratory with animals or human tissues.
For example, cachexia in cancer patients clinically
manifests as anorexia, weight loss, and wasting of
skeletal muscles that decrease patients’ functioning
and quality of life. Cachexia is a very complex process
that has been studied for many years, but the knowl-
edge of the specific causes and progression of this
condition is still evolving. Thus, additional basic
research is needed to examine the pathology of cancer
cachexia with skeletal muscle wasting. For example,
Byerley et al. (2010) conducted a basic study to
examine the causes of body fat loss early in the
 CHAPTER 3  Introduction to Quantitative Research 35
development of cancer cachexia in male rats. Their
study findings are summarized in the following
excerpt:
“We provide evidence that a factor other than the
previously identified lipid mobilizing factor, zinc
alpha-2 glycoprotein, promotes lipolysis in the …
sarcoma-bearing cachexia model.… We compared
tumor-bearing ad lib fed (TB) animals to nontumor-
bearing ad lib fed (NTB) animals.… Prior to cachexia,
the TB animals lost more than 10 ± 0.7% of their
body fat before losing protein mass and decreasing
their food intake. Fat loss occurred because adipo-
cyte size, not number, was reduced.… Yet circulating
levels of norepinephrine, epinephrine, and TNF-α
[tissue necrosing factor-alpha], and zinc alpha-2 gly-
coprotein were not increased prior to the loss of fat
mass. We provide evidence for a serum factor(s),
other than zinc alpha-2 glycoprotein, that stimulates
release of glycerol from 3T3-L 1 adipocytes and pro-
motes the loss of stored adipose lipid prior to the loss
of lean body mass in this model.” (Byerley et al.,
2010, p. 484)
Byerley and colleagues’ (2010) study demonstrates
the importance of genetic research in understanding
disease pathology. It provides the basis for further
research to isolate and characterize the lipolytic
protein causing loss of body fat in animals with cancer.
This basic research in animals provides the basis for
human research in this area. A major force in genetic
research is the National Human Genome Research
Institute (NHGRI) (2012), which plans and conducts
a broad program of laboratory research to improve our
understanding of the human genetic makeup, genetics
of diseases, and potential gene therapy. This basic
research provides a basis for conducting applied “clin-
ical research to translate genomic and genetic research
into a greater understanding of human genetic disease,
and to develop better methods for the detection, pre-
vention, and treatment of heritable and genetic disor-
ders” (NHGRI, 2012).
Applied Research
Applied, or practical, research is a scientific inves-
tigation conducted to generate knowledge that will
directly influence or improve clinical practice. The
purpose of applied research is to solve problems, to
make decisions, or to predict or control outcomes in
real-life practice situations. Because applied research
focuses on specific problems, the findings are less
generalizable than those from basic research. Applied
research is also used to test theory and validate
its usefulness in clinical practice. Often, the new
knowledge discovered through basic research is exam-
ined for usefulness in practice by applied research,
making these approaches complementary (Bond &
Heitkemper, 1987; NHGRI, 2012).
Artinian et al. (2007) conducted an applied study
to determine the effectiveness of a nurse-managed
telemonitoring (TM) program on the blood pressure
(BP) of urban African Americans. The TM program
(1) provided BP equipment for patients to monitor
their BPs at home, (2) improved access to care by
sending patients’ BP readings via telephone to health-
care agencies, and (3) increased monitoring of the
patients’ BPs by a care provider with immediate
feedback to the patients. The treatment group received
the nurse-managed TM intervention or treatment, and
the comparison group received usual care (UC). The
TM intervention group had a significant reduction
in systolic BP in comparison with the UC group,
and diastolic BP was greatly reduced but was not
statistically significant from that of the UC group at
12 months. Thus, the TM intervention did have a posi-
tive impact on the BPs of African Americans, and
additional research is needed to determine whether the
intervention has a long-term effect on BPs and
improves hypertension control in this population. The
findings from this applied study do have implications
for practice, because this nurse-managed TM interven-
tion significantly affected BP in a population with a
high incidence of hypertension. On the basis of Artin-
ian and colleagues’ (2007; 2004; Artinian, Washing-
ton, & Templin, 2001) research and the research of
others documenting the importance of home BP moni-
toring, a scientific statement from the American Heart
Association, American Society of Hypertension, and
Preventive Cardiovascular Nurses Association recom-
mended the use of and reimbursement for home BP
monitoring (Pickering et al., 2008). Artinian was a
member of the group making this recommendation
about home BP monitoring. For more information
about the home BP monitoring recommendation,
you can view an article on the American Heart Asso-
ciation (2011) website, at http://my.americanheart
.org/professional/General/Call-to-Action-on-Use-and-
Reimbursement-for-Home-Blood-Pressure-Monito
ring_UCM_423866_Article.jsp.
Many nurse researchers have conducted applied
studies to produce findings that directly affect clinical
practice. Usually, applied studies focus on developing
and testing the effectiveness of nursing interventions
in the treatment of patient and family health problems.
36 UNIT ONE  Introduction to Nursing Research
In addition, most previous federal funding for nursing
has been granted for applied research. However, the
National Institute of Nursing Research (NINR, 2012)
recognizes the importance of basic research to nursing
and has made it a funding priority. Basic research is
needed to expand our understanding of several patho-
physiological variables, such as impaired oxygenation
and perfusion, fluid and electrolyte imbalance, altered
neurological function, impaired immune system,
nutritional disorders, and sleep disturbance. NHGRI
(2012) priorities include the identification of the genes
responsible for numerous human genetic diseases and
the generation of animal models essential for the study
of human inherited disorders. Because the future of
any profession rests on its research base, both basic
and applied studies are needed to develop knowledge
for evidence-based practice (EBP) in nursing (Brown,
2009; Melnyk & Fineout-Overholt, 2011).
Rigor in Quantitative Research
Rigor is the striving for excellence in research and
involves discipline, scrupulous adherence to detail,
and strict accuracy. A rigorous quantitative researcher
constantly strives for more precise measurement
methods, structured treatments, representative sam­
ples, and tightly controlled study designs (Borglin &
Richards, 2010; Shadish, Cook, & Campbell, 2002).
Characteristics valued in these researchers include
critical examination of reasoning and attention to
precision.
Logistic reasoning and deductive reasoning are
essential to the development of quantitative research.
The research process consists of specific steps that are
developed with meticulous detail and logically linked
together. These steps are critically examined and reex-
amined for errors and weaknesses in areas such as
design, intervention development and implementa-
tion, measurement, sampling, statistical analysis, and
generalization. Reducing these errors and weaknesses
is essential to ensure that the research findings are an
accurate reflection of reality and can be generalized.
Generalizing research findings involves applying the
findings from a particular study sample to a large
population of similar individuals (Borglin & Richards,
2010). For example, the findings from the study by
Artinian et al. (2007) could be generalized from the
African American study participants with hyperten-
sion to the larger population of African Americans
with hypertension.
Another aspect of rigor is precision, which encom-
passes accuracy, detail, and order. Precision is evident
in the concise statement of the research purpose, the
detailed development of the study design, and the
formulation of explicit treatment protocols. The most
explicit use of precision, however, is evident in the
measurement of the study variables (Borglin &
Richards, 2010). Measurement involves objectively
experiencing the real world through the senses: sight,
hearing, touch, taste, and smell. The researcher con-
tinually searches for new and more precise ways to
measure elements and events of the world (Waltz,
Strickland, & Lenz, 2010).
Control in Quantitative Research
Control occurs when the researcher imposes “rules”
to decrease the possibility of error and thus increase
the probability that the study’s findings are an accurate
reflection of reality. The rules used to achieve control
are referred to as design. Through control, the
researcher can reduce the influence or confounding
effect of extraneous variables on the study variables.
For example, in a study focused on the effect of relax-
ation therapy on the perception of incisional pain, the
extraneous variables, such as type of surgical incision
and the timing, amount, and type of pain medication
administered after surgery, would have to be con-
trolled to prevent them from influencing the patient’s
perception of pain.
Controlling extraneous variables enables the
researcher to identify relationships among the study
variables accurately and examine the effects of one
variable on another. Researchers can control extrane-
ous variables with sampling criteria by selecting a
certain type of subject, such as only those individuals
who are having abdominal surgery or those with a
certain medical diagnosis. Using a random sampling
method also improves the control of extraneous vari-
ables and reduces the potential for bias in the study
sample. The setting can also be structured to control
extraneous variables such as temperature, noise, and
interactions with other people. The data collection
process can be sequenced to control extraneous vari-
ables such as fatigue and discomfort (Borglin &
Richards, 2010).
Quantitative research requires varying degrees of
control, ranging from minimal control to highly con-
trolled, depending on the type of study (Table 3-1).
Descriptive studies are usually conducted with
minimal control of the study design, because subjects
are examined as they exist in their natural setting, such
as home, work, or school. However, the researcher
still hopes to achieve the most precise measurement
of the research variables as possible. Experimental
studies are highly controlled and often conducted on
animals in laboratory settings to determine the under-
lying mechanisms for and effectiveness of a treatment.
 CHAPTER 3  Introduction to Quantitative Research 37
TABLE 3-1  Control in Quantitative Research
Level of Control in
Development of the
Type of Research Research Design
Descriptive research Minimal or partial control
Correlational research Minimal or partial control
Quasi-experimental research Moderate to high control
Experimental research High control
Some common areas in which control might be
enhanced in quantitative research are (1) selection of
subjects (sampling), (2) reduction of subject or partici-
pant attrition, (3) selection of the research setting, (4)
development and implementation of the intervention,
(5) measurement of study variables, and (6) subjects’
knowledge of the study (Borglin & Richards, 2010;
Forbes, 2009; Shadish et al., 2002). Nurses are being
encouraged to develop more powerful, controlled, rig-
orous quantitative studies (NINR, 2012).
Sampling and Attrition
Sampling is a process of selecting subjects, events,
behaviors, or elements for participation in a study. In
performing quantitative research, you will use a
variety of random and nonrandom sampling methods
to obtain study samples. Random sampling methods
usually provide a sample that is representative of a
population, because each member of the population
has a probability greater than zero of being selected
for a study. Thus, random or probability sampling
methods require greater researcher control and rigor
than nonrandom or nonprobability sampling methods.
Sample sizes in quantitative studies are usually deter-
mined with a power analysis to ensure adequate
numbers of study participants throughout the study.
Researchers are rigorous in reducing attrition, or loss
of study subjects needed to describe variables, examine
relationships, and determine the effect of interventions
(Aberson, 2010; Thompson, 2002). Chapter 15 pro-
vides a detailed discussion of the sampling process
and determining sample size for quantitative studies.
Research Settings
There are three common settings for conducting
research: natural, partially controlled, and highly con-
trolled. Natural settings are uncontrolled, real-life
settings where studies are conducted (Fawcett &
Garity, 2009; Kerlinger & Lee, 2000). Descriptive
and correlational types of quantitative research are
often conducted in natural settings. A partially con-
trolled setting is an environment that the researcher
manipulates or modifies in some way. A growing
number of quasi-experimental studies are being con-
ducted to test the effectiveness of nursing interven-
tions, and these studies are often conducted in partially
controlled settings. For example, hospitals, clinics, or
rehabilitation centers might be manipulated in selected
ways to control for extraneous variables, such as type
of care, medications, and family interactions. Highly
controlled settings are artificially constructed envi-
ronments that are developed for the sole purpose of
conducting research. Laboratories, experimental
centers, and research units are highly controlled set-
tings often used for the conduct of experimental
research. Chapter 15 discusses the process for select-
ing a setting for the conduct of quantitative research.
Development and Implementation of Study
Interventions or Treatments
Quasi-experimental and experimental studies examine
the effect of an independent variable or intervention
on a dependent variable or outcome. More interven-
tion studies are being conducted in nursing to establish
an EBP. Controlling the development and implemen-
tation of a study intervention increases the validity of
the study design and the credibility of the findings. A
study intervention needs to be (1) clearly and precisely
developed, (2) consistently implemented, and (3)
examined for effectiveness through quality measure-
ment of the dependent variables (Forbes, 2009;
Morrison et al., 2009; Santacroce, Maccarelli, & Grey,
2004). The detailed development of a quality interven-
tion and the consistent implementation of this inter-
vention are known as intervention fidelity. Chapter
14 provides detailed directions for the development
and implementation of a study intervention. Artinian
et al. (2007) provided the following detailed descrip-
tion of the implementation of the nurse-managed TM
(telemonitoring) intervention to improve the BPs of
African Americans:
“Participants in the TM group received UC [usual
care] plus nurse-managed TM. Specially trained reg-
istered nurses delivered the intervention. During a
prescheduled appointment, the intervention nurse
delivered the BP monitor and TM link device (device
that links BP monitor to the telephone) to the par-
ticipant’s home. At the time of the home visit, an
intervention nurse taught participants how to self-
monitor BP in accordance with The Seventh Report
of the Joint National Committee on Prevention,
Detection, Evaluation, and Treatment of High Blood
38 UNIT ONE  Introduction to Nursing Research
Pressure (JNC-VII) guidelines (Chobanian et al.,
2003), set up the home TM system, demonstrated
the system, had participants practice using the BP
monitor, and answered questions. Given the memory
in the BP monitor and that all BPs recorded by the
monitor were telephonically sent to care providers
and the principal investigator, participants received
verbal and written reminders that the BP monitor was
exclusively for their use.… LifeLink Monitoring, Inc.
(Bearsville, NY) provided TM services for this
study.… Telemonitoring participants were also asked
to telephonically send their BP readings to the inter-
vention nurse and their care providers.… Once the
intervention nurses received the BP reports, they
telephoned each participant to provide feedback in
relation to the target goals and to provide telecounsel-
ing about lifestyle modification and medication adher-
ence in accordance with JNC-VII guidelines (Chobanian
et al., 2003).” (Artinian et al., 2007, p. 315)
Measurement of Study Variables
When you are conducting a quantitative study, you
will attempt to use the most precise instruments avail-
able to measure the study variables. Using a variety
of quality measurement methods promotes an accurate
and comprehensive understanding of the study vari-
ables. In addition, researchers want to rigorously
control the process for measuring study variables to
improve the design validity and quality of the study
findings (Waltz et al., 2010). Measurement concepts,
process, and strategies are the foci of Chapters 16
and 17.
Nursing studies often include the measurement of
biophysical variables, which require precise, accurate
physiological measures (Ryan-Wenger, 2010). For
example, Artinian et al. (2007) described their precise
measurement of the dependent physiological variable,
BP, with an accurate, nationally standardized device
as follows:
“The outcome measure of the BP was measured with
electronic BP monitor (Omron HEM-737 Intellisense,
Omron Healthcare, Inc., Vernon Hills, IL) that has
been validated in accordance with the criteria of the
British Hypertension Society and the Association of
the Advancement of Medical Instrumentations (Dabl
Educational Trust, 2005).” (Artinian et al., 2007,
p. 316)
Subjects’ Knowledge of a Study
Subjects’ knowledge of a study could influence their
behavior and possibly alter the research outcomes.
This possibility threatens the validity or accuracy of
the study design. An example of this type of threat to
design validity is the Hawthorne effect, which was
identified during the classic experiments at the Haw-
thorne plant of the Western Electric Company during
the late 1920s and early 1930s. The employees at this
plant exhibited a particular psychological response
when they became research participants: They changed
their behavior simply because they were subjects in a
study, not because of the research treatment. In these
studies, the researcher manipulated the working con-
ditions (altered the lighting, decreased work hours,
changed payment, and increased rest periods) to
examine the effects on worker productivity (Homans,
1965). The subjects in both the treatment group (whose
work conditions were changed) and the control group
(whose work conditions were not changed) increased
their productivity. The subjects seemed to change
their behaviors (increase their productivity) solely in
response to being part of a study. In the study by
Artinian et al. (2007, p. 321), both the treatment and
the comparison groups experienced decreases in their
BPs, and the researchers indicated “the Hawthorne
effect may have been a factor, with participants paying
more attention to their BP and hypertension self-care
behaviors because they were aware of their participa-
tion in the study.”
There are several ways to strengthen a study,
decreasing the threats to design validity and selecting
the strongest design for the proposed study. Chapter
10 addresses design validity, and Chapter 11 focuses
on the process for selecting an appropriate study
design. Your understanding of rigor and control
provide the basis for the implementation of the steps
of the quantitative research process, which are pre-
cisely executed in descriptive, correlational, quasi-
experimental, and experimental research.
Steps of the Quantitative
Research Process
The quantitative research process consists of concep-
tualizing a research project, planning and implement-
ing that project, and communicating the findings.
Figure 3-1 identifies the steps of the quantitative
research process and shows the logical flow of this
process as each step progressively builds on the
 CHAPTER 3  Introduction to Quantitative Research 39

FORMULATING A RESEARCH PROBLEM also contains a feedback arrow, indicating that the
AND PURPOSE research process is cyclical, for each study provides a
basis for generating further research in the develop-
ment of knowledge for EBP.
REVIEW OF RELEVANT LITERATURE
In this chapter, you are briefly introduced to the
steps of the quantitative research process that are pre-
DEVELOPING A FRAMEWORK sented in detail in Unit Two, The Research Process,
and Unit Four, Collecting and Analyzing Data to
Determine Research Outcomes for Dissemination.
Making Assumptions Explicit
The descriptive correlational study conducted by Gill
and Loh (2010), on the relationships of perceived
FORMULATING RESEARCH OBJECTIVES, stress, optimism, and health-promoting behaviors in
QUESTIONS, OR HYPOTHESES new primiparous mothers, is presented as an example
for introducing the steps of the quantitative research
process. Quotations from this study appear throughout
DEFINING STUDY VARIABLES
this section, in shaded boxes, to clarify the steps of the
quantitative research process. The bracketed words are
SELECTING A RESEARCH DESIGN inserted in the quotations to clarify the key concepts
and steps of the quantitative research process.
DEFINING THE POPULATION AND SAMPLE Formulating a Research Problem
and Purpose
SELECTING METHODS OF MEASUREMENT A research problem is an area of concern or phenom-
enon of interest in which there is a gap in the knowl-
edge base needed for nursing practice. The problem
DEVELOPING A PLAN FOR DATA COLLECTION
AND ANALYSIS identifies an area of concern or phenomenon of inter-
est for a particular population and often indicates the
concepts to be studied. The major sources for nursing
IMPLEMENTING THE RESEARCH PLAN research problems include (1) nursing practice, (2)
literature review, (3) research priorities for funding
Pilot Study agencies and professional organizations, (4) researcher
and peer interactions, and (5) theory testing. The
GENERATING research problem usually indicates significance to
Data Collection FURTHER nursing, background knowledge in the area, and
RESEARCH
problem statement of what is not known (Fawcett &
Data Analysis Garity, 2009). As a researcher, you will use deductive
reasoning to generate a research problem from a
research topic or a broad problem area of personal
Interpreting Research Outcomes interest that is relevant to nursing.
The research purpose is generated from the
COMMUNICATING RESEARCH FINDINGS problem and identifies the specific focus or aim
of the study. The focus of the study might be to
Figure 3-1  Steps of the quantitative research process. identify, describe, explain, or predict a solution to a
situation. The purpose often indicates the type of
study to be conducted (descriptive, correlational,
quasi-experimental, or experimental) and usually
previous steps. This research process is also flexible includes the variables, population, and setting for
and fluid, with a flow back and forth among the steps the study. Chapter 5 provides a background for for-
as researchers strive to clarify the steps and strengthen mulating a research problem and purpose. Gill and
the proposed study. This back-and-forth flow among Loh (2010) identified the following problem and
the steps is indicated in the figure by the two-way purpose for their study of new primiparous mothers
arrows connecting the steps of the process. Figure 3-1 as follows:
40 UNIT ONE  Introduction to Nursing Research
Problem
“The transition to motherhood is considered to be
one of the most stressful and disruptive life transitions
(Chen, Kuo, Chou, & Chen, 2007; Warren, 2005).…
Additional competing priorities may include learning
the maternal role, experiencing physical changes, and
learning how to care for their own needs.… These
different roles and priorities may be perceived by the
new mother as highly stressful, particularly as many
of these stressors occur simultaneously (Warren,
2005).… Specifically, perceived stress can occur
when the demands of a particular situation are
assessed by the individual as greater than their coping
attributes [problem significance].…
“Past research has found that perceived stress
may lead a new mother to neglect personal health-
promoting behaviors in her effort to balance the com-
peting priorities of being a first-time mother (Chen
et al., 2007; Walker, 1989).… Optimism and perceived
stress have an inverse relationship (Brissette, Scheier,
& Carver, 2002).… Optimistic people cope better
with stressful situations and thereby report fewer
physical health symptoms [problem background].…
“Perceived stress has been associated with fewer
health-promoting behaviors in new primiparous
mothers, but less is known about the mechanisms
responsible for such effects” [problem statement].
(Gill & Loh, 2010, pp. 348–349)
Purpose
“This investigation was conducted to examine the
potential role of optimism as a partial mediator of the
relationship between perceived stress and health-
promoting behaviors in new primiparous mothers.”
(Gill & Loh, 2010, p. 350)
The research problem is significant and is based on
pervious research. The problem statement indicates
what is not known and provides a basis for the study
purpose. The research purpose clearly indicates that
the focus of this study is to describe and examine
relationships among the variables optimism, perceived
stress, and health-promoting behavior in new first-
time mothers.
Review of Relevant Literature
A review of relevant literature is conducted to gener-
ate an understanding of what is known about a particu-
lar situation, phenomenon, or problem and to identify
the knowledge gaps that exist. Relevant literature
refers to those sources that are pertinent or highly
important in providing the in-depth knowledge needed
to study a selected problem and purpose. This back-
ground enables researchers to build on the works of
others. The concepts and interrelationships of the con-
cepts in the problem usually guide the selection of
relevant theories and studies presented in the review
of literature. Theories are presented to clarify the defi-
nitions of concepts and to develop and refine the study
framework.
By reviewing relevant studies, researchers are able
to clarify (1) which problems have been investigated,
(2) which require further investigation or replication,
and (3) which have not been investigated. In addition,
the literature review can direct researchers in design-
ing the study and interpreting the outcomes. Chapter
6 provides details for conducting a review of relevant
literature. Gill and Loh’s (2010) literature review
covered the concepts of perceived stress, optimism,
and health practice. They also described theoretical
models relevant to these concepts. The following
excerpt from the study identifies the headings and key
ideas covered by the review of relevant literature:
“Perceived Stress and Health Practices
In a longitudinal study from midpregnancy to 12
months postpartum, American primiparous mothers
were found to decline gradually in healthy behaviors
after birth, with exercise and stress management
behaviors performed least by these women (Patteson
& Killien, 2001).… There are also significant health
benefits in practicing health-promoting behaviors. For
example, exercise has been shown to have a dual
effect of promoting weight loss and reducing reported
stress in new mothers (Groth & David, 2008).…
Theoretical Models
The transactional model for understanding stress
emphasizes that the environmental event (e.g., new
motherhood) and the cognitive faculties of the indi-
vidual are constantly in transaction through the
process of appraisal (Lazarus & Folkman, 1984).…
According to the transactional model of stress and
coping, individuals adopt different coping mechanisms
to deal with the perceived stress they encounter.…
There are widespread observations that people
with optimism generally tend to use less avoidant and
more goal-directed coping mechanisms.… In this
study, optimism (i.e., dispositional optimism) refers to
a personality variable where individuals believe that
good outcomes rather than bad will occur in one’s life
 CHAPTER 3  Introduction to Quantitative Research 41
(Scheier & Carver, 1992). Optimists tend to have an
internal locus of control and believe that events result
primarily from their own behaviors and actions.…
Perceived Stress and Optimism
Another study assessing constructive thinking during
pregnancy found that optimism helped in the adjust-
ment to stressful situations during this time and
reduced the level of stress experienced (Park, Moore,
Turner, & Adler, 1997). An optimistic disposition can
therefore be seen to reduce perceived stress and
encourage more positive appraisals of situations.…
Optimism and Health Practices
Optimism was found also to diminish the impact of
depression in pregnant women and up to 3 weeks
postpartum (Carver & Gaines, 1987).…
Optimism as a Mediator
Substantial research exists where optimism has
served as a mediator of a stress relationship.… In a
study with pregnant women, optimism was found to
confer positive benefits through constructive thinking
in decreasing perceived stress, anxiety, and substance
abuse.… Despite these findings, previous studies
have not investigated how optimism might mediate
the relationships between perceived stress and
health-promoting behaviors in new mothers. This is
important because health-promoting behaviors have
the potential to help new mothers cope with the daily
stress of parenthood and define maternal identity
(Walker, 1989).” (Gill & Loh, 2010, pp. 348–350)
Gill and Loh (2010) clearly cover relevant studies
related to the concepts of stress, health practices, and
optimism in pregnant women and new mothers. They
also cover relevant theories that describe the relation-
ships among key concepts. The literature review sum-
marizes what is known in the areas of stress, health
practices, and optimism in new mothers and clearly
indicates what is not known. The study purpose
focuses on what is not known, which is how optimism
mediates the relationship between perceived stress and
health-promoting behaviors.
Developing a Framework
A framework is the abstract, logical structure of
meaning that will guide the development of a study
and enable the researcher to link the findings to the
body of nursing knowledge. In quantitative research,
the framework is often a testable midrange theory that
has been developed in nursing or in another discipline,
such as psychology, physiology, or sociology (Smith
& Liehr, 2008). The framework may also be devel-
oped inductively from clinical observations.
The terms related to frameworks are concept, rela-
tional statement, theory, and framework model or
map. A concept is a term to which abstract meaning
is attached. A relational statement or proposition
declares that a relationship of some kind exists between
or among two or more concepts. A theory consists of
an integrated set of defined concepts and propositions
that present a view of a phenomenon and can be used
to describe, explain, predict, or control the phenome-
non. The propositions or relationship statements of the
theory, not the theory itself, are tested through research.
A study framework can be expressed as a model or
a diagram of the relationships that provide the basis
for a study and/or can be presented in narrative format.
The steps for developing a framework are described
in Chapter 7. The framework for Gill and Loh’s (2010)
study is presented in Figure 3-2 and is described in the
following quote. The framework model identifies
the relationships that are examined in this study, and
the description of the framework identifies the propo-
sition that was tested by this study.
“It was speculated that maternal perceived stress
would inversely affect a new primiparous mother’s
health-promoting behaviors. However, this relation-
ship would be mediated partially by a new mother’s
sense of optimism. In other words, on the basis of
work by Scheier and Carver (1992) the protective
role of optimism would enable a new mother to
self-regulate her perception of stress positively and
help her to engage in health-promoting behaviors to
maintain her sense of well-being and control [see
Figure 3-2].
It is important to note that partial mediation (as
opposed to full mediation) was proposed because
optimism was hypothesized to be only one of a
number of possible variables that contribute to the
associate between maternal perceived stress and
health-promoting behaviors” [proposition]. (Gill &
Loh, 2010, p. 350)
Making Assumptions Explicit
Assumptions are statements that are taken for granted
or are considered true, even though they have not been
scientifically tested. Assumptions are often embedded
(unrecognized) in thinking and behavior, and uncover-
ing them requires introspection. Sources of assump-
tions include universally accepted truths (e.g., all
42 UNIT ONE  Introduction to Nursing Research
Optimism
*LOT-R
Perceived
Stress
*PSS-10
Figure 3-2  Theoretical model outlining the relationships between predictor variables and proposed outcomes.
*Instruments used to measure study variables: HPLPII, Health-Promoting Lifestyle Profile II; LOT-R, Life Orientation Test–Revised; PSS-10, Perceived
Stress Scale. (From Gill, R. M., & Loh, J. M. (2010). The role of optimism in health-promoting behaviors in new primiparous mothers. Nursing Research,
59(5), 350.)
humans are rational beings), theories, previous
research, and nursing practice (Myers, 1982).
In studies, assumptions are embedded in the philo-
sophical base of the framework, study design, and
interpretation of findings. Theories and instruments
are developed on the basis of assumptions that the
researcher may or may not recognize. These assump-
tions influence the development and implementation
of the research process. Because researchers’ assump-
tions influence the logic of the study, their recognition
leads to more rigorous study development.
Researchers often do not identify assumptions that
provide a basis for their study in their research report;
but if assumptions are included, they are usually part
of the framework discussion. Gill and Loh (2010) did
not clearly identify the assumptions for their study, but
the following assumptions seem to provide a basis for
it: (1) stress needs to be managed to promote health,
(2) people’s psychological perspectives influence their
actions, and (3) health is a priority for most people.
Formulating Research Objectives, Questions,
or Hypotheses
Research objectives, questions, and hypotheses bridge
the gap between the more abstractly stated research
problem and purpose and the study design and plan
for data collection and analysis. Objectives, questions,
and hypotheses are narrower in focus than the research
purpose and often (1) specify only one or two research
variables, (2) identify the relationship between the
variables, and (3) indicate the population to be studied.
Some quantitative studies do not include objec-
tives, questions, or hypotheses; the development of
such a study is directed by the research purpose. Many
descriptive studies include only a research purpose,
and other descriptive studies include a purpose and
Outcomes
Reduced
psychological
and
physiological
health risks
Increased
Health-
positive well-
Promoting
being
Behaviors
*HPLPII
Increased
maternal identity
objectives or questions. Some correlational studies
include a purpose and specific questions or hypothe-
ses. Quasi-experimental and experimental studies
often use hypotheses to direct the development and
implementation of the studies and the interpretation of
findings. Chapter 8 examines the development of
research objectives, questions, and hypotheses. Gill
and Loh (2010) predicted relationships among the
study variables that were examined with a predictive
correlational design (Kerlinger & Lee, 2000; see
Figure 11-9). With this type of design, researchers
often formulate hypotheses to predict the outcomes for
their study as was done by Gill and Loh:
“Four hypotheses were proposed for the current
study.
1. Perceived stress will be related negatively to
health-promoting behaviors in new primiparous
mothers.
2. Perceived stress will be related negatively to
optimism in new primiparous mothers.
3. Optimism will be related positively to health-
promoting behaviors in new primiparous
mothers.
4. Optimism will mediate the relationship between
perceived stress and health-promoting behaviors
partially in new primiparous mothers.” (Gill &
Loh, 2010, p. 350)
Defining Study Variables
The research purpose and the objectives, questions, or
hypotheses identify the variables that are examined in
a study. Study variables are concepts of various
levels of abstraction that are measured, manipulated,
 CHAPTER 3  Introduction to Quantitative Research 43
or controlled in a study. The more concrete concepts,
such as temperature, weight, and blood pressure, are
referred to as “variables.” The more abstract concepts,
such as creativity, empathy, and social support, are
sometimes referred to as “research concepts.”
The variables or concepts in a study are operation-
alized when they are conceptually and operationally
defined. A conceptual definition provides a variable
or concept with theoretical meaning (Fawcett &
Garity, 2009) and either is derived from a theorist’s
definition of the concept or is developed through
concept analysis. An operational definition indicates
how a variable will be measured or manipulated in a
study. The knowledge you gain from studying the
variable will increase your understanding of the
concept that the variable represents (see Chapter 8).
Gill and Loh (2010) provided conceptual and oper-
ational definitions of the study variables identified in
their purpose and hypotheses: perceived stress, health-
promoting behaviors, and optimism. The conceptual
definitions for these variables are provided in the
review of literature and are reflective of the framework
that includes Lazarus and Folkman’s (1984) transac-
tional model and the role of optimism by Scheier and
Carver (1992). The operational definitions indicate
how the variables were measured and are linked to the
study variables in the framework model (see Figure
3-2) and detailed in the measures section of the study
methodology.
Perceived Stress
Conceptual Definition
“Perceived stress may be defined as the degree to
which individuals appraise events in their life as over-
whelming and insurmountable. Specifically, perceived
stress can occur when the demands of a particular
situation are assessed by the individual as greater than
their coping attributes.” (Gill & Loh, 2010, p. 348)
Operational Definition
Perceived stress was measured by the Perceived
Stress Scale (PSS-10) (Cohen & Williamson, 1988).
Health-Promoting Behavior
Conceptual Definition
“Health-promoting behaviors are defined as a pattern
of actions and cognitions that aim to augment the
level of wellness, self-actualization, and fulfillment of
an individual.” (Gill & Loh, 2010, p. 348)
Operational Definition
The Health-Promoting Life-Style Profile II (HPLPII)
(Walker & Hill-Polerecky, 1996) was used to measure
the new primiparous mothers’ health-promoting
behaviors.
Optimism
Conceptual Definition
“In this study, optimism (i.e., dispositional optimism)
refers to a personality variable where individuals
believe that good outcomes rather than bad will
occur in one’s life (Scheier & Carver, 1992).” (Gill &
Loh, 2010, p. 349)
Operational Definition
“The revised Life Orientation Test (Scheier et al.,
1994) was used to measure optimism exhibited by
the mothers in this study.” (Gill & Loh, 2010, p. 351)
Selecting a Research Design
A research design is a blueprint for maximizing
control over factors that could interfere with a study’s
desired outcome. The type of design directs the selec-
tion of a population, sampling process, methods of
measurement, and a plan for data collection and analy-
sis. The choice of research design depends on the
researcher’s expertise, the problem being examined,
the purpose for the study, and the desire to generalize
the findings.
Designs have been developed to meet unique
research needs as they emerge; thus, a variety of
descriptive, correlational, quasi-experimental, and
experimental designs have been generated over time.
In descriptive and correlational studies, no treatment
is administered, so the study design centers on describ-
ing variables, examining relationships, and improving
the precision of measurement (Waltz et al., 2010).
Quasi-experimental and experimental study designs
usually involve treatment and control groups and
focus on achieving high levels of control as well as
precision in measurement (Cook & Campbell, 1979;
Kerlinger & Lee, 2000; Shadish et al, 2002). Chapter
10 covers the purpose of a design and the threats to
design validity. Chapter 11 presents models and
descriptions of several types of descriptive, correla-
tional, quasi-experimental, and experimental designs.
Gill and Loh (2010) conducted a descriptive cor-
relational study that involved a cross-sectional, pre-
dictive correlational design. Cross-sectional design is
used to examine groups of subjects in various stages
44 UNIT ONE  Introduction to Nursing Research

of development, trends, patterns, and changes simul-
taneously with the intent to describe changes in the
phenomenon across stages. The researchers provide
very little detail of the cross-sectional aspect of this
study design but do mention examining primiparous
mothers with babies varying in age from 2 to 12
months. The predictive correlational design is clearly
addressed in the study. Relationships among study
variables perceived stress, optimism, and health-
promoting behaviors are examined. The predictor or
independent variables, perceived stress and optimism,
are used to predict the criterion or dependent variable,
health-promotion behavior.
Defining the Population and Sample
The population is all the elements (individuals,
objects, or substances) that meet certain criteria
for inclusion in a given universe (Kaplan, 1964;
Thompson, 2002). For example, suppose you wanted
to conduct a study to describe patients’ responses to
nurse practitioners as their primary care providers.
You could define the population in different ways: It
could include all patients being seen for the first time
in (1) a single clinic, (2) all clinics in a specific network
in one city, or (3) all clinics in that network nation-
wide. Your definition of the population would depend
on the sample criteria and the similarity of subjects in
these various settings. The researcher needs to deter-
mine which population is accessible and can be best
represented by the study sample.
A sample is a subset of the population that is
selected for a particular study, and sampling defines
the process for selecting a group of people, events,
behaviors, or other elements with which to conduct a
study. As mentioned earlier in this chapter, nursing
studies use both probability (random) and nonproba-
bility (nonrandom) sampling methods, and sample
size is usually determined by conduction of a power
analysis. The following quote identifies the sample
size, population, sample criteria, and sample charac-
teristics for the study conducted by Gill and Loh
(2010). The study would have been strengthened by
identification of the sampling method, which appears
to be a nonprobability sample of convenience, whereby
subjects are selected because they happen to be in a
given place at a given time.
“Participants
Participants were 174 new primiparous mothers
[sample size; population] recruited from online baby
forums [setting]. Tabachnick and Fidell’s (2007) rule
of thumb was applied to calculate the power required
for this study. With two predictors, 106 participants
were required, and the study was therefore ade-
quately powered [power analysis]. Only primiparous
mothers with babies up to and including 12 months
of age were eligible to complete the study [sample
criteria]. Demographic information for the partici-
pants is displayed in a table” [see Table 3-2] [sample
characteristics]. (Gill & Loh, 2010, p. 350)
Selecting Methods of Measurement
Measurement is the process of assigning “numbers to
objects (or events or situations) in accord with some
rule” (Kaplan, 1964, p. 177). A component of mea-
surement is instrumentation, which is defined as the
application of specific rules to the development of a
measurement device or instrument. An instrument is

TABLE 3-2  Means, Standard Deviations, Potential Range, Obtained Range, and Cronbach’s Alpha of
the Scores for Main Study Variables
Measure n M SD Potential Range Obtained Range Cronbach’s Alpha
1.  Perceived stress 174 13.62 6.83 0-40 0-29 .75
2.  Optimism 174 17.74 4.51 0-24 0-24 .89
3.  Health-promoting behavior 174 2.74 0.58 1-4 1.54-3.92 .97
4.  Income* 174 4.44 0.89 1-5 1-5 —
5.  Age 174 31.5 4.44 18-40+ 18-42 —
6.  Months since birth 174 6.17 3.28 0-12 –.20-12 —

*1 = <$20,000; 2 = $20,001-35,000; 3 = $35,001-50,000; 4 = $50,001-75,000; 5 = >$75,000. Income reported in Australian dollars.

n = Sample size for calculations.
M = Mean.
SD = standard deviation.
Cronbach’s alpha (α) = reliability of the measurement methods for this study.
From Gill, R. M., & Loh, J. M. (2010). The role of optimism in health-promoting behaviors in new primiparous mothers. Nursing Research, 59(5), 348–355.
The table is on page 352.
 CHAPTER 3  Introduction to Quantitative Research 45
selected to measure a specific variable in a study. Data
generated with an instrument are at the nominal,
ordinal, interval, or ratio level of measurement. The
level of measurement, with nominal being the lowest
form of measurement and ratio being the highest,
determines the type of statistical analyses that you can
perform on the data (Grove, 2007).
Selection of an instrument requires extensive
examination of its reliability and validity. Reliability
assesses how consistently the measurement technique
measures a concept. The validity of an instrument is
the extent to which it actually reflects the abstract
concept being examined (Waltz et al., 2010). Chapter
16 introduces the concepts of measurement and
explains the different types of reliability and validity
for instruments and precision and accuracy for physi-
ological measures (Ryan-Wenger, 2010). Chapter 17
provides a background for selecting measurement
methods for a study. Gill and Loh (2010) provided the
following description of the three Likert scales that
were used to measure their study variables:
“Measures
Perceived Stress The Perceived Stress Scale (PSS-10;
Cohen & Williamson, 1988) was used to measure
stress in the mothers. This scale contains 10 items
that measure perceived stress experienced within the
last month. Participants were asked to indicate on a
5-point Likert-type scale (0 = never to 4 = very often)
their response to each item. A sample questions was,
‘In the last month, how often have you found that you
could not cope with all the things that you had to do?’
High scores on the scale indicated higher levels of
perceived stress. Internal reliability of the PSS-10 for
this study was found to be .78, and adequate con-
struct validity was reported (Cohen & Williams,
1988). Cronbach’s alpha (α) for perceived stress in
the current study was .75.
Health-Promoting Behaviors The Health-
Promoting Lifestyle Profile II [HPLPII] (Walker &
Hill-Polerecky, 1996) was used to measure new
primiparous mothers’ health-promoting behaviors.
Participants were asked to indicate on a 4-point scale
(1 = never to 4 = routinely) the frequency with which
they engaged in the behavior indicated. Sample items
include ‘I discuss my health concerns with health pro-
fessionals’ and ‘I follow a planned exercise program.’
In the original study, internal consistency was found
to be .94 for the global scale, and construct validity
was .68 (Walker & Hill-Polerecky, 1996). Cronbach’s
alpha in the current study was .97.
Optimism The revised Life Orientation Test
[LOT-R] (Scheier et al., 1994) was used to measure
optimism exhibited by the mothers in this study. Par-
ticipants were asked to rate their agreement with the
statements on a 5-point Likert scale (0 = strongly
disagree to 4 = strongly agree). A sample item was, ‘In
uncertain times, I usually expect the best.’ In a study
with 2,055 undergraduate students, the revised Life
Orientation Test reported internal consistency was
.78 for the six items used and adequate construct
validity (Scheier et al., 1994). Internal consistency in
the current study was .89 for the six items.” (Gill &
Loh, 2010, p. 351)
The three Likert scales used in this study seemed
to have strong reliability from previous research
(Cronbach’s alpha = 0.78 to 0.94) and were reliable in
this study (Cronbach’s alpha = 0.75 to 0.97). However,
the researchers provide very limited specific informa-
tion about the validity of the scales, especially the
PSS-10 and LOT-R.
Developing a Plan for Data Collection
and Analysis
Data collection is the precise, systematic gathering of
information relevant to the research purpose or the
specific objectives, questions, or hypotheses of a
study. The data collected in quantitative studies are
usually numerical. Planning data collection will enable
you to anticipate problems that are likely to occur and
to explore possible solutions. Usually, detailed proce-
dures for implementing a treatment and collecting
data are developed, with a schedule that identifies
the initiation and termination of the process (see
Chapter 20).
Planning data analysis is the final step before the
study is implemented. The analysis plan is based on
(1) the research objectives, questions, or hypotheses,
(2) the data to be collected, (3) research design,
(4) researchers’ expertise, and (5) availability of com-
puter resources. Several statistical analysis techniques
are available to describe the sample, examine rela­
tionships, or determine significant differences within
studies. Most researchers consult a statistician for
assistance in developing an analysis plan.
Implementing the Research Plan
Implementing the research plan involves intervention
implementation, data collection, data analysis, inter-
pretation of research findings, and, sometimes, a
pilot study.
46 UNIT ONE  Introduction to Nursing Research
Pilot Study
A pilot study is commonly defined as a smaller
version of a proposed study that is conducted to refine
the methodology. It is developed much like the pro-
posed study, using similar subjects, the same setting,
the same treatment, and the same data collection and
analysis techniques. However, you could use a pilot
study to develop various steps in the research process.
For example, you could conduct a pilot study to
develop and refine an intervention, a measurement
method, a data collection tool, or the data collection
process. Thus, a pilot study could be used to develop
a research plan rather than to test an already developed
plan.
Some of the reasons for conducting pilot studies
are as follows (Arain, Campbell, Cooper, & Lancaster,
2010; Feeley et al., 2009; Hertzog, 2008; Prescott &
Soeken, 1989):
• To determine whether the proposed study is feasi-
ble (e.g., are the subjects available, does the
researcher have the time and money to do the
study?).
• To develop or refine a research treatment or inter-
vention (see Chapter 14).
• To develop a protocol for the implementation of an
intervention.
• To identify problems with a study design.
• To determine whether the sample is representative
of the population or whether the sampling tech-
nique is effective.
• To examine the reliability and validity of the
research instruments.
• To develop or refine data collection instruments.
• To refine the data collection and analysis plan.
• To give the researcher experience with the
subjects, setting, methodology, and methods of
measurement.
• To try out data analysis techniques.
Hayward et al. (2007) believed that conducting a
pilot study improved the strength of their study design
and directed their development of a quality proposal
for a large multisite clinical trial that received external
grant support. Thus, conducting pilot studies has the
potential to improve the development, funding, and
implementation of future studies.
Data Collection
In quantitative research, data collection involves
obtaining numerical data to address the research
objectives, questions, or hypotheses. To collect data,
you must obtain consent or permission from the setting
or agency where the study is to be conducted and from
potential subjects. Frequently, study participants are
asked to sign a consent form, which describes the
study, promises the subjects confidentiality, and indi-
cates that they can stop participation at any time (see
Chapter 9).
During data collection, the study variables are mea-
sured through a variety of techniques, such as observa-
tion, interview, questionnaires, scales, and physiological
measurement methods. In a growing number of
studies, nurses measure physiological variables with
high-technology equipment. The data are collected
and recorded systematically for each subject and are
organized to facilitate computer entry or are directly
recorded in the computer (Ryan-Wenger, 2010). The
procedure for data collection is usually identified in
the Methods section of a study report. Gill and Loh
(2010) provided the following description of their data
collection process. This section would have been
strengthened by a more detailed discussion of the
online survey format and the process for collection
and management of data.
“Procedure
New primiparous mothers were asked to complete
an online anonymous survey that could be accessed
from a link available from an online baby forum Web
page. This survey took approximately 20 minutes to
complete. Implied consent was given by the partici-
pant on the completion and submission of the online
questionnaire; this was stated explicitly in the intro-
ductory sheet on the first page of the online survey.
Ethics approval was granted for this study and all
requirements pertaining to this approval were met.”
(Gill & Loh, 2010, pp. 350–351)
Data Analysis
Data analysis reduces, organizes, and gives meaning
to the data. The analysis of data from quantitative
research involves the use of (1) descriptive analysis
techniques (see Chapter 22) to describe demographic
variables and study variables; (2) statistical techniques
to test proposed relationships among variables (see
Chapter 23); (3) analysis techniques to make predic-
tions (see Chapter 24); and (4) analysis techniques to
examine group differences, such as the difference
between the experimental or intervention group and
the control group (see Chapter 25). Computers are
used to perform most statistical analyses, so Chapter
21 provides a discussion of the computer software
programs used for conducting data analysis.
The choice of analysis techniques implemented is
determined primarily by the research objectives,
 CHAPTER 3  Introduction to Quantitative Research 47
questions, or hypotheses; the research design; and the
level of measurement achieved by the research instru-
ments. Gill and Loh (2010) used frequencies and per-
centages to analyze the demographic variables of
marital status, nationality, employment, education,
and income. The age and months since the baby’s birth
were described with the use of means and standard
deviations (see Table 3-2). Perceived stress, optimism,
and health-promoting behavior variables were (1)
described with means, standard deviations, and ranges,
(2) examined for relationships with Pearson product-
moment correlation, and (3) examined for prediction
with regression analysis.
“Analysis Approach
The means [M], standard deviations [SD], ranges, and
Cronbach’s α of the study … variables are presented in
Table 3-2. Correlations between the main study vari-
ables are depicted in Table 3-3. All correlations were in
the predicted direction. [Perceived stress was signifi-
cantly negatively correlated with optimism, −.60* and
health-promoting behaviors, −.67*. Optimism and
health-promoting behaviors were significantly positively
correlated, .70*. The * indicates the results were signifi-
cant at p < .01]. On average, participants had missing
data on 22.8% of the variables. The missing data in the
sample were assessed to be missing at random.…
Fifteen cases were deleted for having excessive …
missing, data on greater than 50% of the variables.” (Gill
& Loh, 2010, p. 351)
“Results
… Results indicated that perceived stress was signifi-
cantly negatively related to health-promoting behaviors,
β = −.60, t(167) = −10.27, and uniquely accounted for
33% of the variance [change] in health-promoting
behaviors, F(6,167) = 25.74, p < .001 [see regression
TABLE 3-3  Correlations among the Main Study Variables (N = 174)
Variable 1 2
1.  Perceived stress —
2.  Optimism –.60*
3.  Health-promoting behavior –.67
4.  Income –.30*
5.  Age –.21*
6.  Months since birth –.08
Asterisk (*) indicates a significant correlation.
From Gill, R. M., & Loh, J. M. (2010). The role of optimism in health-promoting behaviors in new primiparous mothers. Nursing Research, 59(5), 348–355.
Table is on page 352.
The Results Section in the Gill and Loh (2010)
article was clearly organized by the study hypotheses.
All four of the hypotheses were supported, indicating
support for the study framework. These analyses are
somewhat complicated at this point in the text, but we
encourage you to examine each of the tables for key
ideas. Table 3-3 presents a correlation matrix of the
results obtained from Pearson correlation analysis of
the study variables (perceived stress, optimism, and
health-promoting behaviors) and the demographic
variables (income, age, and months since birth) (see
Chapter 23). These results were examined for signifi-
cance using the t-test (t value), and the significant
analysis in Table 3-4]. Hypothesis 1 was therefore sup-
ported because perceived stress was related inversely
to health-promoting behaviors.
Hypothesis 2 was supported because perceived
stress was related inversely to optimism, β = −.55,
t(167) = −8.56, and uniquely accounted for 27% of the
variance in optimism, F(6,167) = 17.50, p < .001.
In the third analysis, optimism positively influenced
health-promoting behaviors, after controlling for the
effects of perceived stress on health-promoting behav-
iors, β = .46, t(166) = 7.36, and uniquely accounted
for 13% of the variance in optimism, F(7,166) = 36.38,
p < .001.… Hypothesis 3 therefore was supported
because optimism was related positively to health-
promoting behaviors after controlling for perceived
stress.
In the fourth analysis, perceived stress was related to
health-promoting behaviors after controlling for the
effects of optimism, β = −.35, t(166) = −5.77, and
explained an additional 8% of the variance in health-
promoting behaviors, optimism, F(7,166) = 36.83, p <
.001 [see Table 3-4]… Hence Hypothesis 4 was sup-
ported.” (Gill & Loh, 2010, p. 352)
3 4 5
—
.70* —
.26* .38* —
–.34* .27* .45* —
–.03 .02 — —
48 UNIT ONE  Introduction to Nursing Research

TABLE 3-4  Standard Multiple Regression Testing the Direct Effects of Perceived Stress and Optimism
on Health-Promoting Behaviors: Test of Mediation (N = 174)
Analysis and Predictor Variable β SE R2 F Criterion Variable
Analysis 1*: –.60 0.42 .48 25.74§ HPB
Perceived stress
Analysis 2†: –.55 3.44 .39 17.50* Optimism
Perceived stress
Analysis 3: .46
Optimism
Analysis 4‡: –.35 0.37 .61 36.83* HPB
Perceived stress

HPB, health-promoting behavior.

*Effect of predictor variable on criterion at df = 6,167.
†
Effect of predictor variable on mediator at df = 6,167.
‡
Effects of predictor variable and mediator on criterion variable at df = 7,166.
§
P < .01 level.
N = Sample size for regression analysis calculation.
df = Degrees of freedoms for regression analysis.
β = Beta coefficients for regression analysis.
SE = Standard error.
R2 = Square of Regression statistic R to determining percent of variance explained by predictor variables.
F = Statistic for Analysis of Variance to determine significance of predictor variables.
From Gill, R. M., & Loh, J. M. (2010). The role of optimism in health-promoting behaviors in new primiparous mothers. Nursing Research, 59(5), 353.

correlations have an asterisk (*). Table 3-4 presents
the regression analysis results, predicting the effects
of perceived stress and optimism on health-promoting
behaviors in new first-time mothers. The significance
of the regression analysis results was demonstrated
with F-values (see Chapter 24). The amount of missing
data for the study variables was a concern, and it is
unclear how this lack might have influenced the study
results.
Interpreting Research Outcomes
The results obtained from data analysis require inter-
pretation to be meaningful. Interpretation of research
outcomes involves (1) examining the results from data
analysis, (2) exploring the significance of the findings,
(3) identifying study limitations, (4) forming conclu-
sions, (5) generalizing the findings, (6) considering the
implications for nursing, and (7) suggesting further
studies. The study results from data analyses are trans-
lated and interpreted to become findings, and these
findings are synthesized to form conclusions. Study
conclusions are influenced by the limitations of the
study. Limitations are restrictions or problems in a
study that may decrease the generalizability of the
findings. Study limitations often include a combina-
tion of theoretical and methodological weaknesses.
Theoretical weaknesses in a study might include a
poorly developed study framework and unclear con-
ceptual definitions of variables. The limited conceptual
definitions of the variables might decrease the opera-
tionalization or measurement of the study variables.
Methodological limitations result from factors such
as nonrepresentative samples, weak designs, a single
setting, limited control over treatment (intervention),
instruments with limited reliability and validity, limited
control over data collection, and improper use of sta-
tistical analyses. These study limitations can limit the
credibility of the findings and conclusions and restrict
the population to which the findings can be general-
ized. The study conclusions provide a basis for iden-
tifying nursing implications and suggesting further
studies (see Chapter 26). In the excerpts that follow,
Gill and Loh (2010) discuss their study findings, limi-
tations, conclusions, suggestions for further research,
and implications for practice for their study. Because
of the study limitations, the researchers limit the gen-
eralization of the findings and provide direction for
further research. The discussion section would have
been strengthened by a clarification of the implications
of the study findings for practice.
“Discussion
… Finally, results indicated that optimism partially
mediated the relationship between perceived stress
and health-promoting behaviors, suggesting a protec-
tive role of dispositional optimism in the sample of
 CHAPTER 3  Introduction to Quantitative Research 49
new primiparous mothers. This finding is consistent
with previous research, in which optimism has served
as a mediator of a stress relationship … and pro-
moted positive psychological and behavioral health
outcomes in mothers.… Caution must be observed
when interpreting the results of this study. Given the
cross-sectional data, a causal relationship cannot be
confirmed between perceived stress and optimism
in new mothers [findings]. In addition, the sample
group predominantly consists of high-income, well-
educated, and computer-literate women who may
not be representative of most new primiparous
mothers. Expanding on this sample group may be an
area for future research. Additional limitations include
using retrospective reporting and self-reporting
scales, which can lead to response bias. A further
limitation is the use of the PSS-10, which measures
perceptions of stress that have occurred during the
last month. Stress relating to the mothers’ daily lives
may be perceived differently, depending on how long
ago they gave birth and subsequently how old their
child is [limitations].… Therefore, although assessing
the perceived stress of the new mother is interesting,
it may be more useful for future research to use a
measure specifically designed to capture stress
response at different milestones of the baby’s
development.…
The results indicated that optimism only partially
mediated the relationship between perceived stress
and health-promoting behaviors in new primiparous
mothers.… Social support received from family or
new mother groups also may influence positive well-
being in new mothers. Further research is therefore
recommended to explore some of these issues.
In summary, optimism serves as a partial mediator
for new primiparous mothers to self-regulate their
perceptions of stress positively and thereby to facili-
tate their involvement in health-promoting behaviors.
Health-promoting actions have many positive benefits
for new mothers [conclusion], including building iden-
tity as a mother, enhancing positive well-being, and
minimizing the risk of future illness outcomes through
exercise” [implications for practice]. (Gill & Loh, 2010,
p. 354)
Communicating Research Findings
Research is not considered complete until the findings
have been communicated. Communicating research
findings involves developing and disseminating a
research report to appropriate audiences; the research
report is disseminated through presentations and pub-
lication (see Chapter 27). The Gill and Loh (2010)
study was published in one of the most prestigious
nursing journals, Nursing Research.
Types of Quantitative Research
This textbook deals with four types of quantitative
research: (1) descriptive, (2) correlational, (3) quasi-
experimental, and (4) experimental. The level of exist-
ing knowledge for the research problem influences the
type of research planned. When little knowledge is
available, descriptive studies are often conducted. As
the knowledge level increases, correlational, quasi-
experimental, and experimental studies are imple-
mented. This section builds on the content in Chapter
2 and identifies the purpose of each quantitative
research approach. The chapter concludes with the
steps of the research process from a published quasi-
experimental study.
Descriptive Research
The purpose of descriptive research is to explore and
describe phenomena in real-life situations. This
approach is used to generate new knowledge about
concepts or topics about which limited or no research
has been conducted. Through descriptive research,
concepts are described and relationships are identified
that provide a basis for further quantitative research
and theory testing. The study by Gill and Loh (2010)
presented earlier in this chapter has a descriptive com-
ponent, which included description of the study vari-
ables optimism, perceived stress, and health-promoting
behaviors of new first-time mothers. The descriptive
results of this study can be clearly identified in
Table 3-2.
Correlational Research
Correlational research examines linear relationships
between or among two or more variables and deter-
mines the type (positive or negative) and degree
(strength) of the relationship. The strength of a rela-
tionship varies from −1 (perfect negative correlation)
to +1 (perfect positive correlation), with 0 indicating
no relationship. The positive relationship indicates
that the variables vary together—that is, the two vari-
ables either increase or decrease together. The nega-
tive or inverse relationship indicates that the variables
vary in opposite directions; thus, as one variable
increases, the other decreases (see the results pre-
sented earlier from Gill & Loh, 2010 study). The
descriptive correlational study conducted by Gill and
Loh (2010), presented earlier in this chapter, provides
an example of the steps of the quantitative research
50 UNIT ONE  Introduction to Nursing Research
process for correlational research with a predictive
correlational design. The results of the relationships
among the study variables are clearly presented in
Table 3-3. Regression analysis results in Table 3-4
indicated that perceived stress and optimism signifi-
cantly predicted health-promoting behaviors of pri-
miparous mothers in this study.
Quasi-Experimental Research
The purpose of quasi-experimental research is
to examine cause-and-effect relationships among
Steps of the Research Process in a Quasi-
Experimental Study
Research Problem
“Nearly one in three, or approximately 65 million adults
in the United States have hypertension, defined as (a)
having systolic blood pressure (SBP) of 140 mm Hg or
higher or diastolic blood pressure (DBP) of at least
90 mm Hg or higher, (b) taking antihypertensive medi-
cation, or (c) being told at least twice by a physician or
other health professional about having high blood pres-
sure (BP) (American Heart Association [AHA], 2004;
AHA Statistics Committee & Stroke Statistics Subcom-
mittee [AHASC], 2006; Fields et al., 2004).… Estimated
direct and indirect costs associated with hypertension
total $63.5 billion (AHA, 2004).… The crisis of high BP
(HBP) is particularly apparent among African Ameri-
cans; their prevalence of HBP is among the highest
in the world.… Unless healthcare professionals can
improve care for individuals with hypertension, approxi-
mately two thirds of the population will continue to
have uncontrolled BP and face other major health risks
(Chobanian et al., 2003) [problem significance].… Self-
monitoring of BP has been shown to lead to improved
BP control.… The efficacy of BP TM has been tested
using a randomized controlled design, but the sample
was small (Artinian et al., 2001) [problem background].…
The influence of TM on BP control warrants further
study” [problem statement]. (Artinian et al., 2007, pp.
312–314)
Research Purpose
“The purpose of this randomized controlled trial with
urban African Americans was to compare usual care
(UC) only with BP telemonitoring (TM) plus UC to
determine which leads to greater reduction in BP from
baseline over 12 months of follow-up, with assessments
selected independent and dependent variables. Quasi-
experimental studies in nursing are conducted to
determine the effects of nursing interventions or treat-
ments (independent variables) on patient outcomes
(dependent variables) (Cook & Campbell, 1979;
Shadish et al., 2002). Artinian et al. (2007) conducted
a quasi-experimental study to determine the effects of
nurse-managed telemonitoring (TM) on the BPs of
African Americans. The steps for this study are
described here and illustrated with excerpts from the
example study.
at 3, 6, and 12 months postbaseline.” (Artinian et al.,
2007, p. 313)
Review of Literature
The literature review for this study included relevant,
current studies that the researchers summarized to
identify what is known and not known about the
impact of TM on BP. The sources were current and
ranged in publication dates from 1998 to 2005, with
the majority of the studies published in the last 5 years.
The study was accepted for publication on May 31,
2007, and published in the September/October 2007
issue of Nursing Research. Artinian et al. (2007, p. 314)
summarized the current knowledge about the effect of
TM on BP by stating, “Although promising, the effects
of TM on BP have been tested in small, sometimes
nonrandomized, samples, with one study suggesting
that patients may not always adhere to measuring their
BP at home. The influence of TM on BP control war-
rants further study.”
Framework
Artinian et al. (2007) developed a model that identified
the theoretical basis for their study. The model pre-
sented in Figure 3-3 indicates that “nurse-managed TM
is an innovative strategy that may offer hope to hyper-
tensive African Americans who have difficulty accessing
care for frequent BP checks.… In other words, TM may
lead to a reduction in opportunity costs or barriers for
obtaining follow-up care by minimizing the contextual
risk factors that interfere with frequent healthcare
visits.… Combined with information about how to
control hypertension, TM may both help individuals gain
conscious control over their HBP and contribute
to feelings of confidence for carrying out hypertension
self-care actions.… Home TM appeared to contribute
to individuals’ increased personal control and

Blood Pressure (BP) Self
Monitoring
• Moves hypertension from state of
silence to salience.
• May help individual:
• connect symptoms of high
BP to actual BP level;
• see effects of lapses from
medical regimen on BP;
• see effects of stress or
worry on BP level;
• gain conscious control
over their high BP;
• gain feelings of confidence
that they are capable of
carrying out hypertension
self-care actions;
• gain sense of increased
personal control, self-efficacy or
self-responsibility for managing
their BP.
Figure 3-3  Theoretical basis for the effects of telemonitoring on blood pressure (BP). (From Artinian, N. T., Flack, J. M., Nordstrom, C. K., Hockman,
E. M., Washington, O. G. M., Jen, K. C., et al. [2007]. Effects of nurse-managed telemonitoring on blood pressure at 12-month follow-up among urban
African Americans. Nursing Research, 56[5], 313.)
self-responsibility for managing their BP, which ulti-
mately led to improved BP control (Artinian et al., 2004;
Artinian et al., 2001).” (Artinian et al., 2007, pp.
313–314)
The framework for this study was based on tentative
theory that was developed from the findings of previous
research by Artinian et al. (2004; 2001) and other inves-
tigators. This framework map (see Figure 3-3) and
description provide a basis for interpreting the study
findings and giving them meaning. The framework mainly
provides a basis for the TM intervention and would be
strengthened by including concepts to represent the
outcomes or more of a discussion of the SBP and DBP.
Hypothesis Testing
“H1: Individuals who participate in UC plus nurse-
managed TM will have a greater reduction in BP from
baseline at 3-, 6-, and 12-month follow-up than would
individuals who receive UC only.” (Artinian et al., 2007,
p. 317)
CHAPTER 3  Introduction to Quantitative Research 51
Transmission of Blood
Pressure Readings Over
Existing Telephone Lines
Using a Toll Free Number
• Facilitates access to healthcare
for hypertension without the
need for clinic or office visits.
• Reduces opportunity costs
or barriers for obtaining
hypertension follow-up care.
Blood Pressure • Minimizes contextual risk
Telemonitoring
factors that interfere with
frequent healthcare visits.
• Provides a greater number
of BP readings than can be
provided by infrequent clinic
visits; healthcare providers
have more data upon which
to base treatment decisions
to achieve better BP
control.
Variables
The independent variable was the TM Program and the
dependent variables were SBP and DBP. Only the TM
Program, SBP, and DBP conceptual and operational defi-
nitions are presented as examples. The conceptual defi-
nitions are derived from the study framework and/or
are included as part of the literature review. In this
study, the operational definitions for the dependent vari-
ables are found in the methods section under “Outcome
Measurement,” and the independent variable is opera-
tionalized under an “Intervention” heading.
Independent Variable: TM Program
Conceptual Definition: TM program is an innovative
strategy that may offer hope to hypertensive African
Americans to reduce their opportunity costs and barri-
ers for obtaining follow-up care for BP management
(Artinian et al., 2007).
Operational Definition: TM “refers to individuals self-
monitoring their BP at home, then transmitting the BP
52 UNIT ONE  Introduction to Nursing Research
readings over existing telephone lines using a toll-free
number” (Artinian et al., 2007, p. 313). The BP readings
were reviewed by the care providers, with immediate
feedback provided to the patients about their treatment
plan.
Dependent Variables: SBP and DBP
Conceptual Definition: SBP and DBP are an indication of
the patient’s blood pressure control and ultimately the
management of his or her hypertension.
Operational Definition: The outcome of SBP and DBP
were measured with the electronic BP monitor (Omron
HEM-737 Intellisense, Omron Health Care, Inc.)
(Artinian et al., 2007).
Design
“A randomized, two-group, experimental, longitudinal
design was used. The treatment group received nurse-
managed TM and the control group received enhanced
UC. Data were collected at baseline and 3-, 6-, and
12-month follow-ups. Follow-up periods were chosen
to allow some comparisons of findings from this study
with those from other TM studies that have been
reported in the literature.…
Because imbalances in the number of participants
assigned to each group could occur with simple random-
ization, block stratified randomization for antihyperten-
sive medication use was performed to ensure an equal
number of participants taking antihypertensive medica-
tions in each group. It was important to avoid an imbal-
ance in the number of participants taking antihypertensive
medications in one of the groups because an imbalance
could confound the effects of the TM intervention.
Sequentially numbered computer-generated random-
ization assignments were determined prior to the start
of data collection, but participants were not notified of
the group assignment until after baseline data were col-
lected. To keep data collectors blinded to group assign-
ment, the study’s project manager informed the
participants of their group assignment by mail or tele-
phone within a week of their baseline interview. Data
collectors were trained not to ask participants about
group assignment and to ask participants not to
reveal their assignment to them.” (Artinian et al., 2007,
pp. 314–315)
This study has a strong design with groups stratified
according to hypertension medication use. The stratifi-
cation ensured that the TM group was similar to the
UC group regarding the number of participants taking
hypertension medications. Thus, differences found
between the groups would be due to the treatment and
not to differences in the group participants. The partici-
pants were randomly assigned to the treatment or com-
parison groups by a computer with the data collectors
blinded to, or kept unaware of, the group assignment.
Sample
“African Americans with hypertension [population] were
recruited through free BP screenings offered at com-
munity centers, thrift stores, drug stores, and grocery
stores located on the east side of Detroit [natural
settings].…
Inclusion criteria were ≥18 years of age; SBP ≥
140 mm Hg or DBP ≥ 90 mm Hg, unless self-identified
as a diabetic or with a history of chronic kidney disease,
then SBP ≥ 130 mm Hg or DBP ≥ 80 mm Hg; access to
a land-based telephone in own residence (owned or
rented) [natural setting]; oriented to person, time, and
place; English speaking; and intent to remain in Detroit
for the next year. Exclusion criteria were arm circumfer-
ence > 17.5 in.; history of dementia, mental illness,
terminal cancer, advanced liver disease, or hemodialysis;
and self-reported illicit drug use or alcohol abuse as
measured by the CAGE (Cut, Annoyed, Guilty, Eye-
opener) questions.” (Artinian et al., 2007, p. 315)
The population and the recruitment process are
clearly described. The sample criteria for including and
excluding subjects from the study were detailed and
provided a means of identifying patients with hyperten-
sion. The sample size was 387 (194 in the TM group
and 193 in the UC group), and there was a 13% attri-
tion, or loss of subjects, over the 12-month study, which
are both study strengths. The original sample was a
nonrandom sample of convenience, and the subjects
were recruited through free BP screenings offered at a
variety of locations. The nonrandom sampling intro-
duces a potential for sampling bias, because not all
potential subjects had an equal chance to participate in
the study. . The subjects participating in free BP screen-
ings might differ in some way from those who do not
seek screening. These study elements are more charac-
teristic of a quasi-experimental study, but the study does
have an experimental type of design (Kerlinger & Lee,
2000; Shadish et al., 2002).
Procedures
Artinian et al. (2007) detailed the nurse-managed TM
intervention that was presented earlier in this chapter
and provided in entirety on pages 315-316 in the
research article. The outcome measure of BP was mea-
sured with the electronic Omron BP monitor “after a
5-minute rest period; at least two BPs were measured,
 CHAPTER 3  Introduction to Quantitative Research 53
and the average of all was used for analyses. Participants
wore unrestrictive clothing and sat next to the inter-
viewer’s table, their feet on the floor; their back sup-
ported; and their arm abducted, slightly flexed, and
supported at heart level by the smooth, firm surface of
a table” (Artinian et al., 2007, pp. 316–317).
“Most of the data were collected during 2-hour struc-
tured face-to-face interviews and brief physical exams,
which were conducted by trained interviewers in a
private room at one of the project-affiliated neighbor-
hood community centers. Mailed postcards provided
interview appointment reminders 1 week before the
scheduled interview; telephone call reminders were
made the evening before the interview.… Participants
were compensated $25.00 after the completion of each
interview” (Artinian et al., 2007, p. 316). The study was
approved by the Wayne State University Human Inves-
tigation Committee, and all participants signed consent
forms indicating their willingness to participate in
this study.
Results
“The hypothesis was supported partially by the data.
Overall, the TM intervention group had a greater reduc-
tion in SBP (13.0 mm Hg) than the UC group did
(7.5 mm Hg; t = −2.09, p = 0.04) from baseline to the
Experimental Research
The purpose of experimental research is to examine
cause-and-effect relationships between independent
and dependent variables under highly controlled con-
ditions (Shadish et al., 2002). Researchers exert high
control over the planning and implementation of
experimental studies, including experimental designs,
random samples, highly structured interventions, and
quality measurement methods. Often these studies are
conducted in highly controlled settings, such as labo-
ratories or research units, with humans or animals. The
study by Byerley et al. (2010) introduced earlier in
this chapter is an experimental study to examine the
causes of cancer-related cachexia using male rats that
was conducted in a laboratory setting. To improve
your understanding of the steps of the research process,
read this study and identify the steps of quantitative
research process outlined in this chapter.
KEY POINTS
• Nurses use a broad range of quantitative ap-
proaches—including descriptive, correlational,
12-month follow-up. Although the TM intervention
group had a greater reduction in the DBP (6.3 mm Hg)
compared with the UC group (4.1 mm Hg), the differ-
ences were not statistically significant (t = −1.56, p =
0.12)” (Artinian et al., 2007, pp. 317–318). The study
results were clearly presented in tables and figures.
Discussion
“The nurse-managed TM group experienced both clini-
cally and statistically significant reductions in SBP
(13.0 mm Hg) and clinically significant reductions in
DBP (6.3 mm Hg) over a 12-month monitoring period
[study conclusions].… The BP reductions achieved here
are important results, which, if maintained over time,
could improve care and outcomes significantly for urban
African Americans with hypertension.… This may mean
that an individual could avoid starting a drug regimen or
may achieve BP control using a one-drug regimen rather
than a two-drug regimen and thus be at risk for fewer
medication side effects [implications of the findings for
nursing practice].… Future research needs to determine
if this intervention effect maintained over time leads to
reducing the number of complications associated with
uncontrolled BP and if it leads to reducing the number
of drugs necessary to achieve BP control.” (Artinian
et al., 2007, pp. 320–321)
quasi-experimental, and experimental—to develop
nursing knowledge.
• Some of the concepts relevant to quantitative
research are (1) basic research, (2) applied research,
(3) rigor, and (4) control.
• Basic, or pure, research is a scientific investigation
that involves the pursuit of “knowledge for knowl-
edge’s sake” or for the pleasure of learning and
finding truth.
• Applied, or practical, research is a scientific inves-
tigation conducted to generate knowledge that will
directly influence or improve clinical practice.
• Rigor involves discipline, scrupulous adherence to
detail, and strict accuracy.
• Control involves the imposing of “rules” by the
researcher to decrease the possibility of error and
thus increase the probability that the study’s find-
ings are an accurate reflection of reality.
• The quantitative research process involves concep-
tualizing a research project, planning and imple-
menting that project, and communicating the
findings.
• The steps of the quantitative research process are
as follows:
54 UNIT ONE  Introduction to Nursing Research
1. Formulating a research problem and purpose
identifies an area of concern and the specific
focus or aim of the study.
2. Reviewing relevant literature allows the
researcher to build a picture of what is known
about a particular situation or phenomenon and
identify the knowledge gaps that exist.
3. Developing a framework guides the develop-
ment of the study and enables the researcher
to link the findings to the body of knowledge
in nursing.
4. Formulating research objectives, questions,
or hypotheses allows the researcher to bridge
the gap between the more abstractly stated
research problem and purpose and the study
design and plan for data collection and
analysis.
5. Operationalizing study variables involves
developing a conceptual definition and opera-
tional definition for each variable.
6. Selecting a research design directs the selec-
tion of a population, sampling procedure,
implementation of the intervention, methods
of measurement, and a plan for data collection
and analysis.
7. Defining the population and sample deter-
mines who will participate in the study.
8. Implementation of the study intervention
involves the detailed development of the steps
or elements of the intervention and the consis-
tent implementation of the intervention during
the study.
9. Selecting methods of measurement involves
determining the best method(s) to measure
each study variable.
10. Developing a plan for data collection and
analysis directs the precise, systematic gather-
ing of information relevant to the research
purpose or the specific objectives, questions,
or hypotheses of a study and involves the
selection of appropriate statistical techniques
to analyze the study data.
11. Implementing the research plan involves inter­
vention implementation, data collection, data
analysis, and interpretation of research
outcomes.
12. Interpreting the research outcomes involves
examining the results from data analysis,
exploring the significance of the findings,
identifying study limitations, forming conclu-
sions, generalizing findings as appropriate,
considering implications for nursing practice,
and suggesting further studies.
13. Communicating findings includes the develop-
ment and dissemination of a research report to
appropriate audiences through presentations
and publication.
• This chapter introduces four types of quantita-
tive research: descriptive, correlational, quasi-
experimental, and experimental. Examples from
published studies are used to illustrate the steps
of the quantitative research process for these
different types of quantitative research.
REFERENCES
Aberson, C. L. (2010). Applied power analysis for the behavioral
sciences. New York, NY: Routledge Taylor & Francis Group.
American Heart Association (AHA). (2004). Heart disease and
stroke statistics–2005 update. Dallas, TX: Author.
American Heart Association (AHA). (2011). Call to action on use
and reimbursement for home blood pressure monitoring. Retrieved
from http://my.americanheart.org/professional/General/Call-to-
Action-on-Use-and-Reimbursement-for-Home-Blood-Pressure-
Monitoring_UCM_423866_Article.jsp/.
American Health Association Statistics Committee and Stoke Sta-
tistics Subcommittee. (2006). Heart disease and stroke statistics—
2006 update. Circulation, 113(6), e85–e152.
Arain, M., Campbell, M. J., Cooper, C. L., & Lancaster, G. A.
(2010). What is a pilot or feasibility study? A review of current
practice and editorial policy. BMC Medical Research Methodol-
ogy, 10(1), 67.
Artinian, N. T., Flack, J. M., Nordstrom, C. K., Hockman, E. M.,
Washington, O. G. M., Jen, K. C., et al. (2007). Effects of nurse-
managed telemonitoring on blood pressure at 12-month follow-up
among urban African Americans. Nursing Research, 56(5),
312–322.
Artinian, N. T., Washington, O. G., Klymko, K. W., Marbury,
C. M., Miller, W. M., & Powell, J. L. (2004). What you need to
know about home blood pressure telemonitoring, but may not
know to ask. Home Healthcare Nurse, 22(10), 680–686.
Artinian, N. T., Washington, O. G., & Templin, T. N. (2001). Effects
of home telemonitoring and community-based monitoring on
blood pressure control in urban African Americans: A pilot study.
Heart & Lung, 30(3), 191–199.
Bond, E. F. & Heitkemper, M. M. (1987). Importance of basic
physiologic research in nursing science. Heart & Lung, 16(4),
347–349.
Borglin, G. & Richards, D. A. (2010). Bias in experimental nursing
research: Strategies to improve the quality and explanatory power
of nursing science. International Journal of Nursing Studies,
47(1), 123–128.
Brissette, I., Scheier, M. F., & Carver, C. S. (2002). The role of
optimism in social network development, coping, and psychologi-
cal adjustment during a life transition. Journal of Personality and
Social Psychology, 82(1), 102–111.
Brown, S. J. (2009). Evidence-based nursing: The research-practice
connection. Sudbury, MA: Jones and Bartlett.
Byerley, L. O., Lee, S. H., Redmann, S., Culberson, C., Clemens,
M., & Lively, M. O. (2010). Evidence for a novel serum factor
 CHAPTER 3  Introduction to Quantitative Research 55
distinct from zinc alpha-2 glycoprotein that promotes body fat loss
early in the development of cachexia. Nutrition and Cancer,
62(4), 484–494.
Campbell, D. T. & Stanley, J. C. (1963). Experimental and quasi-
experimental designs for research. Chicago, IL: Rand McNally.
Carver, C. S. & Gaines, J. G. (1987). Optimism, pessimism, and
postnatal depression. Cognitive Therapy and Research, 11(4),
449–462.
Chen, C. M., Kuo, S. F., Chou, Y. H., & Chen, H. C. (2007). Post-
partum Taiwanese women: Their postpartum depression, social
support and health-promoting lifestyle profiles. Journal of Clini-
cal Nursing, 16(8), 1550–1560.
Chobanian, A., Bakris, G., Black, H., Cushman, W., Green, L., Izzo,
J., Jr., et al. (2003). Seventh report of the Joint National Commit-
tee on Prevention, Detection, Evaluation, and Treatment of High
Blood Pressure. Hypertension, 42(6), 1206–1252.
Cohen, S. & Williamson, G. (1988). Perceived stress in a probability
sample of the United States. In S. Spacapam & S. Oskamp (Eds.),
The social psychology of health (pp. 31–67). Newbury Pare, CA:
Sage.
Cook, T. D. & Campbell, D. T. (1979). Quasi-experimentation:
Design and analysis issues for field settings. Chicago, IL: Rand
McNally.
Dabl Educational Trust. (2005). Device table: Upper arm devices
for self-measurement of blood pressure. Retrieved from http://
www.dableducational.com/sphygmomanometers.html/.
Fawcett, J. & Garity, J. (2009). Evaluating research for evidence-
based nursing practice. Philadelphia, PA: F. A. Davis.
Feeley, N., Cossette, S., Côté, J., Héon, M., Stremler, R., Martorella,
G., et al. (2009). The importance of piloting an RCT intervention.
Canadian Journal of Nursing Research, 41(2), 85–99.
Fields, L., Burt, V., Cutler, J., Hughers, J., Roccella, E., & Sorlie,
P. (2004). The burden of adult hypertension in the United States
1999-2000: A rising tide. Hypertension, 44(4), 398–404.
Fisher, Sir, R. A. (1935). The designs of experiments. New York,
NY: Hafner.
Forbes, A. (2009). Clinical intervention research in nursing. Inter-
national Journal of Nursing Studies, 46(4), 557–568.
Gill, R. M. & Loh, J. M. (2010). The role of optimism in health-
promoting behaviors in new primiparous mothers. Nursing
Research, 59(5), 348–355.
Groth, S. W. & David, T. (2008). New mothers’ views of weight
and exercise. American Journal of Maternal Child Nursing, 33(6),
364–370.
Grove, S. K. (2007). Statistics for health care research: A practical
workbook. St. Louis, MO: Saunders Elsevier.
Hayward, K., Campbell-Yeo, M., Price, S., Morrison, D., Whyte,
R., Cake, H., et al. (2007). Cobedding twins: How pilot study
findings guided improvements in planning a larger multicenter
trial. Nursing Research, 56(2), 137–143.
Hertzog, M. A. (2008). Considerations in determining sample
size for pilot studies. Research in Nursing & Health, 31(2),
180–191.
Homans, G. (1965). Group factors in worker productivity. In H.
Proshansky & B. Seidenberg (Eds.), Basic studies in social psy-
chology (pp. 592–604). New York, NY: Holt, Rinehart & Winston.
Kaplan, A. (1964). The conduct of inquiry: Methodology for behav-
ioral science. New York. NY: Chandler.
Kerlinger, F. N. & Lee, H. B. (2000). Foundations of behavioral
research (4th ed.). New York, NY: Harcourt Brace.
Lazarus, R. S. & Folkman, S. (1984). Stress, appraisal and coping.
New York, NY: Springer.
Melnyk, B. M. & Fineout-Overholt, E. (2011). Evidence-based
practice in nursing & healthcare: A guide to best practice (2nd
ed.). Philadelphia, PA: Lippincott, Williams, & Wilkins.
Morrison, D. M., Hoppe, M. J., Gillmore, M. R., Kluver, C., Higa,
D., & Wells, E. A. (2009). Replicating an intervention: The
tension between fidelity and adaptation. AIDS Education and Pre-
vention, 21(2), 128–140.
Myers, S. T. (1982). The search for assumptions. Western Journal
of Nursing Research, 4(1), 91–98.
Nagel, E. (1961). The structure of science: Problems in the logic
of scientific explanation. New York, NY: Harcourt, Brace &
World.
National Human Genome Research Institute (NHGRI). (2012). An
overview of the division of intramural research. Retrieved from
http://www.genome.gov/10001634/.
National Institute of Nursing Research (NINR). (2012). Research
and funding. Retrieved from http://www.ninr.nih.gov/Research
AndFunding/.
Park, C. L., Moore, P. J., Turner, R. A., & Alder, N. E. (1997). The
roles of constructive thinking and optimism in psychological and
behavioral adjustment during pregnancy. Journal of Personality
and Social Psychology, 73(3), 584–592.
Patteson, D. & Killien, M. G. (2001). The working mother’s study:
Health promoting behaviors of employed mothers. Communicat-
ing Nursing Research, 34(9), 179.
Pickering, T. G., Miller, N. H., Ogedegbe, G., Krakoff, L. R.,
Artinian, N. T., & Goff, D. (2008). Call to action on use and
reimbursement for home blood pressure monitoring: A joint
scientific statement from the American Heart Association,
American Society of Hypertension, and Preventive Cardiovascu-
lar Nurses Association. Journal of Cardiovascular Nursing, 23(4),
299–323.
Prescott, P. A. & Soeken, K. L. (1989). Methodology corner: The
potential uses of pilot work. Nursing Research, 38(1), 60–62.
Ryan-Wenger, N. A. (2010). Evaluation of measurement precision,
accuracy, and error in biophysical data for clinical research and
practice. In C. F. Waltz, O. L. Strickland, & E. R. Lenz (Eds.),
Measurement in nursing and health research (4th ed.) (pp. 371–
383). New York, NY: Springer Publishing Company.
Santacroce, S. J., Maccarelli, L. M., & Grey, M. (2004). Methods:
Intervention fidelity. Nursing Research, 53(1), 63–66.
Scheier, M. F. & Carver, C. S. (1992). Effects of optimism on
psychological and physical well-being: Theoretical overview
and empirical update. Cognitive Therapy and Research, 16(2),
201–228.
Scheier, M. F., Carver, C. S., & Bridges, M. W. (1994). Distinguish-
ing optimism from neuroticism (trait anxiety, self-mastery, and
self-esteem): A re-evaluation of the life orientation test. Journal
of Personality and Social Psychology, 67(6), 1063–1078.
Shadish, W. R., Cook, T. D., & Campbell, D. T. (2002). Experimen-
tal and quasi-experimental designs for generalized causal infer-
ence. Chicago, IL: Rand McNally.
Smith, M. J. & Liehr, P. R. (2008). Middle range theory for nursing
(2nd ed.). New York, NY: Springer Publishing Company.
56 UNIT ONE  Introduction to Nursing Research
Tabachnick, B. G. & Fidell, L. S. (2007). Using multivariate statis-
tics (5th ed.). Boston, MA: Allyn & Bacon.
Thompson, S. K. (2002). Sampling (2nd ed.). New York, NY: John
Wiley & Sons.
Walker, L. O. (1989). Stress process among mothers of infants:
Preliminary model testing. Nursing Research, 38(1), 10–16.
Walker, S. & Hill-Polerecky, D. M. (1996). Psychometric evalua-
tion of the Health-Promoting Lifestyle Profile II. (Unpublished
manuscript.) University of Nebraska, Omaha, NE.
Waltz, C. F., Strickland, O. L., & Lenz, E. R. (2010). Measurement
in nursing and health research (4th ed.). New York, NY: Springer.
Warren, P. L. (2005). First-time mothers: Social support and confi-
dence in infant care. Journal of Advanced Nursing, 50(5),
479–488.
Wysocki, A. B. (1983). Basic versus applied research: Intrinsic and
extrinsic considerations. Western Journal of Nursing Research,
5(3), 217–224.
   http://evolve.elsevier.com/Grove/practice/
4
Introduction to Qualitative Research
Q ualitative research is a scholarly approach to
describe life experiences from the perspective
of the persons involved. It is a way to give
significance to the subjective human experience as
well as gain insights to guide nursing practice (ZuZelo,
2012). These insights are gained not through measur-
ing concepts or establishing causality but through
improving our comprehension of a phenomenon
of interest. Within a naturalistic holistic framework,
qualitative research allows us to explore the depth,
richness, and complexity inherent in the lives of
human beings. The insights from this process can
foster an understanding of patient needs and prob-
lems, guide emerging theories, and build nursing
knowledge (Munhall, 2012). Although qualitative
research is somewhat flexible, qualitative researchers
use rigorous and systematic processes that require
conceptualization, imaginative reasoning, and elegant
expression.
To critically appraise studies in publications, use
the findings in practice, and develop the skills needed
to conduct qualitative research, you must comprehend
qualitative research methodologies. Nurse researchers
conducting qualitative studies are contributing impor-
tant information to our body of knowledge unobtain-
able by quantitative means. The terminology used in
qualitative research and the methods of reasoning are
different from those of quantitative research and are
reflections of the philosophical orientations of the dif-
ferent types of qualitative research. Each qualitative
approach flows from a specific philosophical orienta-
tion that directs the methodology and interpretation of
data. This chapter presents a general overview of the
following qualitative approaches: phenomenological
research, grounded theory research, ethnographic
research, exploratory-descriptive qualitative research,
and historical research. Many other approaches to
qualitative research have been developed, but these
are the approaches and methods most frequently used
 
CHAPTER
by nurse researchers. Although each qualitative
approach is unique, there are many commonalities.
These commonalities constitute the perspective of the
qualitative researcher.
Perspective of the
Qualitative Researcher
All scientists approach problems from a philosophical
stance or perspective. The philosophical perspective
of the researcher guides the questions asked and the
methods selected for conducting a specific study. Both
quantitative and qualitative researchers have philo-
sophical perspectives. Quantitative studies are based
primarily on the philosophy of logical positivism that
values logic, empirical data, and tightly controlled
methods (see Chapter 3) (Kerlinger & Lee, 2000;
Shadish, Cook, & Campbell, 2002). Qualitative studies
are based on a wide range of philosophies, such as
phenomenology, symbolic interactionism, construc-
tivism, and hermeneutics, each of which espouses
slightly different approaches to gaining new knowl-
edge (Liamputtong & Ezzy, 2005).
Philosophy Describes a View of Science
Qualitative researchers ascribe to a view of science
that values the uniqueness of the individual and
the holistic approach to understanding human experi-
ences. The philosophical perspective of the researcher
is consistent with research questions that seek the
participant’s perspective of a phenomenon or experi-
ence. Figure 4-1 demonstrates this idea, as the arrow
on the left of the figure (“Philosophy”) shapes and fits
with the next arrow (“View of Science”). Because of
their philosophical perspectives, quantitative research-
ers hold views of science that value tightly controlled
studies and generalizable numerical findings. In con-
trast, the philosophical perspectives of qualitative
57
58 UNIT ONE  Introduction to Nursing Research
View of
Philosophy
Science
Figure 4-1  Valid science is based on congruence from philosophy to rigor.
researchers influence their views of science. As a
result, qualitative researchers value rigorous, but flex-
ible designs to identify study findings. Through the
study findings, we are able to increase our under­
standing of an experience using a discovery process
that allows a complex picture to emerge (Forman,
Creswell, Damschroder, Kowalski, & Krein, 2008).
The primary thinking process used in quantitative
studies is deduction; in contrast, qualitative research-
ers tend to be inductive thinkers (Forman et al., 2008).
Deductive thinking begins with a theory or abstract
principle that guides the selection of methods to gather
data to support or refute the theory or principle.
Forman et al. (2008) call this a “top-down” approach.
Inductive thinking involves perceptually putting
insights and pieces of information together and iden-
tifying abstract themes or working from the bottom
up. From this inductive process, meanings emerge.
Because the perception of each qualitative researcher
is unique, the meanings identified within the data may
vary from researcher to researcher. However, others
should be able to retrace the analysis and thinking
processes that resulted in the researcher’s conclusions.
In order to do this, readers must be aware of the under-
lying philosophical perspective of the study being
reviewed.
Philosophy Guides Methods
Each type of qualitative research is consistent with a
particular philosophical perspective (see Figure 4-1).
The philosophy shapes the view of science that in turn
shapes the approaches and methods selected for the
study, just as the arrows in Figure 4-1 fit together like
the pieces of a puzzle. The philosophical perspective
includes an epistemology, a view of knowing and
knowledge (Munhall, 2012). As a result, the philoso-
phy directs the research questions and the collection
and interpretation of the data. Creswell (2009) empha-
sizes this point by stating that the assumptions of the
specific philosophical approach cannot be separated
from the methods. For example, researchers who hold
to a phenomenological philosophy of science will ask
research questions about the “lived experiences” of
persons. They will analyze the data in such a way as
to develop a rich, deep description of an experience
Approach
Criteria of
and
Rigor
Methods
from the perspective of the participants. Although the
researcher does not always clearly state the philo-
sophical stance on which the study is based, a knowl-
edgeable reader can recognize the philosophy through
the description of the problem, formulation of the
research questions, and selection of the methods to
address the research questions. A well-designed quali-
tative study is congruent at each stage with the under-
lying philosophical perspective as identified by the
researcher.
Qualitative researchers use open-ended methods to
gather descriptions of health-related experiences from
participants (Fawcett & Garity, 2009). These open-
ended methods include interviews, focus groups,
observation, and document review (Speziale & Car-
penter, 2007). When oral methods are used, the
researcher will usually capture the interaction by an
audio or video recording so that a transcript of the
communication can be prepared for analysis. These
methods are discussed in detail in Chapter 12.
Philosophy Guides Criteria of Rigor
Scientific rigor is valued because it is associated with
the worth of the research outcomes. The rigor of quali-
tative studies is appraised differently from the rigor of
quantitative studies because of the differences in the
underlying philosophical perspectives. Quantitative
studies are considered rigorous when the procedures
for the study are prescribed prior to data collection,
the sample is large enough to represent the population,
and researchers tightly control the collection and anal-
ysis of the data. A quantitative researcher expects that
another researcher could replicate or repeat the study
with a similar sample and receive similar results. This
is desirable because quantitative researchers define
rigor to include objectivity and generalizability. Rigor-
ous qualitative researchers, however, are characterized
by openness and dem­onstrate methodological congru-
ence, scrupulous adherence to a philosophical perspec-
tive, thoroughness in collecting data, consideration
of all the data in the analysis process, and self-
understanding. The researcher’s self-understanding is
important because qualitative research is an interactive
process shaped by the researcher’s personal history,
biography, gender, social class, race, and ethnicity as
 CHAPTER 4  Introduction to Qualitative Research 59
well as by those of the study participants (Creswell,
2009; Marshall & Rossman, 2011). Self-understanding
allows the researcher to have insight into his or her
potential biases related to the phenomenon of interest
and prevent these biases from interfering with the
voices of the participants being heard. These charac-
teristics of qualitative researchers are essential to
produce a valid study. Critical appraisal of the rigor of
qualitative studies is discussed in more detail in
Chapter 18.
To reinforce the key points of Figure 4-1 that phi-
losophy shapes one’s view of science, the methods,
and criteria of rigor, a study of families with asthmatic
children are presented as an example.
Meah, Callery, Milnes, and Rogers (2009) explored
how parents and preadolescent children shared the
responsibility for asthma management. The research-
ers had noted a gap in knowledge about the transfer
of responsibility from parents to children. The purpose
of the study was to “examine meaning of responsibil-
ity in children’s lives and how parents and children
negotiate these responsibilities over time” (Meah
et al., 2009, p. 1953). An exploratory-descriptive qual-
itative study was conducted because quantitative
studies had produced diverse results that did not
provide adequate understanding of the transfer of
responsibility. The study included a framework of
responsibility, and the researchers designed the study
to be consistent with feminist philosophy and socio-
logical perspectives:
“Interpretation of the data drew upon both feminist
epistemology and sociological understandings of chil-
dren, health, and the body which relocate subjective
experience at the heart of scientific inquiry.” (Meah
et al., 2009, p. 1954)
Consistent with sociological perspectives, the
familial unit was the focus, instead of only the child
or only the parents. The researchers’ methods included
conducting “open-ended, conversational-style inter-
views” with 18 children and their parents (Meah
et al., 2009, p. 1954). The researchers also inter-
viewed the parents separately. The recordings of the
interviews were transcribed and “descriptive codes
were developed from the transcripts” (p. 1954). The
researchers briefly described how they displayed
excerpts from the transcripts on charts that “enabled
comparison of the children’s and adults’ perceptions
while simultaneously facilitating comparison within
children’s accounts and those of their parents”
(p. 1954). The study results revealed the complexity
of allowing children to manage their asthma, as they
grow older:
“For many of the children in our study, their ability to
take responsibility for their condition was often
dependent on what they were both allowed and
encouraged to do by their parents/carers. Clearly
then, we must understand this as a negotiated process
and to appreciate its full complexity, we must also
look at how parents approached the sharing of
responsibility regarding their children’s asthma.”
(Meah et al., 2009, p. 1955)
The researchers concluded that their findings were
consistent with the theoretical framework of responsi-
bility that they introduced at the beginning of the study
report. The implementation of the study and its find-
ings were also consistent with the philosophical per-
spectives that the researchers explicitly identified. The
approach and methods allowed for children and
parents to share their thoughts and feelings in a non-
threatening setting. The data analysis involved strate-
gies that allowed the perspectives of the participants
to be described. The researchers could have strength-
ened the report by providing details about how rigor
was ensured. Additional information about the rigor
of the methods would increase the validity of the
findings.
This example confirms that philosophy shapes
one’s view of science, which in turn shapes the
methods used in a study and the criteria by which the
rigor of the study will be evaluated (see Figure 4-1).
Because qualitative studies emerge from several phi-
losophies, an understanding of different approaches to
qualitative research is needed as a foundation for
appraising the rigor of research and making appropri-
ate application of the findings.
Approaches to
Qualitative Research
Five approaches to qualitative research commonly
conducted and published in the nursing literature are:
phenomenological research, grounded theory research,
ethnographic research, exploratory-descriptive quali-
tative research, and historical research (see Table 4-1).
Although the five approaches have the commonalities
already discussed, these approaches are different, in
great part because researchers in different disciplines
60 UNIT ONE  Introduction to Nursing Research
TABLE 4-1  Philosophical Orientations
Supporting Qualitative Approaches
to Nursing Research
Philosophical and
Theoretical Orientations Qualitative Approach
Phenomenology Phenomenological research
Symbolic interaction theory Grounded theory research
Naturalism and ethical Ethnographic research
principles
Interpretive and naturalistic Exploratory-descriptive
perspectives; pragmatism qualitative research
Historicism, pragmatism Historical research
developed them. For example, the social and psycho-
logical scientists developed the approaches known as
phenomenological research and grounded theory
research (Giorgi, 2010; Reed, 2010). Anthropologists
developed ethnography with its focus on culture
(Morse & Richards, 2002). Exploratory-descriptive
qualitative research has emerged from the disciplines
of nursing and medicine and has a pragmatic focus on
using the knowledge gained to benefit patients and
families and improve health outcomes. Historians
developed methods to analyze source documents, arti-
facts, and interviews of witnesses to summarize the
knowledge gained by studying the past (Lewenson &
Herrmann, 2008). Nurse researchers adapted these
methods to understand changes within nursing and
health care. The common purpose among the methods,
however, is to examine the meaning of human experi-
ences from the perspective of the “knower,” the person
(or persons) to whom the experiences belong (Bassett,
2004).
Although the data are gathered with openness to
the individual experiences of study participants, this
fact does not mean that the interpretation is value-free.
Each approach is based on a philosophical orientation
that influences the study design from the wording of
the research question through the interpretation of the
data (see Figure 4-1). Thus, you as a consumer of
qualitative research must be aware of the guiding prin-
ciples of the philosophical perspective of a study and
use these principles to critically appraise a qualitative
study. The discussion of each approach will cover its
philosophical perspective or orientation, methodol-
ogy, and contribution to nursing knowledge.
Phenomenological Research
Phenomenology is both a philosophy and a research
method. The purpose of phenomenological research
is to describe experiences (or phenomena) as they are
lived-in phenomenological terms, to capture the “lived
experience” of study participants (Fawcett & Garity,
2009; Munhall, 2012). The philosophical positions
taken by phenomenological researchers are very dif-
ferent from those common in the culture and research
traditions of the nursing field, which value large
samples and generalizable findings.
Philosophical Orientation
Phenomenologists view the person as integral with the
environment. The world is shaped by the self and
shapes the self. At this point, however, phenomenolo-
gists diverge in their beliefs according to adherence to
a particular phenomenological philosopher. The key
philosophers used by nurse researchers adhering to
phenomenology are Husserl and Heidegger.
Philosopher Edmund Husserl (1859-1938) is con-
sidered the father of modern phenomenology (Phillips-
Pula, Strunk, & Pickler, 2011). Husserl wrote Logical
Investigations (1901/1970), in which he developed his
ideas about phenomena in an effort to resolve the
conflict in thought between human sciences (primarily
psychology) and the basic sciences (such as physics).
Phenomena make up the world of experience. These
experiences cannot be explained by examining causal
relations but need to be studied as the very things they
are. A phenomenon occurs only when there is a person
who experiences it. Thus, the experience must be
described, not studied using statistics or the research-
er’s preconceived ideas. To describe the experience,
the researcher must be open to the participant’s world-
view, set aside personal perspectives, and allow mean-
ings to emerge. Husserlian phenomenologists believe
that although self and world are mutually shaping, it
is possible to separate oneself from one’s beliefs or set
aside one’s beliefs to see the world firsthand in a naive
way (Dowling, 2007). Setting aside one’s beliefs
during qualitative research is called “bracketing.”
Researchers who follow the philosophy proposed
by Heidegger do not agree, taking the position
that bracketing is not possible. Martin Heidegger
(1889-1976) was a student of Husserl but expanded
the goal of phenomenology from description of lived
experience to the interpretation of lived experiences
(Earle, 2010). Heidegger’s seminal work was Being
and Time (1927/1962). Heideggerian phenomenolo-
gists believe that the person is a self within a body, or
is embodied (Munhall, 2012). Munhall describes
embodiment as “the unity of body and mind” that
eliminates the “the idea of a subjective and objective
world” (p. 127). She goes on to say, “the world is
knowable only through the subjectivity of being in the
world” (p. 127). For example, the interaction between
the sensory input that your body sends to your brain
 CHAPTER 4  Introduction to Qualitative Research 61
and your brain’s thinking processes determines your
perceptions of the world in which you exist. A
depressed person may not notice the color of sunset
or the joy in a child’s laughter. Another person may
not be able to see the sunset because of blindness but
may find special meaning in the nuances of the child’s
laughter. In each situation, the person’s being in the
world is shaped by the unity of body and mind.
Heideggerian phenomenologists posit that the
person is situated in specific context and time that
shapes his or her experiences, paradoxically freeing
and constraining the person’s ability to establish
meanings through language, culture, history, purposes,
and values (Munhall, 2012). For example, a 50-year-
old man diagnosed with aggressive cancer in the
United States would experience the diagnosis differ-
ently from a 20-year-old woman in rural Ethiopia
receiving the same diagnosis. The body, the world,
and the concerns, unique to each person, are the
context within which that person can be understood.
Heideggerians believe that the person experiences
being within the framework of time, also called being-
in-time. The contexts of the man and the woman with
cancer would have different meanings, depending on
whether they lived during the 1960s or the 2010s. The
availability of treatment, financial resources, and
gender roles are only a few of the factors that would
shape the cancer experience for these individuals.
Each of them has only situated freedom, not total
freedom. The man has the freedom to choose physi-
cians from among those who will accept his insurance.
The woman may have the freedom only to choose
whether she will use a traditional healer or not seek
treatment at all. Until a disruption such as an expected
diagnosis of cancer occurs, the person may not have
considered the limits on meaning imposed by the
context and the time.
Other philosophers have built on Husserl and Hei-
degger’s perspectives and refined phenomenological
methods. Merleau-Ponty (1908-1961) was among the
French philosophers who further developed the Hei-
deggerian concepts of being-in-time and embodiment
(Munhall, 2012). He wrote The Phenomenology of
Perception (1945/2002). Colaizzi (1973), Giorgi
(1985), and van Manen (1990) proposed research
guidelines for phenomenological research focused on
procedural interpretations of the phenomenological
method (Speziale & Carpenter, 2007). The novice
nurse researcher interested in phenomenology should
expand his or her knowledge in this area through
immersion in the original writings of these philoso-
phers (Munhall, 2012).
All phenomenologists agree that there is not a
single reality; each individual has his or her own
reality. Reality is considered subjective, and as a
result, unique to the individual. The researcher’s expe-
riences in collecting data and analyzing the data during
a study are also unique. The researcher must invest
considerable time exploring the various philosophical
stances within phenomenology to select one compat-
ible with his or her perspective and the research ques-
tion being asked. More detail on the conduct of
phenomenological research is provided in Chapter 12.
Phenomenology’s Contribution to Nursing Science
Phenomenology is the philosophical base for three
nursing theories and studies that have explored these
theories. These theories include the theory of human-
istic nursing (Paterson & Zderad, 1976), the theory of
human becoming (Parse, 1981, 1992), and the theory
of caring (Watson, 1999). Paterson and Zderad empha-
sized in their theory of humanistic nursing that phe-
nomenology shapes the nurse and the patient as they
share experiences in the context of health care. The
theory is applicable to research because the researcher
meets the participant with openness and respect for
uniqueness.
On the basis of the values of Paterson and Zderad’s
(1976) theory, Lesniak (2010) conducted a study with
adolescent females who deliberately injure themselves
by cutting. She interviewed six young women who
described their experiences, including what they
would like emergency nurses to know about self-
cutting. From their interviews, the researchers identi-
fied “an element of loneliness,” “an overall meaning
of angst and desperation,” and “a recovery process
after the cutting is over” (Lesniak, 2010, p. 146).
Emergency nurses adhering to Paterson and Zderad’s
caring values who are alert to signs of self-cutting will
assess the adolescent and identify strategies to “protect
the adolescent from future self-injury” (p. 146).
Lesniak noted that the sample was limited to white,
middle-class adolescents and could not be generalized
to adolescents of other races/ethnicities or income
levels.
In 1981, Parse described her theory of man-living-
health as evolving from the foundation of existential
phenomenology. As she conducted studies congruent
with the theory, she refined her ideas into the human
becoming theory (Parse, 1992). According to her
theory, human beings co-create reality with the envi-
ronment and structure meaning through images,
words, and actions. Parse proposes research methods
consistent with phenomenology, the philosophical
foundation for her theory (Parse, 2001, 2011). Using
Parse’s research methods, Chen (2010) conducted a
study of the lived experience of moving forward
among persons who had experienced a spinal cord
62 UNIT ONE  Introduction to Nursing Research
injury. Chen collected data from 15 persons through
“dialogical engagement, in which the true presence of
the researcher was with the participant” (p. 1134).
From the stories of the participants and the recordings
of the research interactions, Chen extracted three core
concepts that she reframed within Parse’s theory:
“confronting difficulties,” “going on and finding self-
value and confidence,” and “co-creating successes
amid opportunities and restrictions” (pp. 1137-1138).
The use of Parse’s theory during the interpretation of
the data was consistent with her research approach but
somewhat of a departure from remaining open to the
meanings that emerge from qualitative data.
Watson’s (1999) theory of caring is also congruent
with phenomenology. She describes values of nurses
that produce caring actions and an intersubjective
experience shared by the nurse and patient. She pro-
poses 10 carative factors that provide a structure for
caring as the core of nursing. Transpersonal caring
relationships and the carative factors “potentiate thera-
peutic healing processes and relationships: they affect
the one caring and the one-being-cared-for” (Watson,
1997). Byers and France (2008) conducted a phenom-
enological study using Watson’s theory with nurses
who care for patients with dementia hospitalized on
medical-surgical units. From their interviews with
nine nurses, they extracted a “synthesis of unity,” a
sentence that captured the experience, which they
stated as “They stay with you: they come home with
you every day” (p. 47). High nurse-patient ratios and
the needs of dementia patients challenged the nurses
and resulted in stress. Because of their commitment to
caring, the nurses were frustrated by giving what they
perceived to be care that did not meet their personal
standards. Byers and France conclude by noting that
“the essence of caring is revealed as it is manifested
in the RN caring for the patients with dementia yet it
is also the RN who needs to be cared for” (p. 48). An
appraisal can be made of this study similar to that of
the Chen (2010) study. The use of a theory within
a phenomenological approach has the potential to
interfere with the emergence of the participants’
perspectives.
Shorter and Stayt (2010) conducted a phenomeno-
logical study of the experiences of grief and coping
mechanisms of nurses in an adult intensive care unit.
The philosophical perspective, the methodology, and
the contributions to nursing knowledge of this study
are presented as an example of a phenomenological
study. Shorter and Stayt designed their study to be
consistent with the Heideggerian phenomenological
perspective (Johnson, 2000) and to “explore critical
care nurses’ experiences of grief and their coping
mechanisms when a patient dies” (p.160). The
researchers’ methodology included interviewing eight
nurses and analyzing the narratives of the interviews
according to Colaizzi’s guidelines, a specific approach
to phenomenological data analysis (Dowling, 2007).
The researchers found clusters of data around two
phases in the experience—“the death experience” and
“the death thereafter.” In describing the death experi-
ence, one nurse commented “I hate it when the patient
dies suddenly—everything seems just so disorganized.
… It makes me feel awful when that happens.… I like
to feel that I have done everything that I can” (p. 162).
Another nurse talking about the death thereafter said,
“You do have to keep a distance in a way, because
death happens so often on our unit. You’d be an emo-
tional wreck if you let it bother you or affect you all
the time” (Shorter & Stayt, 2010, p. 164).
The death experience was described with the
phrases “expectedness, control and good care,” “strik-
ing a chord,” and “meaningful engagement” (Shorter
& Stayt, 2010, pp. 162-163). The phrases “formal
support,” “informal support,” “normalization of
death,” and “emotional dissociation” (pp. 163-164)
described the death thereafter. Shorter and Stayt con-
cluded that the grief and coping strategies of nurses
are complex and may vary by type of care setting.
Shorter and Stayt (2010) clearly identified the phil-
osophical perspective and methods of their study,
although they did not cite primary sources for either
Heidegger or Colaizzi. Consistent with the Heideg-
gerian perspective, they collected and analyzed data
simultaneously. They used Colaizzi’s framework to
cluster common meanings and identify themes from
the transcripts of the interviews they had conducted.
Shorter and Stayt explicitly noted adherence to the
Heidegerrian tenets when they discussed the rigor of
the study to include “co-construction of knowledge
between researcher and participant” (p. 161). The
researchers proposed that additional studies are needed
to explore whether the coping strategies of nurses
affect the quality of care they provide.
Grounded Theory Research
Grounded theory research is an inductive research
technique developed by Glaser and Strauss (1967)
through their study of the experience of dying. The
method’s name means that the findings are grounded
in the concrete world as experienced by the partici-
pants and are interpreted at a more abstract theoretical
level. The desired outcome of grounded theory studies
is a middle range or substantive theory (Fawcett &
Garity, 2009; Marshall & Rossman, 2011; Munhall,
2012).
 CHAPTER 4  Introduction to Qualitative Research 63
Philosophical Orientation
Grounded theory is congruent with symbolic interac-
tion theory, which holds many views in common with
phenomenology. George Herbert Mead (1863-1931),
a social psychologist, developed the principles of
interaction theory that were posthumously published
(Mead, 1934). His principles were shaped and refined
by other social psychologists and became known as
symbolic interaction theory (Crossley, 2010). Sym-
bolic interaction theory explores how perceptions of
interactions with others shape one’s view of self and
subsequent interactions. One’s view of self influences
subsequent interactions and creates meaning. People
create reality by attaching meanings to situations.
Meaning is expressed in terms of symbols, such as
words, religious objects, and clothing. These symbolic
meanings are the basis for actions and interactions.
Symbolic meanings are different for each indi-
vidual. We cannot completely know the symbolic
meanings of another individual. In social life, groups
share meanings. They communicate these shared
meanings to others through socialization processes.
Group life is based on consensus and shared mean-
ings. Interaction among people may lead to redefini-
tion of experiences, new meanings, and possibly a
redefinition of self. Because of its theoretical impor-
tance, the interactions among the person and other
individuals in social contexts are the focus of observa-
tion in grounded theory research.
Grounded Theory’s Contribution to Nursing Science
Grounded theory researchers have contributed to our
understanding of the patient experience across a wide
range of settings. Giske and Artinian (2008) helped us
understand how gastroenterology patients undergoing
diagnostic testing were balancing between hope and
despair. Balancing was the primary theoretical code in
the resulting theory of preparative waiting. On similar
themes of waiting and balance, Trimm and Sanford
(2010) described the family experience of waiting
during surgery as maintaining balance during the wait.
In cardiac care settings, researchers have described the
inner strength of women following placement of a
coronary artery stent (Mendes, Roux, & Ridosh, 2010)
and the “dialogue around maintaining and renegotiat-
ing normality” following a coronary artery bypass
graft procedure (Banner, 2010, p. 3123). Noiseux and
Ricard (2008) conducted a grounded theory study to
develop a theory describing the experience of recov-
ery from schizophrenia from the perspectives of
patients, their families, and healthcare professionals.
Maliski, Connor, Williams, and Litwin (2010) con-
ducted a grounded theory study to understand the
experiences of low-income, uninsured African Ameri-
can men diagnosed with prostate cancer. These
researchers noted that African American men were
disproportionately affected by prostate cancer and
often were faced with a “death sentence” due to an
unwelcomed and unexpected diagnosis of cancer. The
methodology of the study was secondary analysis of
transcripts from a larger study (Maliski, Rivera,
Connor, Lopez, & Litwin, 2008). The sample for the
original study included 60 Latina and 35 African
American men who had undergone treatment for pros-
tate cancer. When the transcripts were analyzed for the
original study, the researchers noted that spiritual con-
cepts emerged in the interviews with some of the
African American men. Thus, Maliski et al. (2010, p.
471) analyzed the transcripts from in-depth interviews
with 18 African American men from the original
study, “focusing on the role of faith in coping with
their prostate cancer diagnosis, treatment, and adverse
effects.” The researchers did not specify the philo-
sophical perspective upon which the study was devel-
oped; however, they did indicate that they used
grounded theory analytic strategies to analyze the
data.
Listen to the voice of one participant as he used his
faith to conquer his fear. “As long as my heart and
spirit are good, I can keep going. Good things can
happen in the future, and I was put on Earth for a
reason so I’m not afraid of death” (Maliski et al.,
2010, p. 474).
As you can see, grounded theory research examines
experiences and processes with a breadth and depth
not usually possible with quantitative research. The
reader can intuitively verify these findings through her
or his own experiences. Through the rigorous conduct
of grounded theory research, these investigators were
able to conclude that faith was a spiritual coping strat-
egy for those men of limited resources (Maliski et al.,
2010). The contribution to nursing science is that a
clear, cohesive description of the phenomenon allows
greater understanding. Although a theory was not
developed, the improved understanding can guide
nursing interventions that meet the needs of patients
in ways that they value.
Ethnographic Research
Ethnographic research provides a framework for
studying cultures. The word ethnography is derived by
combining the Greek roots ethno (folk or people) and
graphy (picture or portrait). Ethnographies are the
written reports of a culture from the perspective of
insiders. These reports were initially the products of
anthropologists who studied primitive, foreign, or
64 UNIT ONE  Introduction to Nursing Research

remote cultures. Now, however, a number of other TABLE 4-2  Four Types of Ethnography
disciplines, including social psychology, sociology,
political science, education, and nursing, promote cul- Other
tural research (Wolf, 2012). Type Labels Purpose
Ethnography does not require travel to another Classic Traditional Describe a foreign culture
country or region; however, it requires spending con- through immersion in the
culture for an extended
siderable time in the setting observing and gathering
period
data. A specific group or subculture is identified for Systematic Institutional Describe the social
study, such as women giving birth at home in Haiti or organizational structure
male nurses working in acute care settings. Ethnogra- influencing a specific
phy can be used to describe and analyze aspects of the group of people
ways of life of a particular culture, even your own. Interpretative Interpret the values and
McGibbon, Peter, and Gallop (2010) conducted an attitudes shaping the
ethnographic study with pediatric intensive care nurses behaviors of members of a
focusing on the social organization and the institu- specific group, in order to
tional context that contributed to the stress of the promote understanding of
the context of culture
nurses. They identified “six main forms of nurses’
Critical Disrupted Examine the life of a group
stress, including emotional distress; constancy of pres- in the context of an
ence; burden of responsibility; negotiating hierarchi- alternative theory or
cal power; engaging in bodily caring; and being philosophy, such as
mothers, daughters, aunts, and sisters” (p. 1357). feminism or constructivism
An appreciation of other cultures often begins with
an examination of your own culture and identification
of ethnocentric values that may influence your care of explore “the meanings of social actions within cul-
people of other cultures. The formal study of one’s tures” (Wolf, 2012, p. 285).
own culture or social context, autoethnography, Four schools of thought within ethnography have
involves critical reflection on your own life experi- emerged from different philosophical perspectives:
ences as shaped by your culture (Wolf, 2012). For classic, systematic, interpretive, and critical ethnogra-
example, Brooks (2011) conducted an autoethnogra- phy (Speziale & Carpenter, 2007). Classic ethnogra-
phy of the mental illness of obsessive-compulsive dis- phy seeks to provide a comprehensive holistic
order from her unique viewpoints as “an academic description of a culture (Wolf, 2012). For example,
observer and an individual sufferer” (p. 251). Brooks, researchers who conduct classic ethnographic studies
not a nurse, shared her autoethnography in part so live for extended periods outside of their own country
nurses and other healthcare professionals could under- in the environment of their study and write a factual
stand the ability of some obsessive-compulsive disor- description of the culture (see Table 4-2). In contrast,
der sufferers to appear as though nothing is wrong. In systematic ethnography explores and describes the
her words, by “performing ‘appropriately’ and putting structures of the culture with an increased focus on
on my socially acceptable ‘face,’ ” she “suppressed the groups, patterns of social interaction, organizations,
personal and tragic reality of my illness” (p. 259). Her and institutions. The study by McGibbon et al. (2010,
description of her experience reflected the perfor- p. 1356), described earlier, is an example of a system-
mances required by her roles within the cultures of atic ethnography as evidenced by exploring “nurses’
academia and mental health treatment. everyday work with institutional structures that shape
practice.”
Philosophical Orientation Interpretive ethnography has as its goal under-
Anthropologists seek to understand people: their ways standing the values and thinking that result in behav-
of living, believing, acquiring information, transform- iors and symbols of the people being studied.
ing knowledge, and socializing the next generation. Alexander (2003) identified his study of the black
Studying a culture begins with the philosophical barbershop as an interpretive ethnography. Through
values of respecting, appreciating, and seeking to pre- observation, repeated interaction, and recall of child-
serve the values and ways of life of the culture (Wolf, hood experiences, Alexander clarified the values of
2012). The philosophical bases of ethnography are socializing young black men and creating community
naturalism and respect for others. The purpose of within a physical space. Critical ethnography has a
anthropological research is to describe a culture and political purpose of relieving oppression and
 CHAPTER 4  Introduction to Qualitative Research 65
empowering a group of people to take action on their
own behalf. Wolf (2012) calls this type of ethnography
“disrupted” or “disruptive,” and identifies its philo-
sophical foundation to be critical social theory
(Boutain, 1999; Fontana, 2004). Gardezi and col-
leagues (2009) conducted a critical ethnography of
interprofessional communication in the operating
suite and analyzed data from 700 surgical procedures
performed from 2005 through 2007. An unanticipated
finding was the importance of silence, especially those
silences that emerged during the analysis of episodes
of ineffective communication. Consistent with critical
ethnography, the researchers noted that policies and
procedures are needed that encourage team members
to speak up to protect the patients, but they also
acknowledged the complex interaction of silence and
speech in the operating room.
Ethnography’s Contribution to Nursing Science
Madeline Leininger (1970) brought ethnography into
nursing science by writing the first book linking
nursing with anthropology. Leininger was first a nurse
and then earned her doctoral degree in anthropology.
In the 1950s, she began developing a framework for
culture care that became the Sunrise Model (Clarke,
McFarland, Andrews, & Leininger, 2009). The Sunrise
Model identifies factors that affect health and illness,
such as religion, income, kinship, education, values,
and beliefs. Chapter 7 contains more information
about the Theory of Culture Care developed by
Leininger, so this section focuses on the qualitative
method that she developed to be consistent with
ethnonursing.
Ethnonursing research values the unique per-
spective of groups of people within their cultural
context that is influenced at the macro-level by
geographical location, political system, and social
structures (see Table 4-1). Multiple levels of factors
affect the culture and, consequently, the care expres-
sions of the people. A Vietnamese family who is the
only family in a small rural community in Georgia
may have different care practices from those who live
in New York City in a predominantly Vietnamese
community. Leininger developed “enablers,” sets of
questions to guide the researcher’s study of the culture
(Leininger, 1997; 2002). The enablers provide a flex-
ible framework for the researcher to use to collect and
analyze the qualitative data. For example, one of the
enablers is “Leininger’s Observation-Participation-
Reflection Enabler” (Leininger, 1997, p. 45), which
reminds the researcher to use these three processes
during a study. The method is naturalistic, meaning
that the research is conducted in a natural setting
without any attempt to control or alter the context. The
researcher can be open to explore the insider perspec-
tive on health and well-being. The primary ethnonurs-
ing data collection method is participant-observation
(Douglas et al., 2010). Participant-observation is
defined as being present and interacting with partici-
pants in routine activities. At the same time, the
researcher is paying attention to what is happening
from the perspective of an outsider.
Schumacher (2010) conducted an ethnonursing
study in a rural village in the Dominican Republic. She
used Leininger’s Observation-Participation-Reflection
(OPR) Enabler (1997) to guide 28 days of data collec-
tion related to the different aspects of the Sunrise
Model, such as technological, cultural, and religious
factors. In addition to observing patterns of behavior
related to health, she interviewed 29 people to learn
more about the meaning of the observations. These
people are called informants. The raw data were orga-
nized into 12 categories consistent with the model.
From the categories, recurring patterns emerged that
eventually led to three themes presented in the follow-
ing quotations from the study:
One informant said, “Family is everything. It is the
most important and central thing in this culture.…
Life is lived for the family.… Family is the basis of our
society …” (Schumacher, 2010, p. 98). This quotation
was used as an exemplar of the first theme that family
presence is necessary for cultural care of the com-
munity members. The second theme was caring as
giving respect to the person and paying attention.
“Respect and attention are central to the meaning of
care for rural Dominicans” (p. 98). Schumacher noted
that the people “both value and use generic (folk)
care practices as well as professional care practices”
(p. 99). This third theme was supported by interview
data about using home remedies and faith healers and
seeking treatment by a healthcare professional.
Congruent with Leininger’s theory and ethnonurs-
ing method, Schumacher (2010, p. 99) discussed
the findings using the theory’s action modes
of “culture care preservation/maintenance, culture
care accommodation/negotiation, and culture care
repatterning/restructuring.” The methods were imple-
mented with rigor and produced quality findings.
Although this study was conducted in the Dominican
Republic, the insights into these health values could
promote culturally appropriate care for immigrants
from that country in the United States.
66 UNIT ONE  Introduction to Nursing Research
Exploratory-Descriptive Qualitative Research
Qualitative nurse researchers have conducted studies
with the purpose of exploring and describing a topic
of interest but, at times, have not identified a specific
qualitative methodology. Other researchers have iden-
tified their studies as grounded theory but have not
identified theoretical elements and relationships in
their findings. Some have identified their studies as
phenomenology but have not provided a thick descrip-
tion of a lived experience in their findings. Qualitative
descriptive research is a legitimate method of research
that may be the appropriate “label” for studies that
have no clearly specified method or in which the
method was specified but that ended with “a compre-
hensive summary of an event in the everyday terms of
these events” (Sandelowski, 2000, p. 336). Labeling a
study as a specific type (grounded theory, phenome-
nology, or ethnographic) implies fixed categories of
research with distinct boundaries, but the boundaries
between methods are more appropriately viewed as
permeable (Sandelowski, 2010). Although the studies
result in descriptions and could be labeled as descrip-
tive qualitative studies, most of the researchers are in
the exploratory stage of studying the subject of inter-
est. To decrease any confusion between quantitative
descriptive studies and the discussion of this qualita-
tive approach, we decided to call this approach
exploratory-descriptive qualitative research. In this
book, studies that the researchers identified as being
qualitative without indicating a specific approach like
phenomenology or grounded theory will be labeled as
being exploratory-descriptive qualitative research.
Exploratory-descriptive qualitative studies are con-
ducted to address an issue or problem in need of a
solution. For example, Rohr, Adams, and Young
(2010) recognized that nurses used inconsistent
approaches to prevent and relieve oral discomfort of
terminally ill patients. These researchers used inter-
views in their exploratory-descriptive qualitative
study to gain the perspective of the patient receiving
palliative care so that standards for oral care could be
established for this unique population. They concluded
that oral care must begin with structured assessment
of patients’ mouths every day and include recognition
of the impact that oral discomfort has on the emotions
and social life of patients receiving palliative care.
Philosophical Orientation
The philosophical orientation that supports
exploratory-descriptive qualitative studies undergirds
most methods of qualitative inquiry. In contrast to the
received view of reality that is the foundation for
quantitative methods, qualitative researchers ascribe
to a perceived view of reality. The perceiver—the
person living the experience—is the source of infor-
mation. Closely aligned with the perceived view is a
general approach to science that Liamputtong and
Ezzy (2005) call the “interpretive orientation.” The
interpretive orientation acknowledges that meaning is
created and maintained in context. Other qualitative
experts call the general qualitative approach “natural-
istic inquiry.” Naturalistic inquiry encompasses
studies designed to study people and situations in their
natural states (Sandelowski, 2000). Another broad
philosophical orientation often inferred by exploratory-
descriptive qualitative researchers is pragmatism.
Pragmatism supports studies designed to gather data
for transformation into information needed to solve a
problem or offer a new strategy. “It [pragmatism] indi-
cates that there is a goal, that what works is defined
in practice and thus must be put into practice”
(McCready, 2010, p. 192). Pragmatism was the philo-
sophical orientation of the study conducted by Rohr
et al. (2010).
Exploratory-Descriptive Qualitative Research’s
Contribution to Nursing Science
An exploratory-descriptive qualitative researcher often
indicates that a study is needed with a specific popula-
tion to understand the needs of, desired outcomes of,
or views on appropriate interventions held by the
members of the group. The goal is to create a program
or an intervention to benefit the population. Exploratory-
descriptive qualitative researchers identify a specific
lack of knowledge that can be addressed only through
seeking the viewpoints of the people most affected.
After citing numerous quantitative studies on the topic,
Swanlund (2010) wrote about the need for this
approach in her study of medication adherence of the
older cardiovascular patient:
“Additionally, the research questions for these studies
were derived from the healthcare provider’s per-
spective. These studies have left many questions
unanswered regarding the reasons for nonadherence
from the perspective of the older adult. Information
is needed that will help community-dwelling older
adults be successful with their medication manage-
ment.” (Swanlund, 2010, p. 23)
Researchers who value the perspectives of
participants may begin a program of research with
qualitative methods to: (1) begin development of
inter­ventions, (2) evaluate the appropriateness of an
intervention following implementation, or (3) develop
participants’ definitions of concepts that researchers
 CHAPTER 4  Introduction to Qualitative Research 67
would like to measure. For example, qualitative data
were collected during a federally funded study of
stress management interventions with healthy adults
and persons living with HIV infection (Tuck &
Thinganjana, 2007). The researchers sought to under-
stand the participants’ perspectives on spirituality as a
first step in developing an operational definition for
measurement of spirituality. Qualitative data have
also been collected through interviews or focus groups
to evaluate the cultural appropriateness of and partici-
pants’ satisfaction with an intervention at the conclu-
sion of a study. For example, Vincent (2009) conducted
a focus group with Mexican-Americans with diabetes
who had been the recipients of a tailored intervention
for health behavior change. She found that the inter-
vention group participants were satisfied with the
intervention.
Health promotion within chronic obstructive pul-
monary disease (COPD) was the focus of a qualitative
study conducted by Caress, Luker, and Chalmers
(2010). The philosophical orientation, methodology,
and results of the study are discussed as an example
of the exploratory-descriptive qualitative approach.
Caress et al. described their study as an “exploratory,
descriptive design … using semistructured interviews
with patients and family caregivers” (p. 565). They did
not specify the philosophical orientation; however,
they noted that all interviews took place in the partici-
pants’ homes, an approach consistent with a natural-
istic orientation. In addition, the aim of the study,
provided in the following quotations and summary,
was consistent with a pragmatic philosophical
perspective:
“The aim of this study was to generate in-depth
insights into patients’ and family members’ under-
standing of the causation, progression, and preven-
tion of COPD and the role of health promotion with
this population” (Caress et al., 2010, p. 565). The
interviews were digitally recorded and transcribed.
Data analysis included measures to “ensure the con-
gruence of the analysis amongst the investigators”
such as “independent review and coding, discussion
of the emerging themes, identification of the key
themes, re-examination of the full data set to ensure
the ‘fit’ of the codes to the data, and finally discussion
of any discrepancies with a goal of achieving consen-
sus” (Caress et al., 2010, p. 566). Additional measures
that they used to enhance the rigor of the study and
to protect the ethical rights of the participants were
described in detail in this study.
From the transcribed interviews of 14 patients and
12 family members, Caress et al. (2010) identified
three themes. The first theme, “Health promotion:
What’s that?” emerged from the limited spontaneous
information about health promoting behaviors pro-
vided by the participants. The behaviors that were
mentioned were more preventive in nature, rather
than health promoting. The second theme, “Commu-
nity resources for health promotion” (Caress et al., p.
569), was derived from participant statements that
indicated limited knowledge of community resources,
with the exception of pulmonary rehabilitation. The
theme “It wasn’t just the smoking: Patients’ and
family members’ views on the causation of COPD”
(p. 569) encompassed several participants’ statements
about the role tobacco played in the development of
the disease. The researchers concluded the study’s
report by proposing a new approach to promoting
health among the patients with COPD and their
families:
“Our data suggest that a more wide-ranging approach,
encompassing aspects of health promotion, might
be welcomed by many patients and their family
carers.… The findings from this study highlight gaps
in patients’ and carers’ understanding of the potential
role of health promotion in COPD and areas of inter-
vention for health professionals.” (Caress et al., 2010,
pp. 571-772)
This study is a typical example of exploratory-
descriptive qualitative research, in that the researchers
reported the data without transforming the data into
more abstract concepts or constructs. The study would
have been strengthened by a clear discussion of the
philosophical perspective of the study, which seems to
be a mixture of naturalistic and pragmatic perspec-
tives. The research report clearly identified the
exploratory-descriptive qualitative methodology used
to conduct this study. The researchers took extensive
steps to ensure rigor in the data collection and analysis
phase and the development of study findings. An
unexpected finding was that patients reported anxiety
and fear more frequently than depression. Prior to this
study, researchers had focused their attention almost
exclusively on depression. On the basis of these find-
ings, the researchers would return to the literature to
examine studies on anxiety and fear. They may have
identified a gap in knowledge that warrants additional
research. In addition to addressing anxiety and fear of
patients with COPD, health professionals can use
these findings as the justification for the development
68 UNIT ONE  Introduction to Nursing Research
of a health promotion program for such patients and
their family members.
Historical Research
Historical research examines events of the past. His-
torians describe events in the context of time, social
structures, concurrent events, and key individuals.
These descriptions can increase understanding and
raise awareness of the forces shaping current events.
Historical nursing research can do the same for
the profession and its role in society (Lewenson &
Herrmann, 2008). Nurse researchers using historical
methods have examined the events and people that
shaped health in different settings and countries as
well as nursing as a profession. For example, Wood
(2009) examined the role of the surgical nurse in Aus-
tralia, Britain, and New Zealand in the era before
antibiotics. Her data sources were nursing articles and
textbooks published between 1895 and 1935. She
noted that the surgical nurse’s role was described in
relation to the surgeon rather than the patient. The
surgical nurse’s role was described as supporter,
scrubber, sentry, saboteur, and sloven. The last two
descriptors were applied in cases of patient sepsis fol-
lowing a surgery. The relationship between the surgeon
and the nurses went beyond simple subservience to
include admiration for each other’s skills and a con-
scientious approach to protecting the patient.
Wagner and Whaite (2010, p. 230) examined
selected writings of Florence Nightingale using
content analysis and identified five themes related to
caring: “attend to, attention to, nurture, competent,
and genuine.” Other historical researchers examined
recent events; for example, D’Antonio (2004) studied
the push for higher education of nurses as affected by
women’s opportunities as members of racial/ethnic
groups. These studies demonstrate the importance of
historical research in describing the past to provide
direction for the future in nursing education and prac-
tice (Lewenson & Herrmann, 2008).
Philosophical Orientation
History is a very old science that dates back to the
beginnings of humankind, when people and groups of
people asked, “Where have we come from?” Others
asked, “Who are we?” and “Where are we going?”
These questions have been asked throughout time.
Although the questions do not change, the answers
have changed because of the influence of social, cul-
tural, and personal forces. History provides the context
of experience (Lewenson & Herrman, 2008).
A primary assumption of historical philosophy is
that we can learn from the past and that this
knowledge can increase our understanding of the
present and future. The philosophy of history is a
search for wisdom. The historian examines what has
been, what is, and what ought to be. Influenced by the
values of the profession and the philosophy of prag-
matism, historical nurse researchers often apply the
lessons of the past to current events. Wagner and
Whaite (2010, p. 233) end the report of their study of
the writings of Florence Nightingale with these sen-
tences: “These research findings and others encourage
nurses to examine and reflect on how caring relation-
ships impact their practice … helping students to
understand how caring relationships in nursing can
occur in all practice settings.”
Historical Research’s Contribution
to Nursing Science
Historical research, although viewed as a legitimate
source of knowledge, has received less attention by
funding agencies and nurse researchers in the past 10
years because of a shift toward studies that produce
evidence to improve patient outcomes. The American
Association for the History of Nursing is an organiza-
tion with a focus on historical methodologies. The
association’s journal, Nursing History Review, serves
as vehicle for continued knowledge development
in this area. In this journal, Irwin (2011, p. 79) pub-
lished “a study of U.S. nurses who worked across
continents.”
The philosophical orientation and methodology of
the study received minimal attention, but the researcher
noted the role documents played in the study. “The
archives, sources, and figures that are central to the
field provide a ready means to trace the spread of U.S.
global influence in the early twentieth century” (Irwin,
2011, p. 80). Irwin could have strengthened the report
by providing more detail on methods used to ensure
rigor. Like most historical researchers, Irwin focused
the report on the product and provided little informa-
tion on the processes she used.
Irwin (2011) described four nurses who provided
overseas service through the American Red Cross
during World War I and in the immediate post-war
era. The nurses cared for military personnel but were
very involved also in relief efforts for citizens, refu-
gees, and families. They participated in efforts to
improve hygiene, care for those with tuberculosis,
and open training schools for nurses. Irwin’s contri-
bution to nursing knowledge was to increase nurses’
awareness of their potential influence. Building on
this history of nurse reformers, nurses today can
conceive of roles that extend beyond political
borders.
 CHAPTER 4  Introduction to Qualitative Research 69
KEY POINTS
• Qualitative research is a scholarly approach used to
describe life experiences from the perspective of
the persons involved.
• The philosophical foundation of qualitative re-
search describes a view of science and guides both
the selection of methods and the criteria of rigor.
• Quantitative research is guided by the philosophy
of logical positivism. In contrast, qualitative re-
search is guided by a wider range of philosophies.
• Qualitative researchers use open-ended methods to
gather data, such as interviews, focus groups,
observation, and examination of documents.
• The goal of phenomenological research is to
describe experiences as they are lived. Phenome-
nology is the philosophy guiding these studies, a
philosophy that began with the writings of Husserl.
• The goal of grounded theory research is to produce
findings grounded in the concrete world as experi-
enced by the participants that can be interpreted at
a more abstract level of a theory. Symbolic inter-
actionism is the underlying philosophical and theo-
retical perspective.
• Ethnographic research is the investigation of cul-
tures through an in-depth study of the members of
the culture. Nurse anthropologist Leininger devel-
oped the ethnonursing research method.
• Exploratory-descriptive qualitative research elicits
the perceptions of participants to provide insights
for understanding patients and groups, influencing
practice, and developing appropriate programs for
specific groups of people. The philosophy of prag-
matism and the naturalistic and interpretive orien-
tations guide exploratory-descriptive qualitative
studies.
• Historical research is designed to analyze the inter-
action of people, events, and social context that
occurred in the remote or recent past. The method-
ologies of historical research include interviewing
people with knowledge of past events and examin-
ing documents describing the events.
REFERENCES
Alexander, B. K. (2003). Fading, twisting, and weaving: An inter-
pretive ethnography of the Black barbershop as cultural space.
Qualitative Inquiry, 9(1), 105–128.
Banner, D. (2010). Becoming a coronary artery bypass graft surgery
patient: A grounded theory study of women’s experiences. Journal
of Clinical Nursing, 19(21/22), 3123–3133.
Bassett, D. (2004). Qualitative research. In C. Bassett (Ed.), Quali-
tative research in health care (pp. 1–5). London, England: Whurr
Publishers Ltd.
Boutain, D. M. (1999). Critical nursing scholarship: Exploring criti-
cal social theory with African American studies. Advances in
Nursing Science, 21(4), 37–47.
Brooks, C. F. (2011). Social performance and secret ritual: Battling
against obsessive-compulsive disorder. Qualitative Health
Research, 21(2), 249–261.
Byers, D. C., & France, N. E. M. (2008). The lived experience
of registered nurses providing care to patients with dementia in
the acute care setting: A phenomenological study. International
Journal of Caring, 12(4), 44–49.
Caress, A., Luker, K., & Chalmers, K. (2010). Promoting health of
persons with chronic obstructive pulmonary disease: Patients’ and
carers’ views. Journal of Clinical Nursing, 19(19), 564–573.
Chen, H. (2010). The lived experience of moving forward for clients
with spinal cord injury: A Parse research method study. Journal
of Advanced Nursing, 66(5), 1132–1141.
Clark, P. N., McFarland, M. R., Andrews, M. M., & Leininger, J.
(2009). Caring: Some reflections on the impact of the culture care
theory by McFarland & Andrews and a conversation with Lein­
inger. Nursing Science Quarterly, 22(3), 233–239.
Colaizzi, P. F. (1973). Reflection and research in psychology: A
phenomenological study of learning. Dubuque, IA: Kendall Hunt.
Creswell, J. W. (2009). Research design: Qualitative, quantitative,
and mixed methods approaches (3rd ed.). Los Angeles, CA: Sage.
Crossley, N. (2010). Networks and complexity: Directions for inter-
actionist research? Symbolic Interaction, 33(3), 341–363.
D’Antonio, P. (2004). Women, nursing, and baccalaureate education
in 20th century America. Journal of Nursing Scholarship, 36(4),
379–384.
Douglas, M. K., Kemppainen, J. K., McFarland, M. R., Papadopou-
los, I., Ray, M. A., Roper, J. M., Scollan-Koliopoulos, M., Shapira,
J., & Tsai, H-M. (2010). Chapter 10: Research methodologies for
investigating cultural phenomena and evaluating interventions.
Journal of Transcultural Nursing, 21(Suppl. 1), 3737–4055.
Dowling, M. (2007). From Husserl to van Manen: A review of dif-
ferent phenomenological approaches. International Journal of
Nursing Studies, 44(1), 131–142.
Earle, V. (2010). Phenomenology as research method or substantive
metaphysics? An overview of phenomenology’s uses in nursing.
Nursing Philosophy, 11(4), 286–296.
Fawcett, J., & Garity, J. (2009). Evaluating research for evidence-
based nursing practice. Philadelphia, PA: F. A. Davis.
Fontana, J. S. (2004). A methodology for critical science in nursing.
Advances in Nursing Science, 27(2), 93–101.
Forman, J., Creswell, J. W., Damschroder, L., Kowalski, C. P.,
Krein, S. L. (2008). Qualitative research methods: Key features
and insights gained from use in infection prevention research.
American Journal of Infection Control, 36(10), 764–771.
Gardezi, F., Lingard, L., Espin, L. S., Whyte, S., Orser, B., & Baker,
R. (2009). Silence, power, and communication in the operating
room. Journal of Advanced Nursing, 65(7), 1390–1399.
Giorgi, A. (1985). Phenomenology and psychological research.
Pittsburg, PA: Duquesne University Press.
Giorgi, A. (2010). Phenomenological psychology: A brief history
and its challenges. Journal of Phenomenological Psychology,
41(2), 145–179.
Giske, T., & Artinian, B. (2008). Patterns of “balancing between
hope and despair” in the diagnostic phase: A grounded theory
70 UNIT ONE  Introduction to Nursing Research
study of patients on a gastroenterology ward. Journal of Advanced
Nursing, 62(1), 22–31.
Glaser, B. G., & Strauss, A. (1967). The discovery of grounded
theory: Strategies for qualitative research. Chicago, IL: Aldine.
Heidegger, M. (1927/1962). Being in time (J. Macquarrie & E.
Robinson, Trans.). New York, NY: Harper.
Husserl, E. (1901/1970). Logical investigations: Vol. 1 (N. Findlay,
Trans.). New York, NY: Routledge.
Irwin, J. F. (2011). Nurses without borders: The history of nursing
as U.S. international history. Nursing Historical Review, 19(1),
78–102.
Johnson, M. (2000). Heidegger and meaning: Implications for phe-
nomenological research. Nursing Philosophy, 1(2), 134–146.
Kerlinger, F. N., & Lee, H. P. (2000). Foundations of behavioral
research (4th ed.). Fort Worth, TX: Harcourt College.
Leininger, M. M. (1970). Nursing and anthropology: Two worlds
to blend. New York, NY: Wiley.
Leininger, M. M. (1997). Overview of the Theory of Culture Care
with the ethnonursing research method. Journal of Transcultural
Nursing, 8(2), 32–54.
Leininger, M. M. (2002). Culture care theory: A major contribution
to advance transcultural nursing knowledge and practices. Journal
of Transcultural Nursing, 13(3), 189–192.
Lesniak, R. G. (2010). The lived experience of adolescent females
who self-injure by cutting. Advanced Emergency Nursing Journal,
32(2), 137–147.
Lewenson, S. B., & Herrmann, E. K. (2008). Why do historical
research? In S. B. Lewenson & E. K. Hermann (Eds.), Capturing
nursing history (pp. 1–10). New York, NY: Springer.
Liamputtong, P., & Ezzy, D. (2005). Qualitative research methods
(2nd ed.). Melbourne, VIC, Australia: Oxford University Press.
Maliski, S. L., Connor, S. E., Williams, L., & Litwin, M. S. (2010).
Faith among low-income, African American/Black men treated
for prostate cancer. Cancer Nursing, 33(6), 470–478.
Maliski, S., Rivera, S., Connor, S., Lopez, G., & Litwin, M. (2008).
Renegotiating masculine identity after prostate cancer treatment.
Qualitative Health Research, 18(12), 1609–1620.
Marshall, C., & Rossman, G. B. (2011). Designing qualitative
research (5th ed.). Los Angeles, CA: Sage.
McCready, J. S. (2010). Jamesian pragmatism: A framework for
working towards unified diversity in nursing knowledge develop-
ment. Nursing Philosophy, 11(3), 191–203.
McGibbon, E., Peter, E. & Gallop, R. (2010). An institutional eth-
nography of nurses’ stress. Qualitative Health Research, 20(10),
1353–1378.
Mead, G. H. (1934). Mind, self, and society. Chicago, IL: University
of Chicago Press.
Meah, A., Callery, P., Milnes, L., & Rogers, S. (2009). Thinking
“taller”: Sharing responsibility in the everyday lives of children
with asthma. Journal of Clinical Nursing, 19(13/14), 1952–
1959.
Mendes, B., Roux, G., & Ridosh, M. (2010). Phenomenon of inner
strength in women post-myocardial infarction. Critical Care
Nursing Quarterly, 33(3), 248–258.
Merleau-Ponty, M. (1945/2002). Phenomenology of perception (C.
Smith, Trans.). London, England: Routledge Classics.
Morse, J., & Richards, L. (2002). Readme first for a user’s guide to
qualitative methods. Thousand Oaks, CA: Sage.
Munhall, P. L. (2012). Nursing research: A qualitative perspective
(5th ed.). Sudbury, MA: Jones & Bartlett.
Noiseux, S., & Ricard, N. (2008). Recovery as perceived by persons
with schizophrenia, family members, and health professionals:
A grounded theory. International Journal of Nursing Studies,
45(8), 1148–1162
Parse, R. R. (1981). Man-living-health: A theory of nursing. New
York, NY: Wiley.
Parse, R. R. (1992). Human becoming: Parse’s theory of nursing.
Nursing Science Quarterly, 5(1), 35–42.
Parse, R. R. (2001). Qualitative inquiry: The path of sciencing. New
York, NY: National League for Nursing.
Parse, R. R. (2011). The human becoming modes of inquiry: Refine-
ments. Nursing Science Quarterly, 24(1), 11–15.
Paterson, J. G., & Zderad, L. T. (1976). Humanistic nursing. New
York, NY: Wiley.
Phillips-Pula, L., Strunk, J., & Pickler, R. H. (2011). Understanding
phenomenological approaches to data analysis. Journal of Pedi-
atric Health Care, 25(1), 67–71.
Reed, I. A. (2010). Epistemology contextualized: Social-scientific
knowledge in a post-positivistic era. Sociological Theory, 28(1),
20–39.
Rohr, Y., Adams, J., & Young, L. (2010). Oral discomfort in pallia-
tive care: Results of an exploratory study of the experiences of
terminally ill patients. International Journal of Palliative Nursing,
16(9), 439–444.
Sandelowski, M. (2000). What happened to qualitative description?
Research in Nursing & Health, 23(4), 334–340.
Sandelowski, M. (2010). What’s in a name? Qualitative description
revisited. Research in Nursing & Health, 33(1), 77–84.
Schumacher, G. (2010). Culture care meanings, beliefs, and prac-
tices in rural Dominican Republic. Journal of Transcultural
Nursing, 21(2), 93–103.
Shadish, W. R., Cook, T. D., & Campbell, D. T. (2001). Experimen-
tal and quasi-experimental designs for generalization causal
inference. Chicago, IL: Rand McNally.
Shorter, M., & Stayt, L. C. (2010) Critical care nurses’ experiences
of grief in an adult intensive care unit. Journal of Advanced
Nursing, 66(1), 159–167.
Speziale, H. J. S., & Carpenter, D. R. (2007). Qualitative research
in nursing: Advancing the humanistic perspective (4th ed.). Phila-
delphia, PA: Lippincott Williams & Wilkins.
Swanlund, S. L. (2010). Successful cardiovascular medication man-
agement processes as perceived by community-dwelling adults
over age 74. Applied Nursing Research, 23(1), 22–29.
Trimm, D. R., & Sanford, J. T. (2010). The process of family
waiting during surgery. Journal of Family Nursing, 16(4),
435–461.
Tuck, I., & Thinganjana, W. (2007). An exploration of the meaning
of spirituality voiced by persons living with HIV disease and
health adults. Issues in Mental Health Nursing, 28(2), 151–166.
van Manen, M. (1990). Researching lived experience: Human
science for an action sensitive pedagogy. Ontario, Canada:
Althouse Press.
Vincent, D. (2009). Culturally tailored education to promote life-
style change in Mexican Americans with type 2 diabetes. Journal
of the American Academy of Nurse Practitioners, 21(9),
520–527.
 CHAPTER 4  Introduction to Qualitative Research 71
Wagner, D. J., & Whaite, B. (2010). An exploration of the nurse of
caring relationships in the writings of Florence Nightingale.
Journal of Holistic Nursing, 28(4), 225–234.
Watson, J. (1997). Theory of human caring: Retrospective and pro-
spective. Nursing Science Quarterly, 10(1), 49–52.
Watson, J. (1999). Nursing: Human science and human care:
A theory of nursing. Boston, MA: Jones & Bartlett.
Wolf, Z. E. (2012). Ethnography: The method. In P. L. Munhall
(Ed.). Nursing research: A qualitative perspective (5th ed.) (pp.
285–338). Sudbury, MA: Jones & Bartlett.
Wood, P. J. (2009). Supporting or sabotaging the surgeon’s efforts:
Portrayals of the surgical nurse’s role in preventing wound sepsis,
1895-1935. Journal of Clinical Nursing, 18(19), 2739–2746.
Zuzelo, P. R. (2012). Evidence-based nursing and qualitative
research: A partnership imperative for real-world practice. In
P. L. Munhall (Ed.). Nursing research: A qualitative perspective
(5th ed.) (pp. 533–552). Sudbury, MA: Jones & Bartlett.
This page intentionally left blank
UNIT TWO
The Research Process
Research Problem and Purpose
  http://evolve.elsevier.com/Grove/practice/
W
e are constantly asking questions to better
understand ourselves and the world around
us. This human ability to wonder and ask
creative questions about behaviors, experiences, and
situations in the world provides a basis for identifying
research topics and problems. Identifying a problem
is the initial step, and one of the most significant, in
conducting quantitative, qualitative, outcomes, and
intervention research. The research purpose evolves
from the problem and directs the subsequent steps of
the research process.
Research topics are concepts, phenomena of inter-
est, or broad problem areas that researchers can focus
on to enhance evidence-based nursing. Research
topics contain numerous potential research problems,
and each problem provides the basis for developing
many purposes. Thus, the identification of a relevant
research topic and a challenging, significant problem
can facilitate the development of numerous study
purposes to direct a lifetime program of research.
However, the abundance of research topics and poten-
tial problems frequently is not apparent to nurses
struggling to identify their first study problem.
This chapter differentiates a research problem from a
purpose, identifies sources for research problems, and pro-
vides a background for formulating a problem and pur­
pose for study. The criteria for determining the feasibility
of a proposed study problem and purpose are described.
The chapter concludes with examples of research topics,
problems, and purposes from current quantitative, quali­
tative, outcomes, and intervention studies.
What Is a Research Problem
and Purpose?
A research problem is an area of concern where there
is a gap in the knowledge base needed for nursing
5  
CHAPTER
practice. Research is conducted to generate knowledge
that addresses the practice concern, with the ultimate
goal of providing evidence-based health care. A research
problem can be identified by asking questions such as
the following: What is wrong or is of concern in this
clinical situation? What knowledge is needed to improve
this situation? Will a particular intervention work in this
clinical situation? What is known about this interven-
tion’s effectiveness? Would another intervention be
more effective in producing the desired outcomes?
By questioning and reviewing the literature,
researchers begin to recognize a specific area of
concern and the knowledge gap that surrounds it. The
knowledge gap, or what is not known about this clini-
cal problem, determines the complexity and number
of studies needed to generate essential knowledge
for nursing practice (Craig & Smyth, 2012; Creswell,
2009). In addition to the area of concern, the research
problem identifies a population and sometimes a
setting for the study.
A research problem includes significance, back-
ground, and a problem statement. The significance of
a problem indicates the importance of the problem to
patients and families, nursing, healthcare system, and
society. The background for a research problem
briefly identifies what we know about the problem
area. The problem statement identifies the specific
gap in the knowledge needed for practice. A research
problem from the study by Grady, Entin, Entin, and
Brunye (2011) is presented as an example. This study
was conducted to examine the effectiveness of educa-
tional messages or information on the knowledge,
attitudes, and behaviors of people with diabetes.
“Diabetes prevalence has reached epidemic propor-
tions in this country. The health and economic
73
74 UNIT TWO  The Research Process
consequences for Americans with this disease are
overwhelming and expected to grow as our popula-
tion continues to age. Approximately 23.6 million
people in the United States have diabetes and, despite
the disease being underreported as a cause of death,
diabetes was listed as the seventh leading cause of
death in the United States in 2006 (Centers for
Disease Control and Prevention, 2008a). The direct
medical costs of diabetes care and complications of
$116 billion, together with indirect costs of $58 billion
related to disability and reduced productivity, resulted
in an estimated economic cost of diabetes totaling
$174 billion in 2007 (American Diabetes Association,
2009).… Complications contribute to a risk of death
among individuals with diabetes that is about 2 times
higher than that of individuals without diabetes
(Centers for Disease Control and Prevention, 2008a).
Amputations and foot ulcerations are the most
common consequences of diabetic neuropathy and
the major causes of morbidity and disability in people
with diabetes. Approximately 2% to 3% of individu-
als with diabetes develop one or more foot ulcers
each year, and an estimated 15% will develop a foot
ulcer during their lifetime (Singh, Armstrong, & Lipsky,
2005) [problem significance].…
As the cornerstone of diabetes treatment and an
integral part of a self-management regime, education
of patients with diabetes takes place in both inpatient
and outpatient venues.… Patient education takes time
in the continuum of care that an already overworked
staff is challenged to provide.… The research cited in
the reviews of Boren et al. (2006) and Jackson et al.
(2006) provides evidence that delivery of healthcare
information can be accomplished effectively without
involving diabetes educators or nurses and offers
support for the use of information-technology-based
education as an alternative way to provide informa-
tion and guidance to persons with diabetes [problem
background]. However, regardless of whe­ther the
information is presented in person or via technology,
a relevant and still-open question is how to present
the information so as to foster positive attitudinal and
behavioral change and maximize the long-term effec-
tiveness of health management education [problem
statement].” (Grady et al., 2011, pp. 22-23)
In this example, the research problem identifies an
area of concern (incidence, costs, and complications
of diabetes) for a particular population (persons
with diabetes) in selected settings (inpatient and out-
patient venues). Grady and colleagues (2011) clearly
identified the significance of the problem, which is
extensive and relevant to patients, families, nursing,
healthcare system, and society. The problem back-
ground focuses on key research conducted to exam­ine
the effectiveness of health education on the manage-
ment of diabetes. The last sentence in this example is
the problem statement, which identifies the gap in the
knowledge needed for practice. In this study, there is
limited research on how to present diabetic education
to maximize its effectiveness on attitudinal and behav-
ioral change in people with this chronic illness.
The research problem in this example includes
concepts or research topics such as diabetes preva-
lence, economic consequences, complications of dia-
betes, consequences of diabetic neuropathy, health
management education, self-management, and atti­
tudinal and behavioral changes. Health management
education is an abstract concept, and a variety
of nursing actions or interventions could be imple-
mented to determine their effectiveness in promoting
long-term attitudinal and behavioral changes in
persons with diabetes. Thus, each problem may gen-
erate many research purposes. The knowledge gap
regarding how to present information to foster posi-
tive attitudinal and behavioral changes in persons
with diabetes provides clear direction for formulating
the research purpose.
The research purpose is a clear, concise statement
of the specific focus or aim of the study that is gener-
ated on the basis of the research problem. The purpose
usually indicates the type of study (quantitative, quali-
tative, outcomes, or intervention) to be conducted and
often includes the variables, population, and setting
for the study. The goals of quantitative research
include identifying and describing variables, examin-
ing relationships among variables, and determining
the effectiveness of interventions in managing clinical
problems (Creswell, 2009; Shadish, Cook, & Camp-
bell, 2002). The goals of qualitative research include
exploring a phenomenon, such as depression as it is
experienced by pregnant women; developing theories
to describe and manage clinical situations; examining
the health practices of certain cultures; describing
health-related issues, events, and situations; and deter-
mining the historical evolution of the profession
(Marshall & Rossman, 2011; Munhall, 2012). The
focus of outcomes research is to identify, describe,
and improve the outcomes or end results of patient
care (Doran, 2011). Intervention research focuses on
investigating the effectiveness of nursing interven-
tions in achieving the desired outcomes in natural
settings (Forbes, 2009). Regardless of the type of

research, every study needs a clearly expressed pur­
pose statement to guide it. Grady et al. (2011) clearly
identified their study purpose following their research
problem statement of the gap in the knowledge base.
Thus, the purpose of their study was to “examine the
impact of information framing in an educational
program about proper foot care and its importance for
preventing diabetic complications on long-term
changes in foot care knowledge, attitudes, and behav-
ior” (Grady et al., 2011, p. 23).
This research purpose indicates that these inves-
tigators conducted a quantitative quasi-experimental
study to determine the effectiveness of an indepen-
dent variable or intervention (information framing
educational program about diabetic foot care and
prevention of complications) on the dependent or
outcome variables (foot care knowledge, attitudes,
and behaviors). The researchers also identified two
hypotheses to direct their study, which included the
four variables identified (see Chapter 8 for a discus-
sion of hypotheses). The study findings indicated
that the gain-framed messages focused on the ben-
efits of taking action were significantly more effec-
tive in promoting positive behavioral changes in
people with diabetes than the loss-framed messages
focused on the costs of not taking action. A gain-
framed message might be stated as follows: “Achiev-
ing normal blood sugar increases your feelings of
health and well being and promotes control of your
illness.” A loss-framed message might be worded
as follows: “Poorly controlled blood sugars can lead
to complications of neuropathy, foot lesions, and
amputation.” Grady et al. (2011) also found that
changes in knowledge affected changes in attitudes
and that attitudes were direct predictors of long-term
behavior management of diabetes. The findings from
this study and other research provide evidence of
the effectiveness of information messages in sus­
taining health promoting behavior by people with
diabetes.
Sources of Research Problems
Research problems are developed from many sources,
but you need to be curious, astute, and imaginative to
identify problems from the sources. The sources for
research problems included in this text are (1) clinical
practice, (2) researcher and peer interactions, (3) lit-
erature review, (4) theories, and (5) research priorities
identified by funding agencies and specialty groups.
Researchers often use more than one source to identify
and refine their research problem.
CHAPTER 5  Research Problem and Purpose 75
Clinical Practice
The practice of nursing must be based on knowledge
or evidence generated through research. Thus, clinical
practice is an extremely important source for research
problems. Problems can evolve from clinical observa-
tions. For example, while watching the behavior of a
patient and family in crisis, you may wonder how you
as a nurse might intervene to improve the family’s
coping skills. A review of patient records, treatment
plans, and procedure manuals might reveal concerns
or raise questions about practice that could be the basis
for research problems. For example, you may wonder:
What nursing intervention will open the lines of
communication with a patient who has had a stroke?
What is the impact of home visits on the level of func-
tion, readjustment to the home environment, and
rehospitalization pattern of a child with a severe
chronic illness? What is the most effective treatment
for acute and chronic pain? What is the best pharma-
cological agent or agents for treating hypertension
in elderly, diabetic patients—angiotensin-converting
enzyme (ACE) inhibitor, angiotensin II receptor
blocker (ARB), diuretic, beta blocker, calcium channel
blocker, or alpha antagonist, or a combination of these
drugs? What are the most effective pharmacological
and nonpharmacological treatments for a patient with
a serious and persistent mental illness? What are the
needs of stroke survivors from their perspective?
What are the cultural factors that promote better birth
outcomes in Hispanic women? These clinical ques-
tions could direct you in identifying a significant
research problem and purpose.
Extensive patient data, such as diagnoses, treat-
ments, and outcomes, are now computerized. Analyz-
ing this information might generate research problems
that are significant to a clinic, community, or national
healthcare system. For example, you may ask: Why
has adolescent obesity increased so rapidly in the past
10 years, and what treatments will be effective in
managing this problem? What pharmacological and
nonpharmacological treatments have been most effec-
tive in treating common acute illnesses such as otitis
media, sinusitis, and bronchitis in your practice or
nationwide? What are the outcomes (patient health
status and costs) for treating such chronic illnesses as
type 2 diabetes, hypertension, and dyslipidemia in
your practice? Review of agency patient data often
reveals patterns and trends in a clinical setting and
helps nurses and students to identify patient care
problems.
Because health care is constantly changing in
response to consumer needs and trends in society, the
76 UNIT TWO  The Research Process
focus of current research varies according to these
needs and trends. For example, research evidence is
needed to improve practice outcomes for infants and
new mothers, the elderly and residents in nursing
homes, and persons from vulnerable and culturally
diverse populations. Healthcare agencies would
benefit from studies of varied healthcare delivery
models. Society would benefit from interventions rec-
ognized to promote health and prevent illness. In
summary, clinically focused research is essential if
nurses are to develop the knowledge needed for
evidence-based practice (EBP) (Brown, 2009; Melnyk
& Fineout-Overholt, 2011).
Researcher and Peer Interactions
Interactions with researchers and peers offer valuable
opportunities for generating research problems. Expe-
rienced researchers serve as mentors and help novice
researchers to identify research topics and formulate
problems. Nursing educators assist students in select-
ing research problems for theses and dissertations.
When possible, students conduct studies in the same
area of research as the faculty. Faculty members can
share their expertise regarding their research program,
and the combined work of the faculty and students can
build a knowledge base for a specific area of practice.
This type of relationship could also be developed
between an expert researcher and a nurse clinician.
Because nursing research is critical for designation as
a Magnet facility by the American Nurses Credential-
ing Center© (ANCC, 2012), hospitals and healthcare
systems employ nurse researchers for the purpose of
guiding studies conducted by staff nurses. Building an
EBP for nursing requires collaboration between nurse
researchers and clinicians as well as with researchers
from other health-related disciplines. Interdisciplinary
research teams have the expertise to increase the
quality and quantity of studies conducted. Being a part
of a research team is an excellent way to expand your
understanding of the research process.
Beveridge (1950) identified several reasons for dis-
cussing research ideas with others. Ideas are clarified
and new ideas are generated when two or more people
pool their thoughts. Interactions with others enable
researchers to uncover errors in reasoning or informa-
tion. These interactions are also a source of support in
discouraging or difficult times. In addition, another
person can provide a refreshing or unique viewpoint,
which helps avoid conditioned thinking, or following
an established habit of thought. A workplace that
encourages interaction can stimulate nurses to identify
research problems. Nursing conferences and profes-
sional meetings also provide excellent opportunities
for nurses to discuss their ideas and brainstorm to
identify potential research problems.
The Internet has greatly extended the ability of
researchers and clinicians around the world to share
ideas and propose potential problems for research.
Most colleges or schools of nursing have websites that
identify faculty research interests and provide mecha-
nisms for contacting individuals who are conducting
research in your area of interest. Thus, interactions
with others are essential to broaden your perspective
and knowledge base and to support you in identifying
significant research problems and purposes.
Literature Review
Reviewing research journals, such as Advances in
Nursing Science, Applied Nursing Research, Clinical
Nursing Research, Evidence-Based Nursing, Interna-
tional Journal of Psychiatric Nursing Research,
Journal of Nursing Scholarship, Journal of Advanced
Nursing, Journal of Research in Nursing, Nursing
Research, Nursing Science Quarterly, Research in
Nursing & Health, Scholarly Inquiry for Nursing
Practice: An International Journal, Southern Online
Journal of Nursing Research, and Western Journal of
Nursing Research, as well as theses and dissertations
will acquaint novice researchers with studies con-
ducted in an area of interest. The nursing specialty
journals, such as American Journal of Maternal Child
Nursing, Archives of Psychiatric Nursing, Dimensions
of Critical Care, Heart & Lung, Infant Behavior and
Development, Journal of Pediatric Nursing, and
Oncology Nursing Forum, also place a high priority
on publishing research findings. Reviewing research
articles enables you to identify an area of interest and
determine what is known and not known in this area.
The gaps in the knowledge base provide direction for
future research. (See Chapter 6 for the process of
reviewing the literature.)
At the completion of a research project, an inves-
tigator often makes recommendations for further
study. These recommendations provide opportunities
for others to build on a researcher’s work and
strengthen the knowledge in a selected area. For
example, the Grady et al. (2011, p. 27) study, intro-
duced earlier in this chapter, provided recommenda-
tions for further research to examine “the longer term
eventualities of gain- and loss-framed messages on
preventative behaviors.” They also recommended
examining how long the gain-framed message might
last and when it would be “necessary to provide
another message presentation to bolster effective self-
care behavior” (p. 27). These researchers also encour-
aged others to validate their findings through

replication studies that varied the content and delivery
format of educational messages provided persons with
diabetes.
Replication of Studies
Reviewing the literature is a way to identify a study
to replicate. Replication involves reproducing or
repeating a study to determine whether similar find-
ings will be obtained (Fahs, Morgan, & Kalman,
2003). Replication is essential for knowledge develop-
ment because it (1) establishes the credibility of
the findings, (2) extends the generalizability of the
findings over a range of instances and contexts, (3)
reduces the number of type I and type II errors,
(4) corrects the limitations in studies’ methodologies,
(5) supports theory development, and (6) lessens the
acceptance of erroneous results. Some researchers
replicate studies because they agree with the findings
and wonder whether the findings will hold up in dif-
ferent settings with different subjects over time. Others
want to challenge the findings or interpretations of
prior investigators. Some researchers develop research
programs focused on expanding the knowledge needed
for practice in an area. This program of research often
includes replication studies that strengthen the evi-
dence for practice.
Four different types of replications are important in
generating sound scientific knowledge for nursing: (1)
exact, (2) approximate, (3) concurrent, and (4) system-
atic extension (Haller & Reynolds, 1986). An exact
(or identical) replication involves duplicating the
initial researcher’s study to confirm the original find-
ings. All conditions of the original study must be
maintained; thus, “there must be the same observer,
the same subjects, the same procedure, the same mea-
sures, the same locale, and the same time” (Haller &
Reynolds, 1986, p. 250). Exact replications might be
thought of as ideal to confirm original study findings,
but these are frequently not attainable. In addition, one
would not want to replicate the errors in an original
study, such as small sample size, weak design, or poor-
quality measurement methods.
When conducting an approximate (or operational)
replication, the subsequent researcher repeats the
original study under similar conditions, following the
methods as closely as possible. The intent is to deter-
mine whether the findings from the original study hold
up despite minor changes in the research conditions.
If the findings generated through replication are con-
sistent with the findings of the original study, then the
knowledge is considered more credible and has a
greater probability of accurately reflecting the real
world. If the replication fails to support the original
CHAPTER 5  Research Problem and Purpose 77
findings, the designs and methods of both studies
should be examined for limitations and weaknesses,
and further research must be conducted. Conflicting
findings might also generate additional theoretical
insights and provide new directions for research.
For a concurrent (or internal) replication, the
researcher collects data for the original study and the
replication study simultaneously thereby checking
the reliability of the original study findings. The con-
firmation, through replication of the original study
findings, is part of the original study’s design. For
example, your research team might collect data simul-
taneously at two different hospitals to compare and
contrast the findings. Consistency in the findings
increases the credibility of the study and the likelihood
that others will be able to generalize the findings.
Some expert researchers obtain funding to conduct
multiple concurrent replications, in which a number
of individuals conduct repetitions of a single study, but
with different samples in different settings. Clinical
trials that examine the effectiveness of the pharmaco-
logical management of chronic illnesses, such as dia-
betes, hypertension, and dyslipidemia, are examples
of concurrent replication studies. As each study is
completed, the findings are compiled in a report that
specifies the series of replications that were conducted
to generate these findings. Some outcome studies
involve concurrent replication to determine whether
the outcomes vary for different healthcare providers
and healthcare settings across the United States (Brink
& Wood, 1979; Brown, 2009; Doran, 2011).
A systematic (or constructive) replication is done
under distinctly new conditions. The researchers con-
ducting the replication do not follow the design or
methods of the original researchers; rather, the second
investigative team identifies a similar problem but for-
mulates new methods to verify the first researchers’
findings (Haller & Reynolds, 1986). The aim of this
type of replication is to extend the findings of the
original study and test the limits of the generalizability
of such findings. Intervention research might use this
type of replication to examine the effectiveness of
various interventions devised to address a practice
problem.
Nurse researchers need to actively replicate studies
to develop strong research evidence for practice.
However, the number of nursing studies replicated
continues to be limited. The replications of studies
might be limited because (1) some view replication as
less scholarly or less important than original research,
(2) the discipline of nursing lacks adequate resources
and funding for conducting replication studies, and (3)
editors of journals publish fewer replication studies
78 UNIT TWO  The Research Process
than original studies (Fahs et al., 2003). However, the
lack of replication studies severely limits the genera-
tion of sound research findings needed for EBP in
nursing. Thus, replicating a study should be respected
as a legitimate scholarly activity for both expert and
novice researchers. Funding from both private and
federal sources is needed to support the conduct of
replication studies, with a commitment from journal
editors to publish these studies.
Replication provides an excellent learning oppor-
tunity for the novice researcher to conduct a signifi-
cant study, validate findings from previous research,
and generate new research evidence about different
populations and settings. Students studying for a mas-
ter’s of science in nursing degree could be encouraged
to replicate studies for their theses, possibly to repli-
cate faculty studies. Expert researchers, with programs
of research, implement replication studies to generate
sound evidence for use in practice. When publishing
a replication study, researchers need to designate the
type of replication conducted and the contribution the
study made to the existing body of knowledge.
Theory
Theories are an important source for generating
research problems because they set forth ideas about
events and situations in the real world that require
testing (Chinn & Kramer, 2008). In examining a
theory, you may note that it includes a number of
propositions and that each proposition is a statement
of the relationship of two or more concepts. A research
problem and purpose could be formulated to explore
or describe a concept or to test a proposition from a
theory. Middle range theories are the ones most com-
monly used as frameworks for quantitative studies and
are tested as part of the research process (Smith &
Liehr, 2008). In qualitative research, the purpose of
the study might be to generate a theory or framework
to describe a unique event or situation (Marshall &
Rossman, 2011; Munhall, 2012).
Some researchers combine ideas from different
theories to develop maps or models for testing through
research. The map serves as the framework for the
study and includes key concepts and relationships
from the theories that the researchers want to study.
Frenn, Malin, and Bansal (2003, p. 38) conducted a
quasi-experimental study to examine the effectiveness
of a “4-session Health Promotion/Transtheoretical
Model-guided intervention in reducing percentage of
fat in the diet and increasing physical activity among
low- to middle-income culturally diverse middle
school students.” The intervention was based on the
“components of two behaviorally based research
models that have been well tested among
adults—Health Promotion Model (Pender, 1996) and
Transtheoretical Model (Prochaska, Norcross, Fowler,
Follick, & Abrams, 1992)—but have not been tested
regarding low-fat diet with middle school-aged chil-
dren” (Frenn et al., 2003, p. 36). They developed a
model of the study framework (see Figure 5-1) and
described the concepts and propositions from the
model that guided the development of different aspects
of their study.
“A combined Health Promotion/Transtheoretical
Model guided the intervention designed for this study
[see Figure 5-1]. The first individual characteristic
examined in this study was temptation (low self-
efficacy), defined as the inability to overcome barriers
in sustaining a low-fat diet … and an intervention
helping adolescents develop behavioral control may
enhance self-efficacy and improve health habits.
The second characteristic common to both the
Health Promotion and Transtheoretical Models was
benefits/barriers. In a study of fifth- through seventh-
grade children, Baranowski and Simons-Morton
(1990) found the most common barriers to reducing
saturated fat in the diet were (a) giving up preferred
foods, (b) meals outside the home that contained fat,
(c) not knowing what foods were low in fat, and (d)
not wanting to take the time to read labels.
The last individual characteristic used in this study
was access to low-fat foods. This concept from the
Health Promotion Model is important in a middle
school-aged population, as they are, to some extent,
dependent on others for the types of food available.”
(Frenn et al., 2003, pp. 37-38)
Frenn et al. (2003) used the Pender (1996) Health
Promotion Model and the Transtheoretical Model
(Prochaska et al., 1992), which are middle range theo-
ries, to develop the following research questions to
guide their study:
“(a) Do demographic variables, access to low-fat
foods, perceived self-efficacy, benefits/barriers, and
stages of change predict percentage of fat reported
in the diet by middle school-aged children? (b) Does
the application of a Health Promotion/Transtheoretical
Model intervention in 4 classroom sessions signifi-
cantly improve adoption of a diet lower in fat and
duration of physical activity as compared with a
control group of students not engaged with the
program?” (Frenn et al., 2003, p. 39)

Individual State of Change
Characteristics Transition
Age
Pre-contemplation
Race to contemplation
Gender
Income
Preparation
Temptation to action
(low efficacy)
Benefits/barriers
(pros/cons)
Access to Action to
low-fat foods maintenance
Figure 5-1  The health promotion stage of change model: A synthesis of health promotion and transtheoretical models guiding low-fat diet
intervention for students in an urban middle school.
The findings from a study either support or do not
support the relationships identified in the model. The
study by Frenn et al. (2003) added support to the
Health Promotion/Transtheoretical Model with their
findings that the classroom intervention decreased
dietary fat and increased physical activity for middle
school–age adolescents. Further research is needed to
determine whether classroom interventions over time
reduce body mass index, body weight, and the per-
centage of body fat of overweight and obese adoles-
cents. As a graduate student, you could use this model
as a framework and test some of the relationships in
your clinical setting.
Research Priorities
Since 1975, expert researchers, specialty groups, pro-
fessional organizations, and funding agencies have
identified nursing research priorities. The research pri-
orities for clinical practice were initially identified in
a study by Lindeman (1975). Those original research
priorities included nursing interventions related to
stress, care of the aged, pain management, and patient
education. Developing evidence-based nursing inter-
ventions in these areas continues to be a priority.
Many professional nursing organizations use web-
sites to communicate their current research priorities.
For example, the American Association of Critical-
Care Nurses (AACN) determined initial research
priorities for this specialty in the early 1980s
CHAPTER 5  Research Problem and Purpose 79
Interventions Outcome
Self-reevaluation
Consciousness raising
regarding benefits and
access to low-fat
alternatives
Building efficacy and skill
in preparation of low-fat snacks Decreased
Environmental percentage of
Reevaluation to create fat in diet
options and access
Self-reevaluation as
model for peers
Self-liberation
Reinforcement
Management
Helping relationships
Counterconditioning
(Lewandowski & Kositsky, 1983) and revised these
priorities on the basis of patients’ needs and the
changes in health care. The current AACN (2011)
research priorities are identified on this organization’s
website as (1) effective and appropriate use of technol-
ogy to achieve optimal patient assessment, manage-
ment, or outcomes, (2) creation of a healing, humane
environment, (3) processes and systems that foster the
optimal contribution of critical care nurses, (4) effec-
tive approaches to symptom management, and (5)
prevention and management of complications. AACN
(2011) has also identified future research needs under
the following topics: medication management, hemo-
dynamic monitoring, creating healing environments,
palliative care and end-of-life issues, mechanical ven-
tilation, monitoring of neuroscience patients, and non-
invasive monitoring. If your specialty is critical care,
this list of research needs might help you identify a
priority problem and purpose for study.
The American Organization of Nurse Executives
(AONE, 2012) provides a discussion of their educa-
tion and research priorities online at http://www
.aone.org/education/index.shtml/. For 2011-2012,
AONE identified more than 25 research priorities in
four strategic areas: (1) design of future patient care
delivery systems, (2) healthful practice environments,
(3) leadership, and (4) the positioning of nurse leaders
as valued healthcare executives and managers. To
promote the design of future patient care delivery
80 UNIT TWO  The Research Process
systems, AONE encourages research focused on new
technology, patient safety, and the work environment
that allows strategies for improvement crucial to the
success of the delivery system. In the area of healthful
practice environments, AONE encourages research
focused on practice environments that attract and
retain nurses and that promote professional growth
and continuous learning, including mentoring of staff
nurses and nursing leaders. In the area of leadership,
AONE encourages research focused on evidence-
based leadership capacity, measurement of patient
care quality outcomes, and technology to complement
patient care. To promote the positioning of nurse
leaders as valued healthcare executives and managers,
AONE encourages research focused on patient safety
and quality, disaster preparedness, and workforce
shortages. AONE recognizes the importance of sup-
porting education and research initiatives to create a
healthy work environment, a quality healthcare
system, and strong nurse executives. You can search
online for the research priorities of other nursing orga-
nizations to help you identify priority problems for
study.
A significant funding agency for nursing research
is the National Institute of Nursing Research (NINR).
A major initiative of the NINR is the development
of a national nursing research agenda that involves
identifying nursing research priorities, outlining a
plan for implementing priority studies, and obtaining
resources to support these priority projects. The NINR
has an annual budget of more than $90 million, with
approximately 74% of the budget used for extramural
research project grants, 7% for predoctoral and post-
doctoral training, 6% for research management and
support, 5% for the centers program in specialized
areas, 5% for other research including career develop-
ment, 2% for the intramural program, and 1% for
contracts and other expenses (see NINR at http://
www.ninr.nih.gov/).
The NINR (2011) developed four strategies for
building the science of nursing: “(1) integrating bio-
logical and behavior science for better health; (2)
adopting, adapting, and generating new technologies
for better health care; (3) improving methods for
future scientific discoveries; and (4) developing scien-
tists for today and tomorrow.” The areas of research
emphasis include: (1) promoting health and prevent-
ing disease, (2) improving quality of life, (3) eliminat-
ing health disparities, and (4) setting directions for
end-of-life research (NINR, 2011). Specific research
priorities were identified for each of these four areas
of research emphasis and were included in the NINR
Strategic Plan. These research priorities provide
important information for nurses seeking funding
from the NINR. Details about the NINR mission,
strategic plan, and areas of funding are available
on its website at http://www.ninr.nih.gov/AboutNINR/
NINRMissionandStrategicPlan/.
Another federal agency that is funding healthcare
research is the Agency for Healthcare Research and
Quality (AHRQ). The purpose of the AHRQ is to
enhance the quality, appropriateness, and effective-
ness of healthcare services, and access to such ser-
vices, by establishing a broad base of scientific
research and promoting improvements in clinical
practice and in the organization, financing, and
delivery of healthcare services. Some of the current
AHRQ funding priorities are research focused on
prevention; health information technology; patient
safety; long-term care; pharmaceutical outcomes;
system capacity and emergency preparedness; and
the cost, organization, and socioeconomics of health
care. For a complete list of funding opportunities and
grant announcements, see the AHRQ website at http://
www.ahrq.gov/.
The World Health Organization (WHO) is encour-
aging the identification of priorities for a common
nursing research agenda among countries. A quality
healthcare delivery system and improved patient and
family health have become global goals. By 2020, the
world’s population is expected to increase by 94%,
with the elderly population growing by almost 240%.
Seven of every 10 deaths are expected to be caused
by noncommunicable diseases, such as chronic condi-
tions (heart disease, cancer, and depression) and inju-
ries (unintentional and intentional). The priority areas
for research identified by WHO are to (1) improve the
health of the world’s most marginalized populations,
(2) study new diseases that threaten public health
around the world, (3) conduct comparative analyses
of supply and demand of the health workforce of dif-
ferent countries, (4) analyze the feasibility, effective-
ness, and quality of education and practice of nurses,
(5) conduct research on healthcare delivery modes,
and (6) examine the outcomes for healthcare agencies,
providers, and patients around the world (WHO,
2012). A discussion of WHO’s mission, objectives,
and research policies can be found online at http://
www.who.int/rpc/en.
The Healthy People 2020 website identifies and
prioritizes health topics and objectives for all age
groups over the next decade (U.S. Department of
Health and Human Services, 2012). These health
topics and objectives direct future research in the areas
of health promotion, illness prevention, illness
management, and rehabilitation and can be accessed
 CHAPTER 5  Research Problem and Purpose 81

Observation of real world situations

(Nursing practice)

Identify research topics

Nonresearchable questions Generate questions

RESEARCH PROBLEM
Review of literature Input from others
Clarification and refinement

RESEARCH PURPOSE

Objectives, questions, or hypotheses

Figure 5-2  Formulating a research problem and purpose.

online at http://www.healthypeople.gov/2020/topics research problem. From the problem, the researcher

objectives2020/default.aspx/.
In summary, funding organizations, professional
organizations, and governmental healthcare organiza-
tions, both national and international, are sources
for identifying priority research problems and offer
opportunities for obtaining funding for future research.
Formulating a Research Problem
and Purpose
Potential nursing research problems often emerge
from real-world situations, such as those in nursing
practice. A situation is a significant combination of
circumstances that occur at a given time. Inexperi-
enced researchers tend to want to study the entire situ-
ation, but it is far too complex for a single study.
Multiple problems exist in a single situation, and each
can be developed into a study. A researcher’s percep-
tion of what problems exist in a situation depends on
that individual’s clinical expertise, theoretical base,
intuition, interests, and goals. Some researchers spend
years developing different problem statements and
new studies from the same clinical situation.
The exact thought processes used to extract prob-
lems from a situation have not been clearly identified
because of the abstractness and complexity of the rea-
soning involved. However, in formulating their study
problems, researchers often implement the following
steps: (1) examine a real-world situation, (2) identify
research topics, (3) generate questions, (4) review rel-
evant literature, and (5) ultimately clarify and refine a
develops a specific focus or research purpose for
study. The flow of these steps is presented in Figure
5-2 and described in the following sections.
Examining a Real-World Situation and
Identifying Research Topics
A nursing situation often includes a variety of research
topics or concepts that identify broad problem areas
requiring investigation. Nurses frequently investigate
patient- and family-related topics, such as stress, pain,
coping patterns, the teaching and learning process,
self-care deficits, health promotion, rehabilitation,
prevention of illness, disease management, and social
support. Other relevant research topics focus on the
healthcare system and providers, such as cost-effective
care; advanced practice nurse roles (nurse practitioner,
clinical nurse specialist, midwife, and nurse anesthe-
tist); managed care; and redesign of the healthcare
system. Outcomes research focuses on topics of health
status, quality of life, cost-effectiveness, and quality
of care. A specific outcome study might focus on a
particular condition, such as terminal cancer, and
examine outcomes, such as nutrition, hygiene, skin
integrity, and pain control with a variety of treatments
(Doran, 2011).
Generating Questions and Reviewing
the Literature
Situations encountered in nursing stimulate a constant
flow of questions. The questions fit into three catego-
ries: (1) questions answered by existing knowledge,
(2) questions answered with problem solving, and (3)
82 UNIT TWO  The Research Process
research-generating questions. The first two types of
questions are nonresearchable and do not facilitate the
formulation of research problems that will generate
knowledge for practice. Some of the questions raised
have a satisfactory answer within the nursing profes-
sion’s existing body of knowledge, and these answers
are available in the literature and online, from EBP
guidelines, or from experts in nursing or other disci-
plines. For example, suppose you have questions
about performing some basic nursing skills, such as a
protocol for taking a temperature or giving injections;
you can find answers to questions such as these in the
research literature and procedure manuals (see Figure
5-2). However, suppose your questions focus on
investigating new techniques to improve existing
skills, patient responses to techniques, or ways to
educate patients and families to perform techniques.
Your efforts to answer these types of questions could
add to knowledge needed for EBP.
Some of the questions raised can be answered using
problem-solving or evaluation projects. The problem-
solving process addresses a particular problem situa-
tion, and the goal of the research process is the
generation of knowledge to be generalized to other
similar situations. Many evaluation projects are con-
ducted with minimal application of the rigor and
control required with research. These projects do not
fit the criteria of research, and the findings are relevant
for a particular situation. For example, quality assur-
ance is an evaluation of the patient care implemented
by a specific healthcare agency; the results of this
evaluation project are usually relevant mainly to the
agency conducting the review.
The type of question that can initiate the research
process is one that requires further knowledge to
answer it. Some of the questions that come to mind
about situations include the following: Is there a need
to explore or describe concepts, to know how they are
related, or to be able to predict or control some event
within the situation? What is known and what is not
known about the concepts? What are the most urgent
factors or outcomes to know? Is there a need to gener-
ate or test theory in an area important to practice? Is
the patient’s perspective in this situation needed to
develop an intervention? Which intervention is most
effective in achieving quality patient outcomes?
Research experts have found that asking the right
question is frequently more valuable than finding the
solution to a problem. The solution identified in a
single study might not withstand the test of time or
might be useful in only a few situations. However, one
well-formulated question can generate numerous
research problems, direct a lifetime of research
activities, and significantly contribute to a discipline’s
body of knowledge.
Clarifying and Refining a Research Problem
Fantasy and creativity are part of formulating a
research problem, so you need to imagine prospective
studies related to the situation. You also need to
imagine the difficulties likely to occur with each study,
but avoid being too critical of potential research prob-
lems at this time. Which studies seem the most work-
able? Which ones appeal intuitively? Which problem
is the most significant to nursing? Which study is of
personal interest? Which problem has the greatest
potential to provide a foundation for further research
in the field (Fawcett & Garity, 2009)?
The problems investigated need to have profes-
sional significance and potential or actual significance
for society. A research problem is significant when it
has the potential to generate or refine knowledge to
build an EBP for nursing (Craig & Smyth, 2012;
Melnyk & Fineout-Overholt, 2011). Nurse researchers
believe that significant research problems need to
focus on real-world concerns, to be methodologically
sound, to build knowledge for nursing, to develop and/
or test theory, and to focus on current or timely con-
cerns (Alligood, 2010; Chinn & Kramer, 2008; Craig
& Smyth, 2012; Smith & Liehr, 2008). The problems
that are considered significant vary with time and the
needs of society. The priorities identified earlier indi-
cate some of the current, significant nursing research
topics and problems.
Personal interest in a problem influences the quality
of the problem formulated and the study conducted. A
problem of personal interest is one that an individual
has pondered for a long time or one that is especially
important in the individual’s nursing practice or per-
sonal life. For example, if you know someone who has
had a mastectomy, you may be particularly interested
in studying the emotional impact of a mastectomy or
strategies for caring for mastectomy patients. This per-
sonal interest in the topic can become the driving force
needed to conduct a quality study.
Answering these questions regarding significance
and personal interest can often assist you in narrowing
the number of problems. Without narrowing potential
problems to only one idea, try some of the ideas out
on colleagues (see Figure 5-2). Let them play the
devil’s advocate and explore the strengths and weak-
nesses of each idea. Then begin some preliminary
reading in the area of interest. Examine literature
related to the situation, the variables within the situa-
tion, measurement of the variables, previous studies
related to the situation, and supportive theories. The

literature review often enables you to refine the
problem and clearly identify the gap in the knowledge
base. Once you have identified the problem, you must
frame it or ground it in past research, practice, and
theory. The discussion of the problem must culminate
in a problem statement that identifies the gap in the
knowledge base that your proposed study will address.
Thus, the refined problem has documented signifi-
cance to nursing practice, is based on past research
and theory, and identifies a gap in nursing knowledge
that directs the development of the research purpose.
Research Purpose
The purpose is generated from the problem, identifies
the focus, aim, or goal of the study, and directs the
development of the study. In the research process, the
purpose is usually stated after the problem, because
the problem identifies the gap in knowledge in a
selected area and the purpose clarifies the knowledge
to be generated by the study. The research purpose
must be stated objectively, that is, in a way that does
not reflect particular biases or values of the researcher.
Investigators who do not recognize their values might
include their biases in the research. This situation can
lead them to generate the answers they want or
believe to be true and might add inaccurate informa-
tion to a discipline’s body of knowledge (Kaplan,
1964). Therefore, on the basis of your research
purpose, you can develop specific research objec-
tives, questions, or hypotheses to direct your study
(see Chapter 8).
Example of Problem and
Purpose Development
You might have observed the women receiving treat-
ment at a mental health clinic and noted that many
were withdrawn, depressed, and unable to discuss
certain events in their lives. Their progress in therapy
was usually slow, and they seemed to have similar
physical and psychological symptoms. Often, after
developing a rapport with a therapist, they would
reveal that they were victims of sexual abuse as a
child. This situation could lead you to identify research
topics and generate searching questions. Research
topics of interest include childhood sexual abuse,
history and nature of abuse, parental characteristics,
adult survivors of child sexual abuse, physical and
psychological symptoms in adulthood, assessment and
diagnosis of child sexual abuse history, therapeutic
interventions to manage abuse history and symptoms,
and psychological adjustment. Possible questions
CHAPTER 5  Research Problem and Purpose 83
include the following: What are the physical and psy-
chological symptoms demonstrated by someone who
has experienced childhood sexual abuse? How would
one assess the occurrence, frequency, and impact of
rape or incest on a woman? What influences do age,
duration, and nature of abuse have on a woman’s
current behavior and psychological adjustment? How
frequently is childhood sexual abuse a problem in the
mentally disturbed adult female? How does a health-
care provider assess, diagnose, and manage the emo-
tional problems of adult survivors of child sexual
abuse? These are the types of questions that Zinzow,
Seth, Jackson, Niehaus, and Fitzgerald (2010) might
have raised as they developed the following problem
and purpose for their study titled “Abuse and Parental
Characteristics, Attributions of Blame, and Psycho-
logical Adjustments in Adult Survivors of Child
Sexual Abuse”:
Research Problem
“A history of childhood sexual abuse (CSA) has been
consistently related to adult psychological symptom-
atology in women, including anxiety, depression,
posttraumatic stress, interpersonal difficulties, sexual
dysfunction, and somatization.… However, wide
variability among the presence and severity of nega-
tive outcomes has been observed. Factors that have
been found to account for some of this variation in
adult adjustment include abuse characteristics (e.g.,
severity, number of incidents) (…Steel, Sanna,
Hammond, Whipple, & Cross, 2004) and family char-
acteristics such as parenting style and parental psy-
chopathology.… Furthermore, an emerging body of
research has demonstrated a link between cognitive
mechanisms, such as attributions for abuse, and psy-
chological outcomes in abuse survivors (see review
by Valle & Silovsky, 2002).… However, little is known
about the influence of abuse and parental character-
istics on attributional content. Moreover, the relative
contribution of self-blame, family blame, and perpe-
trator blame to the sequelae of CSA is poorly under-
stood.” (Zinzow et al., 2010, p. 80)
Research Purpose
“The purpose of this study was to examine the influ-
ence of abuse and parental characteristics on attribu-
tional content and determine the relative contribution
of different attributions of blame in predicting psycho-
logical symptoms among adult survivors of childhood
sexual abuse.” (Zinzow et al., 2010, p. 79)
84 UNIT TWO  The Research Process
The research problem identified by Zinzow et al.
(2010) included significance because CSA was linked
to several psychological problems in adult women
who suffered abuse. The key findings from previous
research focused on the nature and duration of abuse,
parental characteristics, and attributions of blame.
However, the findings were varied regarding the psy-
chological adjustment and outcomes of women who
were adult survivors of CSA. The contributions of
self-blame, family blame, and perpetrator blame to
adult survivors’ adjustment and symptoms are poorly
understood. These gaps in the knowledge base pro-
vided a basis for the study purpose. The study purpose
clearly identifies the focus or aim of the study. Zinzow
et al. (2010) studied 83 female undergraduates with
a history of CSA. They found that abuse charac­
teristics such as severity and number of incidents
were significantly related to attributions. Family- and
perpetrator-blame accounted for significant variability
in psychological symptoms and adjustments of the
women in the sample, which were greater than the
contributions of abuse characteristics, family environ-
ment, and self-blame. The findings from this study
included implications for treatment of women who
were adult survivors of CSA and directions for further
research.
Feasibility of a Study
As the research problem and purpose increase in
clarity and conciseness, the researcher has greater
direction in determining the feasibility of a study. The
feasibility of a study is determined by examining the
time and money commitment; the researcher’s exper-
tise; availability of subjects, facility, and equipment;
cooperation of others; and the study’s ethical consid-
erations (Creswell, 2009; Fawcett & Garity, 2009;
Munhall, 2012; Rogers, 1987).
Time Commitment
Conducting research frequently takes longer than
anticipated, making it difficult for any researcher,
especially a novice, to estimate the time that will be
involved. In estimating the time commitment, the
researcher examines the purpose of the study; the
more complex the purpose, the greater the time com-
mitment. You can approximate the time needed to
complete a study by assessing the following factors:
(1) type and number of subjects needed, (2) number
and complexity of the variables to be studied, (3)
methods for measuring the variables (Are instruments
available to measure the variables or must they be
developed?), (4) methods for collecting data, and (5)
the data analysis process. Another factor that can
increase the time needed for a study is obtaining insti-
tutional review board (IRB) approval, especially if
more than one clinical agency is used for data collec-
tion in a study. Also, researchers often overlook the
time commitment necessary to write the research
report for presentation and publication. You must
approximate the time needed to complete each step of
the research process and determine whether the study
is feasible.
Most researchers propose a designated time or set
a specific deadline for their project. For example, an
agency might set a 2-year deadline for studying the
turnover rate of staff. The researcher must determine
whether the identified purpose can be accomplished
by the designated deadline; if not, the purpose could
be narrowed or the deadline extended. Researchers
are often cautious about extending deadlines because
a project could continue for many years. The indi-
vidual interested in conducting qualitative research
frequently must make an extensive time commitment
of 2 years or longer to allow for quality collection
and analysis of data (Marshall & Rossman, 2011).
Time is as important as money, and the cost of a
study can be greatly affected by the time required to
conduct it.
Money Commitment
The problem and purpose selected are influenced by
the amount of money available to the researchers.
Sources for nursing research funding include: (1) gov-
ernment funding from such offices as the NINR and
AHRQ; (2) professional organizations such as AACN,
AONE, and the Oncology Nursing Society; and (3)
local clinical agencies, corporations, and universities
(see Chapter 29). Potential sources for funding should
be considered at the time the problem and purpose are
identified. For example, Grady et al. (2011), who
studied the effects of educational messages on health-
related knowledge, attitudes, and behavior of persons
with diabetes, was partially funded by an Office of
Naval Research Award. Federal and private sources of
funding greatly strengthen the feasibility of conduct-
ing a research project.
The cost of a research project can range from a few
dollars for a student’s small study to hundreds of thou-
sands of dollars for complex projects, such as multisite
clinical trials and major qualitative studies. In estimat-
ing the cost of a research project, the following ques-
tions need to be considered, in addition to other areas
of expense based on the study being conducted:

Literature: What will the review of the literature—
including computer searches, copying articles,
and purchasing books—cost?
Subjects: How many subjects or study participants
will need to be recruited for the study, and will
the subjects have to be paid for their participa-
tion in the project? Grady et al. (2011) paid their
approximately 155 study participants $25 at
each of the three data collection periods. This
resulted in an $11,625 expense that would
require funding to accomplish.
Equipment: What will the equipment for the study
cost? Can the equipment be borrowed, rented,
bought, or obtained through donation? Is the
equipment available, or will it need to be built?
What type of maintenance will be required for
the equipment during the study? What will the
measurement instruments cost?
Personnel: Will assistants or consultants, or both,
be hired to collect, computerize, and analyze the
data and assist with the data interpretation? Will
clerical help be needed to type and distribute the
report and prepare a manuscript for publication?
Computer time: Will computer time be required to
analyze the data? If so, what will be the cost?
Transportation: What will be the transportation
costs for conducting the study and presenting
the findings?
Supplies: Will any supplies—such as envelopes,
postage, pens, paper, and photocopies—be
needed? Will a cell phone be needed to contact
the researcher about potential subjects? Will a
survey or reminder postcard by mailed to par-
ticipants? Will long-distance phone calls or
overnight mailing be needed?
Researcher Expertise
A research problem and purpose must be selected on
the basis of the ability of the investigator(s). Initially,
you might work with another researcher (mentor) to
learn the process and then investigate a familiar
problem that fits your knowledge base or experience.
Selecting a difficult, complex problem and purpose
can only frustrate and confuse the novice researcher.
However, all researchers need to identify problems
and purposes that are challenging and collaborate with
other researchers as necessary to build their research
background.
When a team of researchers conducts a study, the
team members often have a variety of research and
clinical experiences that add to the quality of the study
conducted. In the study conducted by Grady et al.
CHAPTER 5  Research Problem and Purpose 85
(2011), these investigators had research and clinical
expertise in nursing, public health, psychology, and
biostatistics. The researchers are all doctorally pre-
pared, and Grady is a seasoned faculty member for the
University of Pittsburgh. Eileen and Elliot Entin are
both senior research psychologists with Aptima, Inc.;
and Brunye is a cognitive psychologist for the U.S.
Army. The credentials and employment sites for the
investigators are identified under the title of the study
article. These researchers all appear to have strong
backgrounds for conducting research in the discipline
of nursing, psychology, and health care. You can
obtain more information about the authors by search-
ing their names online. The researchers also acknowl-
edged the support of the following in conducting their
study: registered nurses, Conemaugh Diabetes Insti-
tute, Flipside Media, Inc., and Telehealth Department,
Mount Aloysius College.
Availability of Subjects or Participants
In selecting a research purpose, you must consider the
type and number of study participants needed. Finding
a sample might be difficult if the study involves inves-
tigating a unique or rare population, such as quadriple-
gic individuals who live alone and are currently
attending college. The more specific the population
selected for study, the more difficult it is to find sub-
jects. In addition, the Health Insurance Portability and
Accountability Act (HIPAA) prevents clinical agencies
from sharing lists of potential subjects with a researcher
without specific stipulations (see Chapter 9). Potential
subjects who are stigmatized, such as persons with
HIV/AIDS, may be more difficult to access.
The money and time available to the researcher will
affect the subjects selected. With limited time and
money, the researcher might want to investigate sub-
jects who are accessible and do not require payment
for participation. Even if you identify a population
with a large number of potential subjects, those indi-
viduals may be unwilling to participate in the study
because of the topic selected. For example, nurses
could be asked to share their experiences with alcohol
and drug use, but many might fear that sharing this
information would jeopardize their jobs and licenses.
Researchers need to be prepared to pursue the attain-
ment of study participants at whatever depth is neces-
sary. Having a representative sample of reasonable
size is critical for generating quality research findings
(Aberson, 2010). Grady et al. (2011) selected a setting
where they could obtain the sample size that they
needed for their study, as identified in the following
quotation:
86 UNIT TWO  The Research Process
“The study was conducted at the diabetes outpatient
facility of an acute care hospital in west central Penn-
sylvania. After receiving institutional review board
approval from the participating facility, participants
were obtained through advertisements in local news-
papers recruiting people older than 18 years, who
have had diabetes for 5 years or more, and who have
no known foot problems.… The final sample was
composed of 64 men and 91 women.” (Grady et al.,
2011, p. 24)
Availability of Facilities and Equipment
Researchers need to determine whether their studies
will require special facilities to implement. Will a
special room be needed for an educational program,
interview, or observations? If the study is conducted
at a hospital, clinic, or college of nursing, will the
agency provide the facilities that are needed? Setting
up a highly specialized laboratory for the conduct of
a study would be expensive and would probably
require external funding. Most nursing studies are
done in natural settings such as a hospital room or unit,
a clinic, or a patient’s home. Grady et al. (2011) con-
ducted all their research activities in the diabetes out-
patient facility (partially controlled setting).
Nursing studies frequently require a limited amount
of equipment, such as a tape or video recorder for
interviews or a physiological instrument, such as a
scale or thermometer. Often you can borrow equip-
ment from the facility where the study is conducted, or
you can rent it. Some companies are willing to donate
equipment if the study focuses on determining the
effectiveness of the equipment and the findings are
shared with the company. If specialized facilities or
equipment are required for a study, you must be aware
of the options available before actively pursuing the
study. Grady et al. (2011) delivered their educational
messages (either gain-framed or loss-framed) to each
subject individually in the diabetes outpatient facility:
“Videos were watched from digital video disk (DVD)
on a 19-inch television monitor. The inventories were
presented one at a time on a personal computer, via
ScoreMD Web-based assessment software. Partici-
pants controlled the pacing of the questions in the
inventories; responses were automatically coded and
stored in a database.… Participants were paid $25
at each of the three data collection periods.” (Grady
et al., 2011, p. 25)
The researchers did not indicate how they obtained
the equipment for their study, making it difficult to
determine the expenses related to the study. They did
identify the software that was used for data collection
and analysis but not the expenses related to using this
software. You could contact the principal investiga-
tor, Grady, and obtain more details related to equip-
ment costs and computer expenses for data collection
and analysis. The contact information for Grady and
E. B. Entin is provided on the first page of the study
article.
Cooperation of Others
A study might appear feasible, but without the coop-
eration of others, it is not. Some studies are conducted
in laboratory settings and require the minimal coop-
eration of others. However, most nursing studies
involve human subjects and are conducted in hospi-
tals, clinics, schools, offices, or homes. Having the
cooperation of people in the research setting, the sub-
jects, and the research assistants involved in data col-
lection is essential. People are frequently willing to
cooperate with a study if they view the problem and
purpose as significant or if they are personally inter-
ested. Grady et al. (2011) acknowledged the support
of nurses and other professional organizations and
corporations that assisted with the study. The research-
ers seemed to have strong agency support and the
support of relevant businesses. Having the coopera-
tion of others can improve the subject participation
and promote the successful completion of the study
(see Chapter 20 for details on the data collection
process).
Ethical Considerations
The purpose selected for investigation must be
ethical, which means that the participants’ rights and
the rights of others in the setting are protected. If
your purpose appears to infringe on the rights of the
participants, you should reexamine that purpose; the
investigation may have to be revised or abandoned.
There are usually some risks in every study, but the
value of the knowledge generated should outweigh
the risks. Grady et al. (2011, p. 24) received insti­
tutional review board approval from the diabetes
outpatient facility in Pennsylvania where the study
was conducted, and “all eligible adults wishing to
participate in the study provided informed consent.”
By taking these steps, the researchers attempted
to implement an ethical study that protected the
rights of the adults with diabetes who participated
(see Chapter 9 for details on ethical conduct in
research).

Example Research Topics,
Problems, and Purposes for
Different Types of Research
Quantitative Research
Quantitative and qualitative research approaches
enable nurses to investigate a variety of research prob-
lems and purposes. Examples of research topics, prob-
lems, and purposes for some of the different types of
quantitative studies are presented in Table 5-1. The
research purpose usually reflects the type of study
that is to be conducted. The purposes of descriptive
research are to identify patterns of variables, to
describe and define variables, to identify initial links
among variables, and to compare and contrast groups
on selected variables (Kerlinger & Lee, 2000). For
example, Trotter, Gallagher, and Donoghue (2011)
conducted their study to describe the patterns of anxiety
and concerns of patients undergoing percutaneous
coronary interventions (PCIs). The research topics,
problems, and purposes for this study are presented in
Table 5-1. Trotter et al. (2011) found that symptoms
of anxiety were common before and after PCIs. They
recommended that patients who had chest pain or were
undergoing a first PCI should be targeted for an inter-
vention to reduce anxiety during the recovery period
and after discharge.
The purpose of correlational research is to examine
the type (positive or negative) and strength of the
relationships among study variables. In their correla-
tional study, Houck, Kendall, Miller, Morrell, and
Wiebe (2011) examined the relationship between
behavior problems and self-concept in adolescents
and children with attention deficit hyperactivity disor-
der (ADHD) (see Table 5-1). Houck et al. (2011,
p. 239) found that it is important “to assess self-
concept in children and adolescents with ADHD,
especially those who are older and have comorbid
conditions.” These individuals with ADHD who have
low self-concept require diagnosis and treatment as
part of their care.
Quasi-experimental studies are conducted to deter-
mine the effect of an intervention or independent vari-
able on designated dependent or outcome variables
(Shadish et al., 2002). For example, Nyamathi et al.
(2009) conducted a quasi-experimental study to
examine the effects of a nurse-managed program
(intervention) on the vaccine completion rates for
hepatitis A and B vaccine series (outcomes) in a popu-
lation of homeless adults. The research topics,
problem, and purpose are identified in Table 5-1.
CHAPTER 5  Research Problem and Purpose 87
Nyamathi and colleagues found that a program that
used nurse case management and tracking was essen-
tial in supporting the adherence and completion of
6-month hepatitis A and B vaccine series. This was
especially important for white homeless persons, who
were the least likely to complete their vaccine series
without an intervention tailored to address their unique
needs.
Experimental studies are conducted in highly con-
trolled settings and under highly controlled conditions
to determine the effect of one or more independent
variables on one or more dependent variables (Shadish
et al., 2002). Sharma, Ryals, Gajewski, and Wright
(2010) conducted an experimental study to determine
the effects of an aerobic exercise program on the pain-
like behavior and neurotrophin-3 (NT-3) in mice with
chronic widespread pain (see Table 5-1). These
researchers found that moderate-intensity aerobic
exercise had the effect of deep tissue mechanical
hyperalgesia on chronic pain in mice. This finding
provides a possible molecular basis for aerobic exer-
cise training in reducing muscular pain.
Qualitative Research
The problems formulated for qualitative research iden-
tify an area of concern that requires investigation. The
purpose of a qualitative study indicates the focus of
the study and whether it is a subjective concept, an
event, a phenomenon, experience, situation, or a facet
of a culture or society (Marshall & Rossman, 2011;
Munhall, 2012). Examples of research topics, prob-
lems, and purposes from some different types of quali-
tative studies are presented in Table 5-2 and in Chapter
12. Phenomenological research seeks an understand-
ing of human experience from the researcher’s per-
spective, such as children’s experiences of living with
asthma conducted by Trollvik, Nordbach, Silen, and
Ringsberg (2011). The research topics, problem, and
purpose for this study are presented in Table 5-2. The
findings from the study by Trollvik et al. (2011,
p. 295) “described two themes with five subthemes:
fear of exacerbation (body sensations, frightening
experiences, and loss of control) and fear of being
ostracized (experiences of being excluded and dilemma
of keeping the asthma secret or being open about it).”
In grounded theory research, the problem identifies
the area of concern and the purpose indicates the focus
of the theory to be developed from the research
(Munhall, 2012). For example, El-Mallakh (2007)
investigated the poverty and self-care among individu-
als with schizophrenia and diabetes mellitus. On the
basis of findings from this grounded theory study,
Text continued on p. 92
 88

TABLE 5-1  Quantitative Research: Topics, Problems, and Purposes

Type of
UNIT TWO  The Research Process

Research Research Topic Research Problem and Purpose

Descriptive Anxiety, concerns, Title of study: “Care of patients with coronary heart disease: Anxiety in patients undergoing percutaneous coronary
research percutaneous interventions” (Trotter, Gallagher, & Donoghue, 2011, p. 185).
coronary Problem: “Coronary artery disease (CAD) is a leading cause of mortality, morbidity, and loss of quality of life globally. One
intervention, of the most common treatments for CAD is percutaneous transluminal coronary angioplasty/stent placement, collectively
coronary heart labeled percutaneous coronary intervention (PCI).… Although PCIs are common and relatively low risk, many patients
disease undergoing these treatments experience clinically relevant anxiety, with an estimated prevalence rate of 24% to 72%.…
Although anxiety levels decrease after a PCI, clinically relevant anxiety may still be common, with one study noting that
21% of patients remained anxious 6 to 8 weeks after the procedure (Astin, Jones, & Thompson, 2005).… Few studies have
investigated anxiety levels and patients’ expressed concerns in the very early recovery period within 24 hours after the
procedure or the early recovery period in the week after discharge” (Trotter et al., 2011, p. 186).
Purpose: The purpose of this study was “to determine the patterns of anxiety and concerns experienced by patients under
going PCI…” (Trotter et al., 2011, 185).
Correlational Self-concept, attention Title of study: “Self-concept in children and adolescents with attention deficit hyperactivity disorder” (Houck, Kendall, Miller,
research deficit hyperactivity Morrell, & Wiebe, 2011, p. 239).
disorder, behavior Problem: “Attention deficit hyperactivity disorder (ADHD) is the most common mental health disorder of childhood, affecting
problems approximately 3%-8.7% of children and adolescents in the United States (National Institutes of Mental Health, 2008).
ADHD is a chronic and stigmatizing neurological disorder with deficits in the neurotransmitter systems that affect executive
functioning.… People with ADHD have impairments in adaptive functioning, which is often manifested in difficult
behaviors, such as aggression, poor rule-regulated behavior, inability to delay gratification, behavioral disinhibition, learning
difficulties, poor impulse control and low motivation.… Findings from previous studies on the relationships between ADHD
and self-concept are mixed, with some studies indicating that self-concept scores are higher in children with ADHD than in
those without ADHD (Treuting & Hinshaw, 2001).… Understanding more specifically how self-concept and behavioral
problems are related, given the behavioral disruption that accompanies ADHD, is important to support a child’s social and
emotional development” (Houck et al., 2011, pp. 239-240).
Purpose: “The purpose of this study was to examine the relationship between behavioral problems and self-concept in children
and adolescents with ADHD” (Houck et al., 2011, p. 241).
Type of
Research Research Topic Research Problem and Purpose

Quasi- Nurse-case-managed Title of study: “Effects of a nurse-managed program on hepatitis A and B vaccine completion among homeless adults”
experimental intervention, hepatitis (Nyamathi et al., 2009, p. 13).
research A and B vaccine Problem: “Hepatitis B virus (HBV) infection poses a serious threat to public health in the United States. Recent estimates
completion rate, place the true prevalence of chronic HBV in the United States at approximately 1.6 cases per 100,000 persons (Centers for
socio-demographic Disease Control and Preventions [CDC], 2008b). It is estimated that there were 51,000 new cases of HBV infection in
factors, risk 2005, a financial burden reaching $1 billion annually (Cohen et al., 2007).… Homeless populations are at particularly high
behaviors, homeless risk of HBV infection due to high rates of unprotected sexual behavior and sharing of needles and other IDU [injection drug
user] paraphernalia. Previous studies have reported that HBV infection rates among homeless populations range from 17%
to 31% (i.e., from 17,000 to 31,000 per 100,000…) compared with 2.1 per 100,000 in the general United States
population.… Vaccination is the most effective way to prevent HBV infection (CDC, 2006).… Improving vaccination
adherence rates among homeless persons is an important step toward reducing the high prevalence of HBV infection in this
population.… Thus, little is known about adherence to HBV vaccination among community samples of urban homeless
persons or about the effect of stronger interventions to incorporate additional strategies, such as nurse case management and
targeted HBV education along with client tracking” (Nyamathi et al., 2009, pp. 13-14.)
Purpose: The purpose of this study was to determine the “effectiveness of a nurse-case-managed intervention compared with
that of two standard programs on completion of the combined hepatitis A virus (HAV) and HBV vaccine series among
homeless adults and to assess socio-demographic factors and risk behaviors related to the vaccine completion” (Nyamathi
et al., 2009, p. 13).
Experimental Chronic widespread Title of study: “Aerobic exercise alters analgesia and neurotrophin-3 [NT-3] synthesis in an animal model of chronic
research pain, aerobic widespread pain” (Sharma, Ryals, Gajewski, & Wright, 2010, p. 714).
exercise, analgesia, Problem: “Chronic widespread pain is complex and poorly understood and affects about 12% of the adult population in
neurotrophin-3 developed countries (Rohrbeck, Jordan, & Croft, 2007).… Management of chronic pain syndromes poses challenges for
synthesis, pain healthcare practitioners, and pharmacological interventions offer limited efficacy.… Exercise training has been long
management, animal suggested to reduce pain and improve functional outcomes (Whiteside, Hansen, & Chaudhuri, 2004).… Surprisingly, the
model current literature is mainly limited to human studies where the molecular basis for exercise training cannot be easily
determined. Relatively few animal studies have addressed the effects and mechanisms of exercise on sensory modulation of
chronic pain” (Sharma et al., 2010, p. 715).
Purpose: “The purpose of the present study was to examine the effects of moderate-intensity aerobic exercise on pain-like
behavior and NT-3 in an animal model of widespread pain” (Sharma et al., 2010, p. 714).
CHAPTER 5  Research Problem and Purpose
89
 90

TABLE 5-2  Qualitative Research: Topics, Problems, and Purposes

Type of
Research Research Topic Research Problem and Purpose
Phenomenological Lived experience of Title of study: “Children’s experiences of living with asthma: Fear of exacerbations and being ostracized” (Trollvik,
research children, asthma, Nordbach, Silen, & Ringsberg, 2011, p. 295).
health promotion, Problem: “Asthma is the most common childhood disease and long-term medical condition affecting children (Masoli,
child health, chronic Fabian, Holt, Beasley, & Global Initiative for Asthma [GINA] Program, 2004). The prevalence of asthma is increasing,
illness, fears of and atopic diseases are considered to be a worldwide health problem and an agent of morbidity in children.… Studies
exacerbations, fears show that children with asthma have more emotional/behavioral problems than healthy children.… It has also been found
of being ostracized that asthma control in children is poor and that healthcare professionals (HCPs) and children focus on different aspects of
having asthma (Price et al., 2002).… Few studies have considered very young children’s, 7-10 years old, perspectives;
this study might contribute to new insights into their lifeworld experiences” (Trollvik et al., 2011, pp. 295-296).
UNIT TWO  The Research Process

Purpose: “The aim of this study was to explore and describe children’s everyday experiences of living with asthma to tailor
an Asthma Education Program based on their perspectives.… In this study, a phenomenological and hermaneutical
approach was used to gain an understanding of the children’s lifeworld” (Trollvik et al., 2011, p. 296).
Grounded theory Self-care, poverty, Title of study: “Doing my best: Poverty and self-care among individuals with schizophrenia and diabetes mellitus”
research schizophrenia, (El-Mallakh, 2007, p. 49).
diabetes mellitus Problem: “Mental health clinicians and researchers increasingly recognize that individuals with schizophrenia have a high
risk of developing diabetes mellitus (DM) (Bushe & Holt, 2004).… Whereas rates of diabetes in the general populations
range from 2% to 6%, prevalence rates of diabetes among individuals with schizophrenia range from 15% to 18%, and up
to 30% have impaired glucose tolerance (Bushe & Holt, 2004; Schizophrenia and Diabetes Expert Consensus Group,
2004).… The recent mental health literature has focused on the screening, diagnosis, and treatment of diabetes in this
population, including discussions of the risks and benefits of atypical antipsychotic use.… However, few researchers have
investigated the influence of social and demographic characteristics on diabetic self-care among individuals with
schizophrenia and diabetes” (El-Mallakh, 2007, pp. 49-50).
Purpose: “A grounded theory study was conducted to examine several aspects of diabetic self-care in individuals with
schizophrenia and DM” (El-Mallakh, 2007, p. 50).
Ethnography Critical illness, Title of study: “Family presence and surveillance during weaning from prolonged mechanical ventilation” (Happ et al., 2007,
research mechanical p. 47).
ventilation, weaning, Problem: “During critical illness, mechanical ventilation imposes physical and communication barriers between family
family presence members and their critically ill loved ones.… Most studies of family members in the intensive care unit (ICU) have
focused on families’ needs for information, access to the patient, and participation in decisions to withdraw or withhold
life-sustaining treatment.… Although numerous studies have been conducted of patient experiences with short- and
long-term mechanical ventilation (LTMV), research has not focused on family interactions with patients during weaning
from mechanical ventilation. Moreover, the importance of family members’ bedside presence and clinicians’ interpretation
of family behaviors at the bedside have not been critically examined” (Happ et al., 2007, pp. 47-48).
Purpose: “With the use of data from an ethnographic study of the care and communication processes during weaning from
LTMV, we sought to describe how family members interact with the patients and respond to the ventilator and associated
ICU bedside equipment during LTMV weaning” (Happ et al., 2007, p. 48).
Type of
Research Research Topic Research Problem and Purpose

Exploratory- Intimate partner Title of study: “Supporting mothering: Service providers’ perspectives of mothers and young children affected by intimate
descriptive violence, abuse of partner violence” (Letourneau et al., 2011, p. 192).
qualitative spouse, supporting Problem: “Estimates of the percent of women with exposure to intimate partner violence (IPV) over their lifetimes by
research mothering, husbands, partners, or boyfriends range between 8% and 66%.… The high concentration of preschool-age children in
parent-child households where women experience IPV… is a major concern.… Indeed, preschool-age children exposed to IPV may
relationships, family share many of the adjustment difficulties experienced by victims of direct physical and psychological abuse (Litrownik,
health, providers’ Newton, Hunter, English, & Everson, 2003). The degree to which children from birth to 36 months of age are affected by
perspective, social IPV, however, is not well understood. Even less is known of effective services and supports that target mothers and their
support young children exposed to IPV” (Letourneau et al., 2011, p. 193).
Purpose: “We conducted a qualitative descriptive study of service providers’ understandings of the impact of IPV on
mothers, young children (birth to 36 months), and mother-infant/child relationships, and of the support needs of these
mothers and young children” (Letourneau et al., 2011, p. 192).
Historical History, Cold War, Title of study: “Planning for mass disaster in the 1950s: Harriet H. Werley and nursing research” (Leifer & Glass, 2008, p.
research mass disaster 237).
preparation, nursing Problem: “Americans were continually aware of the potential for nuclear disaster during the Cold War era.… Because the
research, 1950s, fear of nuclear war was ever present, military and civil defense programs were developed to help Americans prepare for
Harriet H. Werley, disaster.… Military and civilian healthcare personnel were mobilized to prepare for any mass casualties caused by a
Army Nurse Corps nuclear attack.
“In the 21st century, world events remind nurses of the need to be prepared to respond to disaster. Since the events of
September 11, 2001; the Southeast Asian tsunami in 2004; and the Gulf Coast hurricanes in 2005, there has been an
increased emphasis on preparedness and response planning for man-made or natural disasters.… During the turbulent
Cold War era, Harriet H. Werley, an Army Nurse Corps (ANC) major, was a pioneer in mass disaster education and
nursing research. She served as the first nursing consultant in the newly formed Department of Atomic Casualties Studies
(DACS) from 1955-1958.… When working with military officials in Washington, DC, including the ANC, the Office of
the Surgeon General, the Army Institute of Research, and the Walter Reed Army Hospital, Werley shared her vision of an
evolving role for nurses that included increased opportunities for leadership, research, and expanded practice.… Primary
and secondary sources regarding Werley’s work in the DACS and the field of disaster nursing were examined to obtain
data for this historical study” (Leifer & Glass, 2008, pp. 237-238).
Purpose: The purpose of this historical study “was to analyze nurses’ involvement in research and mass disaster preparations
during the Cold War era and to describe the role of Harriet H. Werley and the Army Nurse Corps” (Leifer & Glass, 2008,
p. 237).
CHAPTER 5  Research Problem and Purpose
91
92 UNIT TWO  The Research Process

El-Mallakh (2007, p. 49) developed a “model, Evolv-
ing Self-Care, that describes the process by which
respondents developed health beliefs about self-care
of dual illnesses. One subcategory of the model,
‘Doing My Best,’ was further analyzed to examine the
social context of respondents’ diabetic self-care.”
In ethnographic research, the problem and purpose
identify the culture and the specific attributes of the
culture to be examined, described, analyzed, and inter-
preted. Happ et al. (2007) conducted an ethnographic
study of family presence and surveillance during
weaning of their family member from a ventilator (see
Table 5-2). These researchers concluded that “this
study provided a potentially useful conceptual frame-
work of family behaviors with long-term critically ill
patients that could enhance the dialogue about family-
centered care and guide future research on family
presence in the intensive care unit” (Happ et al., 2007,
p. 47).
Exploratory-descriptive qualitative research is
being conducted by several qualitative researchers to
describe unique issues, health problems, or situations
that lack clear description or definition. This type of
research often provides the basis for future qualitative
and quantitative research (Creswell, 2009). Letour-
neau et al. (2011) conducted an exploratory-descriptive
qualitative study of service provider’s understandings
of the impact of intimate partner violence (IPV) on the
mothers and their young children and to determine
their needs for support (see Table 5-2). These research-
ers found that mothers experiencing such violence and
their children require more support than is currently
available. In addition, the service providers had diffi-
culty in identifying interventions to promote and
protect these mothers and their children.
The problem and purpose in historical research
focus on a specific individual, a characteristic of
society, an event, or a situation in the past and iden-
tify the period in the past that will be examined. For
example, Leifer and Glass (2008) conducted a his-
torical study of nurses’ involvement in mass disaster
preparations and research during the Cold War era in
the 1950s (see Table 5-2). These researchers focused
on the career of Harriet Werley and how her plan-
ning for mass disaster during the 1950s increased
her emphasis on research and interdisciplinary
collaboration. Werley’s vision provides insights for
today’s nurses, who are once again faced with the
challenges and demands of disaster management
preparation.
Outcomes Research
Outcomes research is conducted to examine the end
results of care (Doran, 2011). Table 5-3 summarizes
the topics, problem, and purpose from an outcomes
study by Bae, Mark, and Fried (2010). These research-
ers examined the impact of nursing unit turnover rates
on patient outcomes in hospitals. They found that

TABLE 5-3  Outcomes Research: Topics, Problem, and Purpose

Type of
Research Research Topic Research Problem and Purpose
Outcomes Nurse turnover, Title of study: “Impact of nursing unit turnover on patient outcomes in hospitals” (Bae,
research workgroup Mark, & Fried, 2010, p. 40)
processes, patient Problem: “The adverse impact of nursing turnover on quality of patient care is a long-
outcomes, quality of standing assumption, yet there is little understanding of the turnover-quality relationship
care, patient safety, or its underlying mechanisms. When turnover occurs, the remaining staff must adjust to
patient satisfaction newcomers, and turnover may affect the interaction and integration among staff members
who remain.… Most empirical research on nursing turnover has focused on a direct
relationship between turnover and patient outcomes; the underlying mechanisms of the
turnover-outcomes relationship have not been explored (Alexander, Bloom, & Nuchols,
1994).… In order to understand the mechanisms by which nursing turnover is related to
patient outcomes, it is necessary to explore the impact of nursing turnover on nursing
units, which is the proximal context for individuals and a bounded interactive context
created by nurses’ attributes, interactions, and responses (Kozlowski, Steve, & Bell,
2003)” (Bae et al., 2010, pp. 40-41).
Purpose: “The aim of this study was to examine how nursing unit turnover affects key
workgroup processes and how these processes mediate the impact of nursing turnover on
patient outcomes” (Bae et al., 2010, p. 40).
 CHAPTER 5  Research Problem and Purpose 93

TABLE 5-4  Intervention Research: Topics, Problem, and Purpose

Type of
Research Research Topic
Intervention Stress management, social
research support, nursing
interventions, HIV/AIDS,
psychoneuroimmunology
(PNI), quality of life,
coping, psychosocial
functioning, immune status,
somatic health, viral load
Research Problem and Purpose
Title of study: “Effects of stress management on PNI-based outcomes in persons
with HIV disease” (McCain et al., 2003, p. 102).
Problem: “Although it remains potentially fatal, infection with the human
immunodeficiency virus (HIV) has become eminently more treatable as a chronic
illness with the advent of highly active antiretroviral therapies.… Insights as to
the relationship of psychological and physiological health in HIV and other
disease are emanating from research in psychoneuroimmunology (PNI).… A
growing body of research with persons who have HIV disease, as well as those
who have other chronic and potentially fatal illnesses such as cancer, indicates
that not only can a variety of biobehavioral strategies for stress management
mitigate psychological distress and improve coping skills, they also can enhance
immune function through neuroendocrine–immune system modulation.… More
recent work has continued to support the use of CBSM [cognitive-behavioral
stress management] as an effective strategy in the management of distress
associated with HIV disease.… Little comparative research has been done to
determine the relative effect of these two types of interventions on either
psychological or physiological status” (McCain et al., 2003, pp. 102-105).
Purpose: “This study was undertaken to compare the effects of CBSM groups,
social support groups (SSG), and a wait-listed control group on the outcomes of
psychosocial functioning (perceived stress, coping patterns, social support,
uncertainty, psychological distress), quality of life, neuroendocrine mediation
(salivary cortisol, DHEA levels), and somatic health (disease progression,
HIV-specific health status, viral load, and immune status)” (McCain et al., 2003,
p. 105).

nursing unit turnover had a significant adverse affect
on the workgroup processes on the unit where turn-
over took place. The negative effects on workgroup
processes adversely impact continuity and quality of
patient care and patient outcomes.
variables, which are commonly used to determine the
health status of patients with HIV. The CBSM inter-
vention was found to be the most effective in produc-
ing positive physical and psychological outcomes for
patients with HIV infection.

Intervention Research
Intervention research determines the interventions that
are most effective in managing clinical problems.
Some interventions might focus on risk reduction, pre-
vention, treatment, or resolution of health-related
problems or symptoms; management of a problem or
symptom; or prevention of complications associated
with a practice problem. In intervention research, the
interventions might have more than one purpose
and multiple outcomes. For example, McCain et al.
(2003) examined the effectiveness of two complex
interventions, cognitive-behavioral stress manage-
ment (CBSM) and social support, on the multiple out-
comes of patients with HIV infection. Table 5-4 lists
the topics, problem, and purpose from the study by
McCain et al. (2003). The outcomes measured in this
study were many physiological and psychological
KEY POINTS
• A research problem is an area of concern where
there is a gap in the knowledge base needed for
nursing practice and includes significance, back-
ground, and problem statement.
• The major sources for nursing research problems
include nursing practice; researcher and peer
interactions; literature review; theories; and re-
search priorities identified by individuals, specialty
groups, professional organizations, and funding
agencies.
• Replication is essential for the development of
evidence-based knowledge for practice and con-
sists of four types: exact, approximate, concurrent,
and systematic.
94 UNIT TWO  The Research Process
• The research purpose is a concise, clear statement
of the specific focus or aim of the study and
usually indicates the type of study (quantitative,
qualitative, outcomes, or intervention research) to
be conducted.
• The researcher examines the real-world situation,
identifies research topics, generates questions,
and ultimately clarifies and refines a research
problem.
• From the problem, a specific aim or research
purpose is developed that provides a clear focus for
the study.
• On the basis of the research purpose, specific
research objectives, questions, or hypotheses are
developed to direct the study.
• The feasibility of the research problem and purpose
are determined by examination of the time and
money commitments; researchers’ expertise; avail-
ability of subjects, facility, and equipment; coopera-
tion of others; and the study’s ethical considerations.
• Quantitative, qualitative, outcomes, and interven-
tion studies enable nurses to investigate a variety
of research problems and purposes.
REFERENCES
Aberson, C. L. (2010). Applied power analysis for the behavioral
sciences. New York, NY: Routledge Taylor & Francis Group.
Alexander, J. A., Bloom, J. R., & Nuchols, B. A. (1994). Nursing
turnover and hospital efficiency: An organization-level analysis.
Industrial Relations, 33(4), 505.
Alligood, M. R. (2010). Nursing theory: Utilization & application.
Maryland Heights, MO: Mosby Elsevier.
American Diabetes Association. (2009). Diabetes statistics.
Retrieved from http://www.diabetes.org/diabetes-statistics.jsp/.
American Association of Critical-Care Nurses (AACN). (2011).
AACN’s identified research needs. Retrieved from http://
www.aacn.org/WD/practice/docs/research/identified-research-
needs.pdf.
American Nurses Credentialing Center (ANCC). (2012). Magnet©
Program: Overview. Retrieved from http://www.nursecredential
ing.org/Magnet/ProgramOverview.aspx/.
American Organization of Nurse Executives (AONE). (2012).
AONE 2012 education and research priorities. Retrieved from
http://www.aone.org/education/index.shtml/.
Astin, F., Jones, K., & Thompson, D. R. (2005). Prevalence and
patterns of anxiety and depression in patients undergoing elective
percutaneous transluminal coronary angioplasty. Heart & Lung,
34(6), 393–401.
Bae, S., Mark, B., & Fried, B. (2010). Impact of nursing unit turn-
over on patient outcomes in hospitals. Journal of Nursing Schol-
arship, 42(1), 40–49.
Baranowski, T., & Simons-Morton, B. (1990). A center-based
program for exercise change among Black-Americans. Health
Education Quarterly, 17(3), 179–186.
Beveridge, W. I. B. (1950). The art of scientific investigation. New
York, NY: Vintage.
Boren, S. A., Gunlock, T. L., Santosh, K., & Kramer, T. C. (2006).
Computer-aided diabetes education: A synthesis of randomized
controlled trials. American Medical Informatics Association
Annual Symposium Proceedings (pp. 51–55).
Brink, P. J., & Wood, M. J. (1979). Multiple concurrent replications.
Western Journal of Nursing Research, 1(1), 117–118.
Brown, S. J. (2009). Evidence-based nursing: The research-practice
connection. Sudbury, MA: Jones and Bartlett Publishers.
Bushe, C., & Holt, R. (2004). Prevalence of diabetes and impaired
glucose tolerance in patients with schizophrenia. British Journal
of Psychiatry, 184(Suppl 47), S67–S71.
Centers for Disease Control and Prevention. (2006). A comprehen-
sive immunization strategy to eliminate transmission of hepatitis
B virus infection in the United States. Morbidity and Mortality
Weekly Report, 55(RR-16), 1–25.
Centers for Disease Control and Prevention. (2008a). National dia-
betes fact sheet: General information and national estimates on
diabetes in the United States, 2007. Atlanta, GA: U.S. Department
of Health and Human Services, Author.
Centers for Disease Control and Prevention. (2008b). Surveillance
for acute viral hepatitis—United States, 2006. Morbidity and
Mortality Weekly Report, 57(SS02), 1–24.
Chinn, P. L., & Kramer, M. K. (2008). Integrated theory and knowl-
edge development (7th ed.). St. Louis, MO: Mosby.
Cohen, C. A., London, W. T., Evans, A. A., Block, J., Conti, M. C.,
& Block, T. (2007). Underestimation of chronic hepatitis B in
APIs: A call for advocacy and action [Abstract 149577]. Paper
presented at the 135th American Public Health Association
Annual Meeting & Exposition, Washington, DC.
Craig, J. V., & Smyth, R. L. (2012). The evidence-based practice
manual for nurses (3rd ed.). Edinburgh, Scotland: Churchill
Livingstone.
Creswell, J. W. (2009). Research design: Qualitative, quantitative,
and mixed methods approaches (3rd ed.). Los Angeles, CA:
Sage.
Doran, D. (2011). Nursing-sensitive outcomes: The state of the
science (2nd ed.). Sudbury, MA: Jones & Bartlett Learning.
El-Mallakh, P. (2007). Doing my best: Poverty and self-care among
individuals with schizophrenia and diabetes mellitus. Archives of
Psychiatric Nursing, 21(1), 49–60.
Fahs, P. S., Morgan, L. L., & Kalman, M. (2003). A call for replica-
tion. Journal of Nursing Scholarship, 35(1), 67–71.
Fawcett, J., & Garity, J. (2009). Evaluating research for evidence-
based nursing practice. Philadelphia, PA: F. A. Davis.
Forbes, A. (2009). Clinical intervention research in nursing. Inter-
national Journal of Nursing Studies, 46(4), 557–568.
Frenn, M., Malin, S., & Bansal, N.K. (2003). Stage-based interven-
tions for low-fat diet with middle school students. Journal of
Pediatric Nursing, 18(1), 36–45.
Grady, J. L., Entin, E. B., Entin, E. E., & Brunye, T. T. (2011). Using
message framing to achieve long-term behavioral changes
in persons with diabetes. Applied Nursing Research, 24(1),
22–28.
Haller, K. B., & Reynolds, M. A. (1986). Using research in practice:
A case for replication in nursing: Part II. Western Journal of
Nursing Research, 8(2), 249–252.

Happ, M. B., Swigart, V. A., Tate, J. A., Arnold, R. M., Sereika, S.
M., & Hoffman, L. A. (2007). Family presence and surveillance
during weaning from prolonged mechanical ventilation. Heart &
Lung, 36(1), 47–57.
Houck, G., Kendall, J., Miller, A., Morrell, P., & Wiebe, G. (2011).
Self-concept in children and adolescents with attention deficit
hyperactivity disorder. Journal of Pediatric Nursing, 26(3),
239–247.
Jackson, C. L., Bolen, S., Brancati, F. L., Batts-Turner, M. L., &
Gary, T. L. (2006). A systematic review of interactive computer-
assisted technology in diabetes care. Journal of General Internal
Medicine, 21(2), 105–110.
Kaplan, B. A. (1964). The conduct of inquiry: Methodology for
behavioral science. New York, NY: Harper & Row.
Kerlinger, F. N., & Lee, H. B. (2000). Foundations of behavioral
research (4th ed.). Fort Worth, TX: Harcourt College
Publishers.
Kozlowski, S. W. J., Steve, W. J., & Bell, B. S. (2003). Work groups
and teams in organization. In W. C. Borman, D. R. Ilgen, & R. J.
Klimoski (Eds.), Comprehensive handbook of psychology: Indus-
trial and organizational psychology. New York, NY: Wiley.
Leifer, S. L., & Glass, L. K. (2008). Planning for mass disaster in
the 1950s: Harriet H. Werley and nursing research. Nursing
Research, 57(4), 237–244.
Letourneau, N., Young, C., Secco, L., Stewart, M., Hughes, J., &
Critchley, K. (2011). Supporting mothering: Service providers’
perspectives of mothers and young children affected by intimate
partner violence. Research in Nursing & Health, 34(3),
192–203.
Lewandowski, A., & Kositsky, A. M. (1983). Research priorities for
critical care nursing: A study by the American Association of
Critical Care Nurses. Heart & Lung, 12(1), 35–44.
Lindeman, C.A. (1975). Delphi survey of priorities in clinical
nursing research. Nursing Research, 24(6), 434–441.
Litrownik, A. J., Newton, R., Hunter, W. M., English, D., &
Everson, M. D. (2003). Exposure to family violence in young
at-risk children: A longitudinal look at the effects of victimization
and witnessed physical and psychological aggression. Journal of
Family Violence, 18(1), 59–73.
Marshall, C., & Rossman, G. B. (2011). Designing qualitative
research (5th ed.). Los Angeles, CA: Sage.
Masoli, M., Fabian, D., Holt, S., Beasley, R., & Global Initiative
for Asthma (GINA) Program. (2004). The global burden of
asthma: Executive summary of the GINA Dissemination Commit-
tee Report. Allergy, 59(5), 469–478.
McCain, N. L., Munjas, B. A., Munro, C. L., Elswick, R. K.,
Robins, J. L. W., Ferreira-Gonzales, A., et al. (2003). Effects
of stress management on the PNI-based outcomes in persons with
HIV disease. Research in Nursing & Health, 26(2), 102–117.
Melnyk, B. M., & Fineout-Overholt, E. (2011). Evidence-based
practice in nursing & healthcare: A guide to best practice (2nd
ed.). Philadelphia, PA: Lippincott Williams & Wilkins.
Munhall, P. L. (2012). Nursing research: A qualitative perspective
(5th. ed.). Sudbury, MA: Jones & Bartlett.
National Institute of Nursing Research (NINR). (2011). Strategic
plan National Institute of Nursing Research: Areas of research
emphasis. Retrieved from http://www.ninr.nih.gov/AboutNINR/
NINRMissionandStrategicPlan/.
CHAPTER 5  Research Problem and Purpose 95
National Institutes of Mental Health (2008). Attention deficit hyper-
activity disorder. Retrieved from http://www.nimh.nih.gov/health/
publications/adhd-listing.shtml.
Nyamathi, A., Liu, Y., Marfisee, M., Shoptaw, S., Gregerson, P.,
Saab, S., et al. (2009). Effects of a nurse-managed program on
hepatitis A and B vaccine completion among homeless adults.
Nursing Research, 58(1), 13–22.
Pender, N. J. (1996). Health promotion in nursing practice (3rd ed.).
Stamford, CT: Appleton & Lange.
Price, D., Ryan, D., Pearce, L., Bawden, R., Freeman, D., Thomas,
M., et al. (2002). The burden of pediatric asthma is higher
than health professionals think: Results from the Asthma In Real
Life (AIR) study. Primary Care Respiratory Journal, 11(1),
30–33.
Prochaska, J. O., Norcross, J. C., Fowler, J. L., Follick, M. J., &
Abrams, D. B. (1992). Attendance and outcome in a work site
weight control program: Processes and stages of change as
process and predictor variables. Addictive Behaviors, 17(1),
35–45.
Rogers, B. (1987). Research corner: Is the research project feasible?
AAOHN Journal, 35(7), 327–328.
Rohrbeck, J., Jordan, K., & Croft, P. (2007). The frequency and
characteristics of chronic widespread pain in general practice:
A case-control study. British Journal of General Practice,
57(535), 109–115.
Schizophrenia and Diabetes Expert Consensus Group. (2004).
Consensus summary. British Journal of Psychiatry, 47(2),
S112–S114.
Shadish, W. R., Cook, T. D., & Campbell, D. T. (2002). Experimen-
tal and quasi-experimental designs for generalized causal infer-
ence. Chicago, IL: Rand McNally.
Sharma, N. K., Ryals, J. M., Gajewski, B. J., & Wright, D. E.
(2010). Aerobic exercise alters analgesia and neurotrophin-3 syn-
thesis in an animal model of chronic widespread pain. Physical
Therapy, 90(5), 714–725.
Singh, N., Armstrong, D., & Lipsky, B. (2005). Preventing foot
ulcers in patients with diabetes. Journal of American Medical
Association, 293(2), 217–228.
Smith, M. J., & Liehr, P. R. (2008). Middle range theory for nursing
(2nd ed.). New York, NY: Springer Publishing Company.
Steel, J., Sanna, L., Hammond, B., Wipple, J., & Cross, H. (2004).
Psychological sequelae of childhood sexual abuse: Abuse-related
characteristics, coping strategies, and attributional style. Child
Abuse & Neglect, 18(7), 785–801.
Treuting, J., & Hinshaw, S. (2001). Depression and self-esteem in
boys with ADHD: Associations with comorbid aggression and
explanatory attributional mechanisms. Journal of Abnormal Psy-
chology, 29(1), 23–39.
Trollvik, A., Nordbach, R., Silen, C., & Ringsberg, K. C. (2011).
Children’s experiences of living with asthma: Fear of exacerba-
tions and being ostracized. Journal of Pediatric Nursing, 26(4),
295–303.
Trotter, R., Gallagher, R., & Donoghue, J. (2011). Care of patients
with coronary heart disease: Anxiety in patients undergoing per-
cutaneous coronary interventions. Heart & Lung, 40(3),
185–192.
U.S. Department of Health and Human Services (U.S. DHHS).
(2012). Healthy people 2020: Topics and objectives. Retrieved
96 UNIT TWO  The Research Process
from http://www.healthypeople.gov/2020/topicsobjectives2020/
default.aspx/.
Valle, L. A., & Silovsky, J. F. (2002). Attributions and adjustment
following child sexual and physical abuse. Child Maltreatment,
7(1), 9–25.
Whiteside, A., Hansen, S., & Chaudhuri, A. (2004). Exercise lowers
pain threshold in chronic fatigue syndrome. Pain, 109(3),
497–499.
World Health Organization (WHO). (2012). World Health Organi-
zation research policy. Retrieved from http://who.int/rpc/en/.
Zinzow, H., Seth, P., Jackson, J., Niehaus, A., & Fitzgerald, M.
(2010). Abuse and parental characteristics, attributions of blame,
and psychological adjustment in adult survivors of child sexual
abuse. Journal of Child Sexual Abuse, 19(1), 79–98.
   http://evolve.elsevier.com/Grove/practice/
6
Review of Relevant Literature
Y
ou have been asked to present a lecture about
home health care for undergraduate nursing
students. Maybe you are concerned about the
families of the patients in your critical care unit and
want to devise a program to address their unique needs.
Maybe you are a graduate nursing student and your
teacher said your paper needed to include a review of
the literature. Maybe you are in a Magnet hospital and
the nurses on your unit are developing a proposal for
a study. In each situation, you need to understand how
to review the literature and to present the information
you find in a logical, synthesized manner. By building
on previous knowledge, nurse researchers can add to
the evidence upon which we base our practice.
The rate of new knowledge being generated each
year continues to grow. Early studies in the 1960s
indicated that knowledge doubled every 13 to 15 years
(Larsen & von Ins, 2010). The vast amount of infor-
mation available within seconds implies that knowl-
edge is doubling much more rapidly in the digital age.
Bachrach (2001) noted that each new discipline
launches new journals to develop its disciplinary
knowledge. Computerized bibliographical databases
have made the process of searching for relevant empir-
ical or theoretical literature easier in some ways, but
you are faced with the dilemma of selecting the most
relevant sources from a much larger number of arti-
cles. The task of reading, critically appraising, analyz-
ing, and synthesizing has expanded and can consume
any time gained by more efficient searching. This
chapter provides basic skills and knowledge to iden-
tify evidence for changing nursing practice, develop-
ing a research proposal, preparing a lecture, or writing
a manuscript.
What Is “The Literature”?
“The literature” consists of all written sources rele-
vant to the topic you have selected. The literature
 
CHAPTER
consists of newspapers, monographs, encyclopedias,
conference papers, scientific journals, textbooks,
other books, theses, dissertations, and clinical jour-
nals. Websites and reports developed by government
agencies and professional organizations are also
included. For example, to support the significance of
diabetes mellitus as a topic for a research study, you
could find statistics about the prevalence and cost of
the disease from the Centers for Disease Control and
Prevention (CDC) and the World Health Organization
(WHO). Not every source that you find, however,
will be valid and legitimate for scholarly use. The
website of a company that sells insulin may not be an
appropriate source for diabetes statistics. Online
encyclopedias to which anyone can contribute, such
as Wikipedia, are not considered scholarly sources.
These websites might point you toward professional
sources but should not be cited in a professional or
academic paper. You should use primary, peer-
reviewed, professional literature. New knowledge
develops when researchers and scholars produce
manuscripts for journals and books that are reviewed
by peers to determine whether the manuscripts should
be published.
What Is a Literature Review?
The literature review is an organized written presen-
tation of what you find when you review the literature.
The literature review is “central to scholarly work and
disciplined inquiry” (Holbrook, Bourke, Fairbairn, &
Lovat, 2007, p. 337); it summarizes what has been
published on a topic by scholars and presents relevant
research findings. Developing the ability to write
coherently about what you have found in the literature
requires time and guidance. The review should be
organized into sections that present themes, identify
trends, or examine variables. The purpose is not to
list all the material published but, rather, to synthesize
97
98 UNIT TWO  The Research Process
and evaluate it on the basis of the phenomenon of
interest.
The focus of the review depends on the reason you
are reviewing the literature. This overview describes
common purposes for conducting literature reviews;
however, the goal of this chapter is to provide specific
guidance and practical suggestions related to review-
ing research relevant to a proposed study. The three
major stages of literature reviews are discussed: (1)
searching the literature, (2) processing the literature,
and (3) writing the literature review.
Purposes of Reviewing
the Literature
Writing a Course Paper
For most course papers, your instructor will expect
you to review published sources on the topic of your
paper. Reviews of the literature for a course assign-
ment will vary depending on the level of educational
program, the purpose of the assignment, and the
expectations of the instructor. The literature review for
a graduate course is expected to have greater depth,
scope, and breadth than a review for an undergraduate
course (Hart, 2009). A paper in a nurse practitioner
course might require that you review pharmacology
and pathology reference books in addition to journal
articles. In a nursing education course, you may review
neurological development, cognitive science, and
general education publications to write a paper on
a teaching strategy. For a doctor of nursing practice
course on clinical information systems, your review
might need to extend into computer science and hos-
pital management literature. For a theory course in a
doctor of philosophy in nursing program, your review
may need to include all the publications of a specific
theorist or all the studies based on the theory. For each
of these papers, your professor may specify the
publication years and the type of literature to be
included. Also, you must note the acceptable length of
the written review of the literature to be submitted.
Reviews of the literature for course assignments tend
to focus on what is known, the strength of the evi-
dence, the implications of the knowledge, and what is
not known for the purpose of developing new studies.
Examining the Strength of the Evidence
Evidence-based practice guidelines are developed
through the synthesis of the literature on the clinical
problem. The purpose of the literature review designed
to examine the strength of the evidence is to identify
all studies that provide evidence of a particular inter-
vention, to critically appraise the quality of each study,
and to synthesize all of the studies providing evidence
of the effectiveness of a particular intervention. It is
also important to locate and include previous evidence-
based papers that have examined the evidence of a
particular intervention, because the conclusions of the
authors of such papers are highly relevant. Literature
syntheses related to promoting evidence-based nursing
practice are described in Chapter 19.
Developing a Qualitative Research Proposal
In qualitative research, the purpose and timing of the
literature review depend on the type of study to be
conducted (see Chapter 12). Some phenomenologists
believe that the literature should not be reviewed until
after the data have been collected and analyzed, so that
the literature does not interfere with the researcher’s
ability to suspend what is known and to approach the
topic with openness (Munhall, 2012). In the develop-
ment of a grounded theory study, a minimal review of
relevant studies provides the beginning point of the
inquiry, but this review is only a means of making the
researcher aware of what studies have been conducted.
This information, however, is not used to direct the
collection of data or interpretation of the findings in a
grounded theory study. During the data analysis stage,
a core variable is identified and the researcher theoreti-
cally samples the literature for extant theories that may
assist in explaining and extending the emerging theory
(Munhall, 2012). In historical research, the initial
review of the literature helps the researcher define the
study questions and make decisions about relevant
sources. The data collection is actually an intense
review of published and unpublished documents that
the researcher has found.
The purposes of reviewing the literature for ethno-
graphic studies and for exploratory descriptive quali-
tative research are more similar to that for quantitative
research. The researcher develops a general under-
standing of the concepts to be examined in relation to
the selected culture or topic. The literature review also
provides a background for conducting the study and
interpreting the findings. Chapter 12 describes in more
detail the role of the literature review in qualitative
research.
Developing a Quantitative Study
The review of literature in quantitative research directs
the development and implementation of a study. The
focus of the major literature review at the beginning
of the research process is to identify a gap in what is
known. The study is designed to add knowledge in the
 CHAPTER 6  Review of Relevant Literature 99

TABLE 6-1  The Role of the Literature Review in Developing a Quantitative Research Proposal
Phase of the
Research Process How Literature Is Used and Its Role
Research topic Broad searches using keywords to understand the extent of what is known and what is not known; what
concepts are related to the topic
Statement of the From your synthesis of the literature, the specific gap in knowledge that this study will address
research purpose
Background and Searches of books and articles to provide an overview of the topic
significance Identification of the size, cost, and consequences of the research problem
Research framework Find and read relevant theories
Facilitate development of the framework
Develop conceptual definitions of concepts
Purpose of the study On the basis of your knowledge of the literature, state the purpose of the study
Research objectives, On the basis of the knowledge gained from and examples found in the literature, write the objectives or
questions, or questions of the study
hypotheses If sufficient literature allows a prediction, state the hypotheses of the study
Review of the literature Find sources as evidence for logical argument for why this study and methodology are needed
Summarize current empirical knowledge that is related to the topic
Methodology Compare research designs of reviewed studies to select the most appropriate design for the proposed study
Identify possible instruments or measures of variables
Provide operational definitions of concepts
Describe performance of measures in previous studies
Develop sampling strategies based on what you have learned from the studies in the literature
Findings Refer to statistical textbooks to explain the results of the data analysis
Discussion Compare your findings with those of studies you have previously reviewed
Return to the literature to find new references to interpret unexpected findings
Identify limitations of the study
Refer to theory sources to relate the findings to the research framework
Conclusions On basis of your knowledge of the literature and your study’s findings, draw conclusions
Discuss implications for nursing clinical practice, administration, and education
Propose future studies

area of the identified gap. For example, an intervention
to prevent hospital-acquired infections related to intra-
venous infusions has been shown to reduce the inci-
dence of these infections among postoperative patients
who have no history of diabetes mellitus. After a thor-
ough review of the literature, the researcher identifies
a specific gap in knowledge. What is not known is
whether this intervention will be equally effective for
postoperative patients who have diabetes. After the
data have been analyzed and the findings described,
the researcher will return to the literature in the gen-
eralization phase of the research report to integrate
knowledge from the literature with new knowledge
obtained from the study. The purposes of the literature
review are similar for the different types of
quantitative studies (descriptive, correlational, quasi-
experimental, and experimental).
Table 6-1 describes the role of the literature
throughout the development and implementation of
the study. The types of sources needed and how you
will search the literature will vary throughout the
study. The introduction section uses relevant sources
to summarize the background and significance of the
research problem. The review of the literature section
includes both theoretical and empirical sources that
document the current knowledge of the problem. The
researcher develops the framework section from the
theoretical literature and sometimes from empirical
literature. If little theoretical literature is found, the
researcher may need to develop a tentative theory to
guide the study from the findings of previous research
studies (see Chapter 7 for more information).
The methods section describes the design, sample,
measurement methods, treatment, and data collection
100 UNIT TWO  The Research Process
process of the planned study and is based on previous
research. Thus, previous studies may be cited in the
methods section. In the results section, sources are
included to document the different types of statistical
analyses conducted and the computer software to
conduct these analyses. The discussion section of the
research report begins with what the results mean in
light of the results of previous studies. Conclusions
are drawn that are a synthesis of the cited findings
from previous research and those from the present
study.
Practical Considerations
What Types of Literature Can I Expect
to Find?
Two broad types of literature are cited in the review
of literature for research: theoretical and empirical.
Theoretical literature consists of concept analyses,
models, theories, and conceptual frameworks that
support a selected research problem and purpose.
Empirical literature comprises knowledge derived
from research. The empirical literature reviewed
depends on the study problem and the type of research
conducted. Research problems that have been fre-
quently studied or are currently being investigated
have more extensive empirical literature than new or
unique problems. If searching the empirical literature,
you need to identify seminal and landmark studies.
Seminal studies are the first studies that prompted the
initiation of the field of research. Nurse researchers
studying hearing loss in infants would need to review
the seminal work of Fred H. Bess, an early researcher
on this topic who advocated for effective screening
tools (Gravel, 2009). Critical care nurses comparing
correction formulas for QT intervals on electrocardio-
grams would want to refer to Bazet’s correction
formula. The development of the formula can be
traced to his seminal paper, published in 1920, on
time-relations in electrocardiograms (Roguin, 2011).
Landmark studies are the studies that led to an
important development or a turning point in the field
of research. For example, researchers conducting
studies related to glycemic control must be knowl-
edgeable of the implications of the Diabetes Control
and Complications Trial, a longitudinal study whose
findings changed diabetic care beginning in the mid-
1990s (Everett, Bowes, & Kerr, 2010).
Literature is disseminated in several different
formats. Serials are published over time or may be
in multiple volumes but do not necessarily have pre-
dictable publication dates. Periodicals are subsets of
serials with predictable publication dates, such as jour-
nals, which are published over time and are numbered
sequentially for the years published. This sequential
numbering is seen in the year, volume, issue, and page
numbering of a journal. Monographs, such as books,
hard-copy conference proceedings, and pamphlets, are
usually written once and may be updated with a new
edition as needed. Conference proceedings can help
you identify major researchers in your research area
who have presented findings that may not yet be pub-
lished. Periodicals and monographs are available in a
variety of media, such as print, online, CD-ROM, and
downloadable formats. Textbooks are monographs
written to be used in formal education programs.
Entire volumes of books available in a digital or
electronic format are called eBooks (Tensen, 2010).
You may be familiar with digital books in the mass
publication literature that are available to download to
read on a special reading device, such as a Kindle or
Nook. eBooks are also available for scholarly volumes
and articles that can be downloaded to a reading
device, cell phone, laptop, or other computer. Books
that in the past would have been difficult to obtain
through interlibrary loan are now available 24 hours a
day, 7 days a week as eBooks.
To develop the significance and background section
of a proposal, you may also need to search for govern-
ment reports for the United States (U.S.) and other
countries, if appropriate for your study. A researcher
developing a proposal on task shifting in HIV care
settings in low-resource countries would search the
Ministry of Health websites for those countries to find
official guidelines for this type of practice. Research-
ers developing a proposal in Wisconsin on the smoking
cessation in adolescents would consult the Healthy
People 2020 website for the national goals related
to this topic (http://www.healthypeople.gov/2020/
default.aspx/). They may also explore health-related
agencies in Wisconsin to determine information spe-
cific to their state.
Position papers are disseminated by professional
organizations and government agencies to promote a
particular viewpoint on a debatable issue. Position
papers, along with descriptions of clinical situations,
may be included in the discussion of the background
and significance of the research problem. A researcher
developing a proposal on race-related differences in
HIV treatment outcomes would want to review the
Association of Nurses in AIDS Care position paper,
“Health Disparities,” which the organization’s board
approved in 2009.
Master’s theses and doctoral dissertations are valu-
able literature as well but may not be published. A

thesis is a research project completed as part of the
requirements for a master’s degree. A dissertation
is an extensive, usually original research project
that is completed as the final requirement for a doc-
toral degree. Theses and dissertations can be found
by searching special databases that are available for
these publications, such as ProQuest Dissertations
and Theses (http://www.proquest.com/en-US/default
.shtml/).
The published literature contains primary and sec-
ondary sources. A primary source is written by the
person who originated, or is responsible for generat-
ing, the ideas published. A research publication pub-
lished by the person or people who conducted the
research is a primary source. A theoretical book or
paper written by the theorist who developed the
theory or conceptual content is a primary source. A
secondary source summarizes or quotes content
from primary sources. Thus, authors of secondary
sources paraphrase the works of researchers and the-
orists. The problem with a secondary source is that
its author has interpreted the works of someone else,
and this interpretation is influenced by that author’s
perception and bias. Authors have sometimes spread
errors and misinterpretations by using secondary
sources rather than primary sources. You should use
mostly primary sources to write literature reviews.
Secondary sources are used only if primary sources
cannot be located or if a secondary source contains
creative ideas or a unique organization of informa-
tion not found in a primary source. Citation is the
act of quoting or paraphrasing a source, using it as an
example, or presenting it as support for a position
taken.
How Long Will the Review
of the Literature Take?
The time required to review the literature is influenced
by the problem studied, sources available, and goals
of the reviewer. The literature review for a topic that
is focused and somewhat narrow may require less time
than one for a topic that is broad. The difficulty you
experience identifying and locating sources and the
number of sources to be located also influence the time
involved, as does the intensity of effort. Only through
experience does one become knowledgeable about the
time needed for a literature review.
The novice reviewer requires more time to find the
relevant literature than an experienced searcher, and
the novice frequently underestimates the time needed
for the review. An experienced librarian who works
closely with nursing graduate students on a variety of
course assignments recommends that the reviewer
CHAPTER 6  Review of Relevant Literature 101
estimate the time required on the basis of the number
of sources required and the reviewer’s familiarity with
the library’s databases. The reviewer knows he or she
needs at least 30 sources and that finding these sources
may require 10 hours of searching. The reviewer esti-
mates that it will take another 20 hours to read and
synthesize the sources. The reviewer estimates that the
review will require 30 hours. The reviewer should
then multiply that number by four, making the esti-
mated time required for the review 120 hours. This
longer estimate is often more realistic. As searching
skills are refined, the need to use this expanded esti-
mate reduces. Often, the literature review is limited by
the time that the reviewer can commit to the assign-
ment. The conclusion related to the time issue is to
start as early as possible and stay focused on the
purpose of the review.
How Many Sources Do I Need to Review?
Students repeatedly ask, “How many articles should I
have? How far back in years should I look to find
relevant information?” The answer to both those ques-
tions is an emphatic “It depends.” Course faculty for
masters courses commonly require that you obtain
full-text articles of all studies relevant to the variables
in the proposed study that were published in the previ-
ous 10 years. The instructors often indicate, however,
that the length of time may vary depending on the
topic and the presence of classic studies. Doctoral
students are expected to conduct a more extensive
review for course papers. If you are writing a research
proposal for a thesis or dissertation, the literature
required will be extensive. You need to locate the key
papers in the field of interest. After doing some initial
searches, discuss what you find with your instructor,
thesis chair, or dissertation chair, who will be able to
help you determine a reasonable publication period for
you to use in your review.
Am I Expected to Read Every Word
of the Available Sources?
The answer is “No.” If researchers attempted to read
every word of every source that is somewhat related
to a selected problem, they would be well read but
would probably never complete their searches or
move on to developing study proposals. Some indi-
viduals, even after a thorough literature review, con-
tinue to believe that they do not know enough about
their area of interest, so they persist in their review;
however, this activity ultimately becomes an excuse
for not progressing with their work. The opposite of
this situation is the individual who wants to move
rapidly through the review of literature to reach the
102 UNIT TWO  The Research Process
conclusion or get to the part of the work that is more
enjoyable or important.
With the availability of full-text online articles, the
researcher can easily get “lost in the literature” and
forget the focus of the review. Becoming a skilled
reviewer of the literature involves finding a balance
and learning to identify the most pertinent and relevant
sources. On the other hand, you cannot critically
appraise and synthesize what you have not read. Avoid
being distracted by nonrelevant information provided
by the author. Learn to read with a purpose and involve
multiple senses in your reading. Try reading aloud a
section that you have difficulty understanding. Are
you having difficulty following an author’s presenta-
tion? Try writing an outline. Draw a diagram of key
points. Audio record your thoughts on the content of
an article. Listen to soothing music if you are tense or
anxious. Listen to music with an upbeat tempo if you
are tired and having trouble staying awake. Try reading
in a quiet place outside or standing up with the copy
of the article or your laptop on a counter or high table.
Involving multiple senses while reading may help you
stay awake and focused.
Stages of a Literature Review
The stages of a literature review reflect a systems
model. Systems have input, throughput, and output.
The input consists of the sources that you find through
searching the literature. The throughput is the pro-
cesses you use to read, critically appraise, analyze, and
synthesize the literature you find. The written litera-
ture review is the output of these processes (see Figure
6-1). The quality of the input and throughput will
determine the quality of the output. As a result, each
stage of the literature review is critical to producing a
high-quality literature review. Although these stages
are presented here as sequential, you may go back to
a previous stage. For example, during the analysis and
INPUT
THROUGHPUT
OUTPUT
Searching Writing the
Processing the
the Literature
Literature
Literature
Review
Figure 6-1  Systems model of the review of the literature.
synthesis of your sources, you identify that the studies
you are citing were conducted only in Europe. You
might go back and search the literature again using the
United States or another search term to ascertain that
no studies have been done in that country. As you are
writing your literature review, you may identify a
problem with the logic of your presentation. To resolve
it, you may return to the processing stage to clarify the
presentation.
Searching the Literature
Before writing a literature review, you must first
perform literature searches to identify sources relevant
to your topic of interest. The literature review will help
you narrow your topic and develop a feasible study
(Hart, 2009). Whether you are a student, practicing
nurse, or nurse researcher, your goal is to develop a
search strategy designed to retrieve as much of the
relevant literature as possible given the time and finan-
cial constraints of your project.
Libraries have become gateways to information or
information resource centers, rather than storehouses
of knowledge (Hart, 2009). High-quality libraries
provide access to a large number of electronic data-
bases that supply a broad scope of the literature avail-
able internationally, enabling library users not only to
identify relevant sources quickly but also to read full-
text versions of most of these sources immediately.
When your library does not have the hard copy of a
book or electronic access to a specific journal, the
librarian can usually provide the book or an electronic
copy of the article through interlibrary loan. All librar-
ies, public, private, college, and university, have inter-
library loan capabilities. You may especially need to
use interlibrary loan when sources relevant to your
topic were published prior to the advent of electronic
databases.
Consider consulting with an information profes-
sional, such as a subject specialist librarian, to develop
a literature search approach (Booth, Colomb, & Wil-
liams, 2008; Tensen, 2010). Often these consultations
can be performed via email or a web-based meeting,
so that communication occurs at the convenience of
both the researcher and the information professional.
Many university libraries provide this consultation
service whether or not the library user is affiliated with
the university.
Develop a Search Plan
Before you begin searching the literature, you must
consider exactly what information you are seeking.
Expending time and effort in the early stage of a
review to develop a search strategy is likely to save

TABLE 6-2  Plan and Record for Searching the Literature
Database Searched Date of Search
Cumulative Index to Nursing and
Allied Health Literature (CINAHL)
MEDLINE
Academic Search Premier
Cochrane Library
time and effort later (Hart, 2009). A written plan helps
you to avoid duplication of effort, to return to a previ-
ously searched area with a different set of search terms
or a different range of publication years. Your initial
search should be based on the widest possible inter-
pretation of your topic. This strategy enables you to
envision the extent of the relevant literature. As you
see the results of the initial searches and begin reading
the material, you will refine your topic, and then you
can narrow the focus of your searches.
As you search, add your selected search terms to
your written search plan. As you search, add other
terms that you discover from the references you locate.
For each search, record (1) the name of the database,
(2) the date, (3) search terms and searching strategy,
(4) the number and types of articles found, and (5) an
estimate of the proportion of the citations you retrieved
that were relevant articles. Table 6-2 is an example of
search history that you can use to record what and how
you have searched the literature. Save the results of
each search on your computer or external device.
Some databases allow you to create an account and
save your search history online (i.e., the record of what
and how you searched).
Select Databases to Search
A database is computer data that have been collected
and arranged to be searchable and automatically
retrievable (Tensen, 2010). A bibliographical data-
base is an “an electronic version of a bibliographic
index” (p. 57) or compilation of citations. The data-
base may consist of citations relevant to a specific
discipline or may be a broad collection of citations
from a variety of disciplines. Databases of periodical
citations include the authors, title, journal, keywords,
and usually an abstract of the article. Library data-
bases contain titles and authors of hard copy books
and documents, government reports, and reference
books. Library databases also include a searchable list
of the journals to which the library has a paper or
electronic subscription. For example, your library
CHAPTER 6  Review of Relevant Literature 103
Search Strategy Number and Type Estimate of
and Limiters of Articles Found Relevant Articles
may have received paper copies of a monthly journal
in the mail until 2006. The hard copies of the issues
were bound to create annual volumes of the journal.
Since 2006, the library has subscribed to the elec-
tronic journal or a journal database that provides
access to specific issues.
Bibliographical databases provided by the same
vendor, such as those databases affiliated with EBSCO
Publishing, allow you to search multiple databases
simultaneously to save time. Usually the search engine
will automatically delete duplicates of the same study.
You can also change the order in which the results of
the search are shown. For example, with the EBSCO
Publishing databases, you can sort the citations by
relevance, date descending (most current first), or date
ascending (oldest to more recent).
Older sources of reference indexes are useful for
sources published prior to the electronic databases.
Card catalogs, abstract reviews, and indexes were the
only ways to search for nursing references until 1955,
when the Cumulative Index to Nursing and Allied
Health Literature (CINAHL) began being published.
Because the printed editions had red covers, you still
may hear more experienced scholars fondly refer to
“the Red Books.” The print version of CINAHL is still
available in libraries, and you may find it useful when
searching for citations published before 1982 or when
bibliographic databases are not available (Tensen,
2010). In medicine, the Index Medicus (IM) was first
published in 1879 and is the oldest health-related
index. The Index Medicus includes some citations of
nursing publications, with the number of nursing jour-
nals cited growing. CINAHL contains, however, a
more extensive listing of nursing publications and
uses more nursing terminology as subject headings.
With the greater focus on interdisciplinary research,
nurse researchers must also be consumers of the litera-
ture in the National Library of Medicine (MEDLINE),
other government agencies, and professional organi-
zations. Table 6-3 provides descriptions of commonly
used bibliographical databases.
104 UNIT TWO  The Research Process

TABLE 6-3  Bibliographical Databases

Name of Database Description of the Database by the Publisher*
Cumulative Index of Nursing and “Comprehensive source of full text for nursing & allied health journals, providing full
Allied Health Literature (CINAHL) text for more than 770 journals”
MEDLINE “Information on medicine, nursing, dentistry, veterinary medicine, the health care system,
pre-clinical sciences, and much more”
Created and provided by the National Library of Medicine
Uses Medical Subject Headings (MeSH terms) for indexing and searching of “citations
from over 4,800 current biomedical journals”
PubMed Free access to Medline that provides links to full-text articles when available
PsychARTICLES 15,000 “full-text, peer-reviewed scholarly and scientific articles in psychology”
Limited to journals published by the American Psychological Association (APA) and
affiliated organizations
PsychINFO “Scholarly journal articles, book chapters, books, and dissertations, is the largest resource
devoted to peer-reviewed literature in behavioral science and mental health”
Supported by APA
Covers over 3 million records
Academic Search Complete “Comprehensive scholarly, multi-disciplinary full-text database, with more than 8,500
full-text periodicals, including more than 7,300 peer-reviewed journals”
Health Source Nursing/ Academic “Provides nearly 550 scholarly full text journals focusing on many medical disciplines”
Edition Also includes 1,300 patient education sheets for generic drugs
Psychological and Behavioral “Comprehensive database covering information concerning topics in emotional and
Sciences Collection behavioral characteristics, psychiatry & psychology, mental processes, anthropology,
and observational & experimental methods”
400 journals indexed

*Direct quotations from EBSCO Publishing descriptions of the databases, available at http://www.ebscohost.com/academic/.

Search Strategies A simple way to begin identifying a database’s

Keywords
Keywords are the everyday words and phrases used
for the major concepts or variables that must be
included in your computer search (Tensen, 2010). To
determine keywords, identify the concepts relevant to
your study. Ascertain the populations that are of par-
ticular interest in your area of study or the specific
interventions, measurement methods, or outcomes
that are relevant. These databases have a thesaurus
that the researcher, as well as anyone who reads an
article, can use as keyword search terms. By logging
on to the database, you can access the thesaurus to
select relevant terms. The formal subject terms
included in the thesaurus may encompass a number of
the terms that you have identified and allow you to
expand your search to obtain more references or to
focus your search to be more specific to your interest.
This expansion or focus occurs because someone who
has already read the articles has grouped and linked
all citations with similar concepts together.
standardized subject terms is to search using one of
your keywords and display full records of a few rel-
evant citations. The records are the descriptions of the
articles, not the articles themselves. The subject terms
linked to that article are listed on the full record.
Examine the terminology used to describe these arti-
cles, and use the terms in additional refined searches.
Frequently, word-processing programs, dictionaries,
and encyclopedias are helpful in identifying synony-
mous terms and subheadings. A combination of both
keywords and formal subjects most often retrieves
better search results.
The format and spelling of search terms can yield
different results. Truncating words can allow you to
locate more citations related to that term. For example,
authors might have used intervene, intervenes, inter-
vened, intervening, intervention, or intervenor. To
capture all of these terms, you can use a truncated term
in your search, such as interven, interven*, or inter-
ven$. The form or symbol used to truncate a search
term depends on the rule of the search engine being

used. Avoid shortening a search word to fewer than
four letters. Otherwise, you will get far too many
unwanted citations. Also, pay attention to variant
spellings. You may need to search, for example, by
orthopedic or orthopaedic (British spelling). Consider
irregular plurals, such as woman and women.
Authors
If an author is cited frequently, you can perform a
search using the author’s name. In this case, you
should identify the name as an author term, not a
keyword term. Recognize that some databases list
authors only under first and middle initials, whereas
others use full first names. Identifying and using cita-
tions to seminal studies in various citation indexes or
full-text databases can lead you to other, more current
works that have also cited the seminal studies as refer-
ences. Web of Knowledge, a database developed from
the Science Citation Index and the Social Science
Citation Index, focuses on the relationships among
these citations. These indexes may require that your
library subscribe to their services, however, Web of
Knowledge does have a Facebook page (http://
www.facebook.com/pages/Web-of-Knowledge/11968
7984715358/). Several other databases, depending on
the vendor, may also have a function to search the
references of articles.
Complex Searches
A complex search of the literature combines two or
more concepts or synonyms in one search. There are
several ways to arrange terms in a database search
phrase or phrases. The three most common ways are
by using (1) Boolean operators, (2) locational opera-
tors (field labels), and (3) positional operators. Opera-
tors permit you to group ideas, select places to search
in a database record, and show relationships within a
database record, sentence, or paragraph. Examine the
Help screen of a database carefully to determine
whether the operators you want to use are available
and how they are used.
The Boolean operators are the three words AND,
OR, and NOT. Often they must be capitalized. The
Boolean operators AND and NOT are used with your
identified concepts. Use AND when you want to
search for the presence of two or more terms in the
same citation. Use NOT when you want to search for
one idea but not another in the same citation. NOT is
rarely used because it is too easy to lose good cita-
tions. The Boolean operator OR is most useful with
synonymous terms or concepts. Use OR when you
want to search for the presence of any of a group of
terms in the same search. Figure 6-2 shows the results
CHAPTER 6  Review of Relevant Literature 105
Coping OR Social Support
5855 citations
Coping Social Support
3578 citations Coping 2688 citations
AND
Social
Support
409
citations
Figure 6-2  Example of search using social support and coping with
different Boolean operators.
of searches using coping and social support as key-
words alone and with Boolean operators.
Locational operators (field labels) identify terms
in specific areas or fields of a record. These fields
may be parts of the simple citation, such as the article
title, author, and journal name, or they may be from
additional fields provided by the database, such as
subject headings, abstracts, cited references, publica-
tion type notes, instruments used, and even the entire
article. In some databases, these specific fields can be
selected by means of a drop-down menu in the data-
base input area. In other databases, specific coding
can be used to do the same thing. Do not assume that
the entire article is being searched when you are
using the default search; the default is usually looking
for your terms in the title, abstract, and/or subject
fields. You may choose to search for a concept only
within the abstract of articles, on the basis of the
logic that this strategy is less restrictive than search-
ing for the concept only in article titles but more
restrictive than searching for the concept in all the
text of the article.
Positional operators are used to look for requested
terms within certain distances of one another. Avail-
ability and phrasing of positional operators are highly
dependent on the database search software. Common
positional operators are NEAR, WITH, and ADJ; they
also are often required to be capitalized and may have
numbers associated with them. A positional operator
is most useful in records with a large amount of infor-
mation, such as those with full-text articles attached,
and is often used with locational operators, either in
an implied way or explicitly. For example, ADJ is an
abbreviation for adjacent; it specifies that one term
must be adjacent to another in the order entered.
“ADJ2” commands that there must be no more than
two intervening words between the search terms in the
order entered. NEAR does not define the specific order
of the terms; the command “term1 NEAR1 term2”
106 UNIT TWO  The Research Process

TABLE 6-4  Selected Internet-Only Nursing Journals

Journal
Online Journal of Issues in Nursing
Internet Journal of Advanced Nursing Practice
Australian Electronic Journal of Nursing Education
CONNECT: The World of Critical Care Nursing
Open Nursing Journal
International Journal of Nursing
URL
http://nursingworld.org/MainMenuCategories/ANAMarketplace/
ANAPeriodicals/OJIN.aspx/
http://www.ispub.com/journal/the_internet_journal_of_advanced_nursing_
practice.html/
http://www.scu.edu.au/schools/nhcp/aejne/
http://en.connectpublishing.org/
http://www.benthamscience.com/open/tonursj/
http://www.ijnonline.com/index.php/ijn/index/

requires that the first term occur first and within two
words of the second term. WITH often indicates that
the terms must be within the same sentence, para-
graph, or region (such as subject headings) of the
record.
Limiting Your Search
You can use several strategies to limit your search if,
after performing Boolean searches, you continue to
get too many hits. The limits you can impose vary with
the database. In CINAHL, for example, you may limit
your search to English-language articles. You can also
limit the years of your search. For example, you might
choose to limit the search to articles published in the
past 10 years. Searches can be limited to find only
papers that are research, are reviews, are published in
consumer health journals, include abstracts, or are
available in full text. You may also narrow your search
by adding your population or intervention to the search
strategy.
Searching the Internet
A number of nursing journals are published only in
electronic form. Table 6-4 contains selected online
nursing journals with their URLs. Because of the high
costs of publishing and distributing a printed journal,
a publishing company risks losing money unless there
is a large market for the journal. Most of the electronic
journals are targeted to relatively small specialty audi-
ences. These journals may have more current informa-
tion on your topic than you will find in traditional
journals, because articles submitted by authors to
online journals are reviewed and published within 3
to 4 months. Articles submitted to printed journals are
usually under review for an extended time and, if
accepted, may not be seen in print for over a year.
Faculty members at some universities have estab-
lished online journals in a particular specialty area. In
some cases, you may have to subscribe to the online
journal to gain access to the articles. Some electronic
journals are listed in available bibliographical
databases, and you can access full-text articles from
an electronic journal through the database. However,
many electronic journals are not yet in the biblio-
graphical databases or may not be in the database you
are using. Ingenta Connect (www.ingenta.com) is a
commercial website that allows you to search more
than 13,000 publications from many disciplines. Pub-
lications available through Ingenta include those that
are free to download and those that require the reader
to buy the article.
Metasearch engines, such as Google, allow you to
search the Internet. Google Scholar is a specialized
tool that allows you to focus your search on research
and theoretical publications. With the exception of
articles in online-only journals, scholarly sources are
published first in traditional formats. Thus, what you
find online will be older references. Especially early
in your search, these older articles may point you to
seminal and landmark studies or help you identify
subject terms for new searches. Government reports
and publications by professional organizations may
also be found by searching the Internet. Prior to using
a reference from the Internet, you must evaluate the
reliability of the information and the potential for bias
on the part of the author. There is no screening process
for information placed on the World Wide Web. Thus,
you find a considerable amount of misinformation as
well as some “gems” you might not find elsewhere. It
is important to check the source of any information
you obtain from the Web so that you can judge its
validity.
Locate Relevant Literature
Within each database, conduct your search of relevant
literature using the strategies described in this chapter.
Most databases provide abstracts of the articles in
which the term is cited, allowing you to get some
sense of their content so you may judge whether the
information is useful in relation to your selected topic.
If you find the information to be an important refer-
ence, save it to a file on your computer or in an online

folder maintained by your employer or university, and/
or move it to a reference management program (next
section). At this point in the process, do not try to
examine all of the citations listed.
It is rarely, if ever, possible to identify every rele-
vant source in the literature. The most extensive
retrievals of literature are funded literature review
projects focused on defining evidence-based practice
or developing clinical practice guidelines (see Chapter
19). In these projects, a literature review coordinator
manages the literature review process. The project
employs several full-time, experienced, professional
librarians as literature searchers. When these extensive
literature reviews are completed, the results are pub-
lished so that you may have access to them and to the
citations from the review, either on the Internet or in
journal articles.
Systematically Record References
The bibliographical information on a source should be
recorded in a systematic manner, according to the
format that you will use in the reference list. The
purpose for carefully citing sources is that readers can
retrieve the reference for themselves, confirm your
interpretation of the findings, and gather additional
information on the topic. Many journals and academic
institutions use the format developed by the American
Psychological Association (APA) (2010). Computer-
ized lists of sources usually contain complete citations
for references and should be saved electronically so
you can access complete reference citations. The 6th
edition of the APA’s Publication Manual (2010) pro-
vides revised guidelines for citing electronic sources
and direct quotations from electronic sources. Citing
direct quotations from electronic sources has posed
unique challenges. The APA standard for direct quota-
tions from a print source is to cite the page of the
source on which the quotation appears. The reference
lists in this text are presented in APA format, with the
exceptions that we have not included digital object
identifiers (DOIs) and have slightly modified the cita-
tion of multiple authors.
DOIs have become the standard for the International
Standards Organization (http://www.doi.org/) but have
not yet received universal support. The use of DOIs
seems to be gaining in credibility because the DOI
“provides a means of persistent identification for man-
aging information on digital networks” (APA, 2010,
p. 188). CrossRef is an example of a registration agency
for DOIs that enables citations to be linked across
databases and disciplines (http://www.crossref.org/).
Each citation on the reference list is formatted as a
paragraph with a hanging indent, meaning that the first
CHAPTER 6  Review of Relevant Literature 107
line is on the left margin and subsequent lines are
indented. If you do not know how to format a paragraph
this way, search the Help tool in your word-processing
program to find the correct command to use. When you
retrieve an electronic source in portable document
format (pdf), you cite the source the same as if you had
made a copy of the print version of the article. When
you retrieve an electronic source in html format, you
will not have page numbers for the citation. The
updated APA standard is to provide the URL for the
home page of the journal from which the reader could
navigate and find the source (APA, 2010). Providing
the URL that you used to retrieve the article is not
helpful because it is unique to the path you used to find
the article and reflects your search engines and biblio-
graphical databases. The following are examples for
citing different types of articles:
Citation for the print copy or electronic pdf copy of
an article:
Everett, J., Bowes, A., & Kerr, D. (2010). Barriers to
achieving glycaemic control in CSII. Journal of
Diabetes Nursing, 14(5), 176-181. doi:10.1177/
1043659609357635.
Online full-text version of an article for which pdf was
not available:
Scott, A. J., & Wilson, R. F. (2011). Social determi-
nants of health among African Americans in a rural
community in the Deep South: An ecological study.
Rural and Remote Health, 11, 1634 (Online).
Retrieved from http://www.rrh.org.au/
Use Reference Management Software
Reference management software can make tracking
the references you have obtained through your searches
considerably easier. You can use such software to
conduct searches and to store the information on all
search fields for each reference obtained in a search,
including the abstract. Within the software, you can
store articles in folders with other similar articles. For
example, you may have a folder for theory sources,
another for methodological sources, and a third for
relevant research topics. When you export search
results from the bibliographic database to your refer-
ence management software, all of the needed citation
information and the abstract are readily available to
you electronically when you write the literature
review. As you read the articles, you can also insert
comments into the reference file about each one.
Reference management software has been devel-
oped to interface directly with the most commonly
used word processing software to organize the refer-
ence information using whatever citation style you
stipulate. You may be familiar with APA format but
108 UNIT TWO  The Research Process
want to submit a manuscript to a journal that uses
another bibliographical style. Within these reference
management programs, you can generate the reference
list or bibliography in the format specific to the journal.
You can insert citations into your paper with just a
keystroke or two. The four most commonly used soft-
ware packages, along with websites that contain infor-
mation about them, are as follows:
EndNote (http://www.endnote.com/) has an inclusive
format that is compatible with Windows and Mac
computers.
RefWorks (www.refworks.com/) operates from the
Web and can be accessed free by some universities’
affiliates depending on the license.
Reference Manager (http://www.refman.com/) oper-
ates on your personal computer or you can use it
to make your databases accessible to others in a
Web environment.
Bookends (http://www.sonnysoftware.com/) is refer-
ence manager for Mac users that allows users to
search bibliographical databases and download
citations and full-text articles. Searches can also be
downloaded from iPhone, iTouch, and iPad.
Saved Searches and Alerts
If you are working on a research project in which the
literature review may take months or are engaged in a
field of study that will interest you for years, you
might want to repeat the same search regularly. Many
databases permit you to create an account where you
can save your search strategy so you can redo the same
search with just a few clicks and without having to
enter the entire strategy again. You might want to have
just the new updates of a search strategy sent to you
automatically by email without having to redo the
entire search, even though this redo now entails just a
few clicks. These Saved Search and Alert features may
be available in your favorite databases. However,
review your saved and alert strategies with some regu-
larity to ensure you are obtaining what you really
desire. Many journals also permit a table of contents
to be sent to you automatically when new issues come
out. Examine the database or journal home page help
screens to determine how to create and use these
features.
Processing the Literature
Reading and critically appraising sources promote
understanding of the current knowledge of a research
problem. It involves skimming, comprehending, ana-
lyzing, and synthesizing content from sources. Skills
in reading and critically appraising sources are
essential to the development of a high-quality litera-
ture review.
Reading
Skimming a source is quickly reviewing a source to
gain a broad overview of its content. You would prob-
ably read the title, the author’s name, and an abstract
or introduction for the source. Then you would read
the major headings and sometimes one or two sen-
tences under each heading. Finally, you would review
the conclusion or summary. Skimming enables you to
make a preliminary judgment about the value of a
source and to determine whether it is a primary or
secondary source. Secondary sources are reviewed
and used to locate cited primary sources, but they are
seldom cited in a research proposal or report.
Comprehending a source requires that you read
all of it carefully. Focus on understanding major con-
cepts and the logical flow of ideas within the source.
Highlight the content you consider important or make
notes in the margins. Notes might be recorded on
photocopies or electronic files of articles, indicating
where the information will be used in developing a
research proposal. The kind of information you high-
light or note in the margins of a source depends on the
type of study or source. The information highlighted
on theoretical sources might include relevant con-
cepts, definitions of those concepts, and relationships
among them. The notes recorded in the margins of
empirical literature might include relevant informa-
tion about the researcher, such as whether the author
is a major researcher of a selected problem and com-
parisons with other studies this individual has con-
ducted. For a research article, the research problem,
purpose, framework, major variables, study design,
sample size, data collection, analysis techniques, and
findings are usually highlighted. You may wish to
record quotations (including page numbers) that might
be used in a review of literature section. The decision
to use or paraphrase these quotes can be made later.
You might also record creative ideas about content
that develop while you are reading a source. At this
point, you will identify relevant categories for sorting
and organizing sources. These categories will ulti-
mately guide you in writing the review of literature
section, and some may even be major headings in this
review.
Appraising and Analyzing Sources
Through analysis, you can determine the value of a
source for a particular study. Analysis must take place
in two stages. The first stage involves the critical
appraisal of individual studies. The process of

TABLE 6-5  Literature Summary Table
Author and Year Purpose Framework Sample
appraising individual studies, including the steps, is
detailed in Chapter 18. During the critical appraisal,
relevant content is clearly identified in the articles and
sources are sorted into a sophisticated system of
categories.
Conducting an analysis of sources to be used in a
research proposal requires some knowledge of the
subject to be critiqued, some knowledge of the research
process, and the ability to exercise judgment in evalu-
ation (Pinch, 1995, 2011). However, the critical
appraisal of individual studies is only the first step in
developing an adequate review of the literature. Any
written literature review that simply appraises indi-
vidual studies paragraph by paragraph is inadequate.
A literature review that is a series of paragraphs, in
which each paragraph is a description of a single study
with no link to other studies being reviewed, does
not provide evidence of adequate analysis of the
literature.
Analysis requires manipulation of what you are
finding, literally making it your own (Garrard, 2011).
Pinch (1995) was the first nurse to publish a strategy
to synthesize research findings using a literature
summary table. We modified this table by adding two
columns that are useful in sorting information from
studies into categories for analysis (see Table 6-5). In
2001, Pinch published a modified table to use in trans-
lating research findings into clinical innovations. Hart
(2009) provides examples of table formats that may
be helpful at different points during the review. The
reference management software may allow you to
generate these tables from information you record
about each study.
Another way to manipulate the information you
have retrieved and transform it into knowledge is
mapping (Hart, 2009). Nurse educators teach concep-
tual mapping to their students to encourage students
to make connections among facts and principles that
they are learning (Vacek, 2009). The same strategy
applied to a literature review is to classify the sources
CHAPTER 6  Review of Relevant Literature 109
Measurement Treatment Results Findings
and arrange them into some type of format that
requires you to become familiar with key concepts
(Hart, 2009). The map may connect studies with
similar methodologies or key ideas.
The second stage of analysis involves making com-
parisons among studies. This analysis allows you to
critically appraise the existing body of knowledge in
relation to the research problem. You will be able to
determine (1) theoretical formulations that have been
used to explain how the variables in the problem influ-
ence one another, (2) what methodologies have been
used to study the problem, (3) the methodological
flaws in previous studies, (4) what is known about the
problem, and (5) what the most critical gaps in the
knowledge base are. The information gathered by
using the table format shown in Table 6-5 or displayed
in a conceptual map can be useful in making these
comparisons. Various studies addressing a research
problem have approached the examination of the
problem from different perspectives. They may have
organized the study from different theoretical perspec-
tives, asked different questions related to the problem,
selected different variables, or used different designs.
Pay special attention to conflicting findings as they
may provide clues for gaps in knowledge that repre-
sent researchable problems.
Sorting Your Sources
Relevant sources (theoretical and empirical) are orga-
nized for inclusion in the different chapters of the
research proposal. The sources to be included in the
review of literature chapter are organized to reflect
the current knowledge about the research problem.
Those sources that provide background and signifi-
cance for the study are included in the introduction
chapter. Certain theoretical sources establish the
framework for the study. Other relevant sources
become the basis for defining research variables and
identifying assumptions and limitations. Content
from methodologically strong studies is used to direct
110 UNIT TWO  The Research Process
Claim Reason 1
Evidence 1
Figure 6-3  Building the logical argument. (Adapted from W. C. Booth, G. G. Colomb, & J. M. Williams [2008]. The craft of research [3rd ed.]. Chicago,
IL: University of Chicago Press.)
the development of the research design, guide the
selection of measurement methods, influence data col-
lection and analysis, and provide a basis for interpreta-
tion of finding. (Refer back to Table 6-1 to review
contributions of the literature to each part of the
research process.)
Synthesizing Sources
Synthesis of sources involves clarifying the meaning
obtained from the sources as a whole in the review of
the literature section of your proposal. “Integration is
making connections between ideas, theories, and
experience” (Hart, 2009, p. 8).Through synthesis and
integration, one can cluster and connect ideas from
several sources to develop a personal overall view of
the topic. Garrard (2011) describes this personal level
of knowledge as ownership, as “being so familiar with
what has been written by previous researchers that you
know clearly how this area of research has progressed
over time and across ideas” (p. 7). Synthesis is the key
to the next step of the review process, which is devel-
oping the logical argument that supports the research
problem you intend to address. Booth et al. (2008)
details the process of constructing the argument as
beginning with stating a claim and the supporting
reasons. The reviewer must also include adequate
information so that the reader agrees that the reasons
are relevant to the claim. For each reason, the reviewer
provides evidence to support the reasons. Thinking at
this level and depth prepares you for outlining the
written review. Figure 6-3 provides a visual represen-
tation of an argument that can be developed through
a written review. The writer/reviewer supports each
claim with evidence so that the reader can accept the
conclusion the reviewer has made. For example, the
reviewer has synthesized several sources related to
medication adherence and is presenting the argument
for developing patient-focused medication adherence
Reason 2 Reason 3 Conclusion
Evidence 2 Evidence 3
intervention. The following outline could be devel-
oped for this argument.
Claim 1: Interventions to promote medication adher-
ence must incorporate the hypertensive patient’s
perspective.
Reason 1: Provider-focused interventions have not
resulted in long-term improvement in medica-
tion adherence.
Evidence 1: Description of studies of provider-
focused interventions and their outcomes
Reason 2: Patients who do not adhere to an exter-
nally imposed medication regimen (the target
population) may be less likely to use an inter-
vention that is externally imposed.
Evidence 2: Description of studies in which patients
failed to return for appointments during a trial
of an electronic device to promote adherence
Reason 3: Medication adherence requires behavior
change that must be incorporated into the
patient’s life.
Evidence 3: Theoretical principles of behavior
change that recommend individualization of
interventions to meet unique patient needs
Conclusion 1: Using a participatory approach to
develop individual strategies for promoting med-
ication adherence is an important first step to
improving patient outcomes.
Writing the Review of Literature
Writing Suggestions
Clarity and cohesion are characteristics of scholarly
writing. If you have followed the steps for reviewing
the literature in this chapter, you want to demonstrate
your synthesis and ownership of the literature by
clearly presenting your argument. Rather than using

direct quotes from an author, you should paraphrase
his or her ideas. Paraphrasing involves expressing the
ideas clearly and in your own words. The meanings of
these sources are then connected to the proposed
study. Last, the reviewer combines, or clusters, the
meanings obtained from all sources to determine the
current knowledge of the research problem (Pinch,
1995, 2001). Reviews that lack clarity and cohesion
reflect deficits in synthesis more than deficits in written
abilities.
Start each paragraph with a theme sentence that
describes the main idea of the paragraph or makes a
claim. Present the relevant studies as evidence of the
main idea or claim and end the paragraph with a con-
cluding sentence that transitions to the next claim.
Each paragraph can be compared to a train with an
engine (theme sentence), linked cargo cars (sentences
with evidence), and a caboose (summary sentence
linking to next paragraph).
Organization of Written Reviews
The readability and flow of a literature review are
determined by its structure (Cronin, Ryan, & Cough-
lan, 2008). A review of the literature can be organized
as series of claims with supporting reasons and claims
(Booth et al., 2009) or it can be organized according
to the research framework. Theories supporting the
research framework may be presented in the first
section. This then allows the writer/reviewer to use the
framework as an outline for the section on empirical
literature. In the section on empirical literature, sub-
heading can identify each concept. Then in these sub-
sections, studies most relevant to that concept can be
discussed. The availability of studies will directly
affect the length of the literature review. In newer
areas of research, you may not have as many sources
per concept.
The purpose of the written literature review is to
establish a context for your study. The literature review
for a study may have four major sections: (1) the
introduction, (2) a discussion of theoretical literature,
(3) a discussion of empirical literature, and (4) a
summary. The introduction and summary are standard
sections, but you will want to organize the discussion
of sources in a way that makes sense for the topic. You
may organize the discussion by themes, your logical
argument, or your research framework.
Introduction
The introduction to the literature review indicates the
focus or purpose of the study, identifies the purpose
of the literature review, and presents the organiza-
tional structure of the review. You should make clear
CHAPTER 6  Review of Relevant Literature 111
in this section what you will and will not cover. If
you are taking a particular position or developing a
logical argument for a particular perspective on the
basis of the literature, make this position clear in the
introduction.
Discussion of Theoretical Literature
The theoretical literature section contains concept
analyses, models, theories, and conceptual frame-
works that support the research purpose. Concepts,
definitions of concepts, relationships among concepts,
and assumptions are presented and analyzed to build
a theoretical knowledge base for the study. This
section of the literature review is sometimes used to
present the framework for the study and may include
a conceptual map that synthesizes the theoretical lit-
erature (see Chapter 7 for more detail on developing
frameworks).
Discussion of Empirical Literature
The presentation of empirical literature should be
organized by concepts or organizing topics. In the
past, the researcher was expected to present the
purpose, sample size, design, and specific findings for
each study reviewed with a scholarly but brief critique
of the study’s strengths and weaknesses, This approach
is expected less commonly now. Use the synthesis and
key ideas related to each concept that you developed
in the previous stage of the review to organize this
section. In addition to the synthesis, you want to incor-
porate the strengths and weaknesses of the overall
body of knowledge rather than a detailed presentation
and critical appraisal of each study. The findings from
the studies should logically build on each other so that
the reader can see how the body of knowledge in the
research area evolved.
Evidence from multiple studies is combined to
reveal the current state of knowledge in relation to a
particular concept or study focus (topic area). Con-
flicting findings and areas of uncertainty are explored.
Similarities and differences in the studies should be
identified. Gaps and areas needing more research are
discussed. A summary of findings in the topic area is
presented, along with inferences, generalizations, and
conclusions you have drawn from your review of the
literature. A conclusion is a statement about the state
of knowledge in relation to the topic area. This should
include a discussion of the strength of evidence avail-
able for each conclusion.
The reviewer who becomes committed to a particu-
lar viewpoint on the research topic must maintain the
ethical standard of intellectual honesty. The content
from reviewed sources should be presented honestly,
112 UNIT TWO  The Research Process
not distorted to support the selected problem. Research-
ers frequently read a study and wish that the author
had studied a slightly different problem or that the
study had been designed or conducted differently.
However, the reviewers must recognize their own
opinions and must be objective in presenting
information.
The defects of a study need to be addressed, but it
is not necessary to be highly critical of another
researcher’s work. The criticisms must focus on the
content that is in some way relevant to the proposed
study and to be stated as possible or plausible explana-
tions, so that the criticisms are more neutral and schol-
arly than negative and blaming.
Summary
The summary consists of a concise presentation of the
current knowledge base for the research problem.
Other literature reviews conducted in relation to your
field of research should be discussed. The gaps in the
knowledge base are identified, with a discussion of
how the proposed study will contribute to the develop-
ment of knowledge in the defined field of research.
The summary concludes with a statement of how your
study will contribute to the body of knowledge in this
field of research.
Refining the Written Review
You complete the first draft of your review of the lit-
erature and breathe a sigh of relief before moving onto
the next portion of the research proposal. Before
moving on, you need to read, review, and refine your
review. Set the review aside for 24 hours and then read
it aloud. You may identify awkward sentences in this
way that you might overlook when reading the review
silently. Ask a fellow student or trusted colleague to
read the review and provide constructive feedback.
Use the criteria and guiding questions in Table 6-6 to
evaluate your literature review.
Checking References
Sources that will be cited in a paper or recorded in a
reference list should be cross-checked two or three
times to prevent errors. Questions that will identify
common errors are displayed in Box 6-1. To prevent
these errors, check all the citations within the text of
your literature review and each citation in your refer-
ence list. Typing or keyboarding errors may result in
inaccurate information. You may omit some informa-
tion, planning to complete the reference later, and then
forget to do so. Downloading citations from a database
directly into a reference management system and
using the system’s manuscript formatting functions
TABLE 6-6  Characteristics of High-Quality
Literature Reviews
Criteria Guiding Questions
Coverage Did the writer provide evidence of having
reviewed sufficient literature on the
topic?
Does the review indicate that the writer is
sufficiently well informed about the topic
and has identified relevant studies?
Understanding Does the written review indicate that the
writer has understood and synthesized
what is being studied?
Have similarities and differences of what
was found been described?
Coherence Does the writer make a logical argument
related to the significance of the topic
and the gap to be addressed by the
proposed study?
Accuracy Does the writer’s attention to detail give the
reader confidence in the conclusions of
the review?
Box 6-1 Checking to Avoid Common
Reference Citation Errors
Does every direct quotation have a citation
that includes the author’s name, year, and
page number?
Are authors’ names spelled the same way in
the text and in the reference list?
Are the years on citations in the text the same
as the years for the reference on the
reference list?
Are the citations on the reference list
complete so that the reference can be
retrieved?
Does every source cited in the text have a
corresponding citation on the reference list?
Is every reference on the reference list cited in
the text?
reduces some errors but does not eliminate all of
them. Use your knowledge and skills to enhance your
technology use; relying on technology alone will not
create a quality manuscript.
Example of a Literature Review
An excerpt from the literature review of the research
report of a quasi-experimental study is provided here

“When comfort is enhanced, patients are better able to
be successful in their health-seeking behaviors (HSBs).
Scholtfeldt (1975) included internal behaviors and exter-
nal behaviors in her definition of HSBs, and both of these
behaviors are incorporated into Comfort Theory.
Therefore, it was hypothesized that increased comfort
would lead to a reduction in anxiety, stress, and
depression.
Guided Imagery (GI)
“In the context of this study, GI is defined as the use of
the imagination to bring about positive mind/body
responses (Rossman, 2000). It is a cognitive process that
evokes and uses many senses: sight, sound, smell, taste,
and touch and also the senses of movement. All of these
senses together produce regenerative changes in the
mind and body (Achterberg, 1985). GI is a program of
instructions meant to help people acquire a state of
psychological and physiological ease through muscular
relaxation and positive mental images, relieving the
discomforts provoked by symptoms associated with
mood disorders (Apóstolo, 2007). Increasing adaptive
responses in depressive individuals requires replacing
the negative processes of thinking with a more positive
cognitive style (Achterberg, 1985; Rossman, 2000). GI
is a complementary nursing intervention that can be
implemented in addition to other therapeutic approaches
to mood disorders. Studies show that focusing the imag-
ination in a positive way can result in a state of ease,
encouragement, and mood regulation, all of which allow
the patient to reestablish a state of physical and mental
health (Rossman, 2000).
Guided Imagery and Depression
“Currently, it is thought that good body functioning is
accompanied by positive thoughts, whereas pathological
body functioning is accompanied by negative and repeti-
tive thoughts (Damásio, 2004). In the brain, a thought,
idea, or mental image work as “emotionally competent
stimuli.” These stimuli, whether prescribed by biologic
evolution or learned have the capacity to produce
certain patterns of homeostasis. The state of sadness is
accompanied by a reduced number of positive mental
images and by more excessive attention to those images.
When persons have the experience of positive thoughts,
CHAPTER 6  Review of Relevant Literature 113
their mind represents more than well-being; it also rep-
resents well-thinking. On the contrary, feeling sadness
is associated not only with sickliness but also with an
inefficient way of thinking, concentrated around a
limited number of ideas of loss (Damásio, 2004).
“In GI, positive mental images and positive affective
experiences can counteract the depression rumination
spiral (Folkman & Moskowitz, 2000). This process
works as an adaptive alternative to decompensation,
raising the mood, and relieving depressive symptoms.
Therefore, GI contributes to antirumination strategies
that, as Nolen-Hoeksema (1991, 2000) states, are debil-
itating. Positive mental images have a relaxing effect and,
consequently, a psychophysiologic and cognitive effect
(Singer, 2006). When depressed individuals have access
to positive mental images and to a state of body relax-
ation, they are able to reorient their thoughts away from
unpleasant stimuli. Thus, positive thoughts contribute to
an improvement in feelings about oneself and the world.
“Results from the empirical literature indicated that
GI was effective in improving mood states in individuals
with a variety of illnesses. Sloman (2002) conducted a
community-based nursing study in 56 people with
advanced cancer. Progressive muscle relaxation and GI
training revealed significant decreases in depression.
Campbell-Gillies (2004) used a program including posi-
tive mental images and music with 45 women with
breast cancer. Her findings revealed that GI decreased
depression and anxiety over a six-cycle period of che-
motherapy. McKinney, Antoni, Kumar, Times, and
McCabe (1997) used GI combined with music with 28
healthy adults and reported significant decreases in
depression, fatigue, and total mood disorders between
pretest and posttest sessions. Identical outcomes were
revealed in the study of Watanabe et al. (2006), with a
sample of 148 healthy adults, using relaxation and posi-
tive mental images. After two sessions, positive mood
increased, and negative mood decreased. Finally, in an
experimental design, Kolcaba and Fox (1999) assessed
the effects of GI for increasing comfort over time in
patients with breast cancer going through radiation
therapy. However, no experimental studies have been
conducted for patients in a psychiatric context to
increase their comfort.” (Apostolo & Kolcaba, 2009, pp.
404-405)
114 UNIT TWO  The Research Process
as an example of a review of the literature. Apóstolo
and Kolcaba (2009) conducted a study to determine
the effect of guided imagery on comfort, depression,
anxiety, and stress among 60 hospitalized mental
health patients. The researchers began the review with
a presentation of Kolcaba’s comfort theory, and then
linked that theory to guided imagery and depression.
The guided imagery treatment group showed sta-
tistically significant improvement in their depression,
anxiety, stress, and sense of comfort. Their results
were compared with those in a matched control group.
Apóstolo and Kolcaba (2009) indicated that GI was
an effective intervention, supporting their findings
with links back to their review of the literature. The
strengths of the study were the use of matched controls
that supported internal validity and the implementa-
tion of the study in three clinical facilities that sup-
ported the external validity of the study.
KEY POINTS
• Reviewing the existing literature related to your
study is a critical step in the research process.
• A literature review is a written logical presentation
of knowledge gained from reading and analyzing
selected articles, books, conference proceedings,
and other sources.
• Information from the literature guides the develop-
ment and implementation of quantitative studies
and is incorporated throughout the research
proposal.
• With use of a systems approach, the three major
stages of a literature review are searching the litera-
ture (input), processing the literature (throughput),
and writing the literature review (output).
• The literature consists of all written sources rele-
vant to the topic you have selected.
• Two types of literature are predominantly used in
the review of literature for research: theoretical and
empirical.
• Theoretical literature consists of concept analyses,
models, theories, and conceptual frameworks that
support a selected research problem and purpose.
• Empirical literature comprises relevant studies
in journals and books as well as unpublished
studies, such as master’s theses and doctoral
dissertations.
• Searching the literature begins with a written plan
for the review that is maintained as a search history
during the first stage of the literature review.
• Searching the literature requires use of bibliograph-
ical databases and reference management systems.
• Processing the literature requires the researcher to
read the sources to be able to critically appraise,
analyze, and synthesize the information that has
been retrieved.
• A thorough, organized literature review facilitates
the development of a research proposal.
• The well-written literature review presents a logical
argument for why the research question should
be studied and in the specific way that is being
proposed.
REFERENCES
Achterberg, J. (1985). Imagery in healing: Shamanism and modern
medicine. Boston: Shambhala.
American Psychological Association (2010). Publication manual of
the American Psychological Association (6th ed.). Washington,
DC: Author.
Apóstolo, J. L. A. (2007). O imaginário conduzido no conforto de
doentes em contexto psiquiátrico. Doctoral Dissertation. Porto,
PT: Porto University.
Apóstolo, J. L. A., & Kolcaba, K. (2009). The effects of guided
imagery on comfort, depression, anxiety, and stress of psychiatric
inpatients with depressive disorders. Archives of Psychiatric
Nursing, 23(6), 403–411.
Association of Nurses in AIDS Care (2009). Position statement:
Health disparities. Retrieved from http://www.proquest.com/
en-US/default.shtml/.
Bachrach, S. (2001). Scientific journals of the future. In R. S. Berry
& A. S. Moffat (Eds.), The transition from paper: Where are we
going and how will we get there? Cambridge, MA: American
Academy of Arts & Sciences. Online report. Retrieved from
http://www.amacad.org/publications/trans.aspx/.
Booth, W. C., Colomb, G. G., & Williams, J. M. (2008). The craft
of research (3rd ed.). Chicago, IL: University of Chicago Press.
Campbell-Gillies, L. (2004). Guided imagery as treatment for
anxiety and depression in breast cancer patients: A pilot study
[Online]. A dissertation submitted in partial fulfillment of the
requirements for the degree of M.A. Psychology. Rand Afrikaans
University. Retrieved from http://etd.rau.ac.za/theses/available/
etd-10062004-095533/restricted/GIreviseddissert2003130304.
pdf/.
Cronin, P., Ryan, F., & Coughlan, M. (2008). Undertaking a litera-
ture review: A step-by-step approach. British Journal of Nursing,
17(1), 38–43.
Damásio, A. R. (2004). Ao encontro de Espinosa. As Emoções e a
neurologia do sentir (6th ed.). Mem Martins: Publicações
Europa-América.
Everett, J., Bowes, A., & Kerr, D. (2010). Barriers to achieving
glycaemic control in CSII. Journal of Diabetes Nursing, 14(5),
176–181.
Folkman, S., & Moskowitz, J. T. (2000). Positive affect and the
other side of coping. American Psychologist, 55(6), 647–654.
Garrard, J. (2011). Health sciences literature review made easy: The
matrix method (3rd ed.). Sudbury, MA: Jones & Bartlett.
Gravel, J. (2009). Bess and hearing screening: Portending the chal-
lenges in children. Seminars in Hearing, 30(2), 71–79.

Hart, C. (2009). Doing a literature review: Releasing the social
science imagination. Los Angeles, CA: Sage.
Holbrook, A., Bourke, S., Fairbairn, H., & Lovat, T. (2007). Exam-
iner comment on the literature review in Ph.D. theses. Studies in
Higher Education, 32(3), 337–356.
Kolcaba, K. Y., & Fox, C. (1999). The effects of guided imagery on
comfort of women with early stage breast cancer undergoing
radiation therapy. Oncology Nursing Forum, 26(1), 67–72.
Larsen, P. O., & von Ins, M. (2010). The rate of growth in scientific
publication and the decline in coverage provided by Scientific
Citation Index. Scientometrics, 84(3), 575–603.
McKinney, C. H., Antoni, M. H., Kumar, M., Times, F. C., &
McCabe, P. M. (1997). Effects of guided imagery and music
(GIM) therapy on mood and cortisol in healthy adults. Health
Psychology, 16(4), 390–400.
Munhall, P. L. (2012). Nursing research: A qualitative perspective
(5th ed.). Sudbury, MA: Jones & Bartlett.
Nolen-Hoeksema, S. (1991). Responses to depression and their
effects on the duration of depressive episodes. Journal of Abnor-
mal Psychology, 100(4), 569–582.
Nolen-Hoeksema, S. (2000). The role of rumination in depressive
disorders and mixed anxiety/depressive symptoms. Journal of
Abnormal Psychology, 109(3), 504–511.
Pinch, W. J. (1995). Synthesis: Implementing a complex process.
Nurse Educator, 20(1), 34–40.
Pinch, W. J. (2001). Improving patient care through use of research.
Orthopaedic Nursing, 20(4), 75–81.
CHAPTER 6  Review of Relevant Literature 115
Roguin, A. (2011). Henry Cuthbert Bazett (1885-1950)—The man
behind the QT interval correction formula. Pace, 34(3),
384–388.
Rossman, M. (2000). Guided imagery for self-healing. An essential
resource for anyone seeking wellness (2nd ed.). Tiburon, CA: H.J.
Kramer.
Scholtfeldt, R. (1975). The need for a conceptual framework. In P.
Verhonic (Ed.), Nursing research (pp. 3–25). Boston: Little &
Brown.
Scott, A. J., & Wilson, R. F. (2011). Social determinants of health
among African Americans in a rural community in the Deep
South: An ecological study. Rural and Remote Health, 11, 1634
(Online). Retrieved from http://www.rrh.org.au/.
Singer, J. L. (2006). Imagery in psychotherapy. Washington: Ameri-
can Psychological Association.
Sloman, R. (2002). Relaxation and imagery for anxiety and depres-
sion control in community patients with advanced cancer. Cancer
Nursing, 25(6), 432–435.
Tensen, B. L. (2010). Research strategies for the digital age (3rd
ed.). Boston, MA: Wadsworth.
Vacek, J. E. (2009) Using a conceptual approach with a concept
map of psychosis as an exemplar of promote critical thinking.
Journal of Nursing Education, 48(1), 49–53.
Watanabe, E., Fukuda, S., Hara, H., Maeda, Y., Ohira, H., & Shi-
rakawa, T. (2006). Differences in relaxation by means of guided
imagery in a healthy community sample. Alternative Therapies in
Health and Medicine, 13(2), 60–66.

7
CHAPTER
Frameworks
A
framework is an abstract, logical structure of
meaning that guides the development of the
study and enables you to link the findings to
the body of knowledge in nursing. Frameworks are
used in quantitative research and sometimes in qualita-
tive research. In quantitative studies, the framework
may be a testable theoretical structure or may be devel-
oped inductively from published research or clinical
observations. In most qualitative studies, the researcher
will not identify a theoretical framework. With a
grounded theory study, the researcher is attempting to
develop a theory as an outcome of the study.
Every quantitative study has a theoretical frame-
work, although some researchers do not identify or
describe the theoretical framework in the report of the
study. For example, researchers may use anatomy and
physiology knowledge to guide a study without iden-
tifying a framework. Ideally, the framework of a quan-
titative study is carefully structured, clearly presented,
and well integrated with the methodology. When you
critically appraise studies, you need to identify and
evaluate the extent to which the framework guided the
study methodology. Your ability to understand the
study findings will depend on your ability to under-
stand the logic within the framework and will deter-
mine how you will use the findings. In addition, when
you develop a quantitative study, you will need to
describe the study’s theoretical framework. To help
you build the knowledge and skills needed to critically
appraise studies and describe or develop a framework
for your own studies, this chapter begins by defining
relevant terms. The chapter also describes processes
used to examine and appraise the components of theo-
ries and presents approaches to identifying or develop-
ing a framework to guide a study.
Definition of Terms
The first step in understanding theories and frame-
works is to become familiar with the terms related to
116
  http://evolve.elsevier.com/Grove/practice/
theoretical ideas and their application. This section is
an introduction to key theoretical terms: concept, rela-
tional statement, conceptual model, theory, middle-
range theory, and study framework. In response to the
ongoing debate and varying opinions about the differ-
ences between a conceptual model and a theory or a
conceptual model and a philosophy (Meleis, 2012),
working definitions of these concepts are presented to
facilitate the application of theoretical principles to
research.
Concept
A concept is a term that abstractly describes and
names an object, a phenomenon, or an idea, thus pro-
viding it with a separate identity or meaning. As a
label for a phenomenon or composite of behavior and
thoughts, a concept is a concise way to represent the
experience or state (Meleis, 2012). An example of a
concept is the term anxiety. The concept brings to
mind a feeling of uneasiness in the stomach, a rapid
pulse rate, and troubling thoughts about future nega-
tive outcomes. Consider the concept of hope. What
feelings, expectations, and actions are expressed by
this label?
Concepts can vary in their level of abstraction. At
high levels of abstraction, concepts may have general
meanings and are sometimes referred to as constructs.
For example, a construct associated with the concept
of anxiety might be “emotional responses.” Con-
structs, because they are abstract, can subsume mul-
tiple concepts. For example, in addition to anxiety,
emotional responses could include, grief, frustration,
peace, hope, and confidence.
Relational Statements
A relational statement declares that a relationship of
some kind exists between or among two or more con-
cepts (Walker & Avant, 2011). Relational statements
provide the skeleton of a framework. Clear relational
statements are essential for constructing an integrated

framework for guiding study design. The relationships
expressed in your framework will direct the develop-
ment of your study’s objectives, questions, or hypoth-
eses. The types of relationships described by the
statements will determine the study design and the
statistical analyses needed to address the study objec-
tives, questions, or hypotheses. Mature theories, such
as physiological theories, have measurable concepts
and clear relational statements that can be tested
through research.
Conceptual Models
A conceptual model, also known as a grand theory,
is a set of highly abstract, related constructs. A con-
ceptual model broadly explains phenomena of inter-
est, expresses assumptions, and reflects a philosophical
stance. Nurse scholars have expended time and effort
to debate the distinctions between the definitions of
theory, conceptual model, conceptual framework, and
theoretical framework (Chinn & Kramer, 2011;
Meleis, 2012). For example, Watson’s theory of caring
has been identified as a theory (Meleis, 2012), a phi-
losophy (Alligood, 2010), and a conceptual model
(Fitzpatrick & Whall, 2005). In this textbook, we are
using the terms conceptual model, conceptual frame-
work and grand nursing theory interchangeably
because our purpose is not to classify theoretical
thinking. The purpose of this chapter is to provide the
information needed to use concepts, relational state-
ments, and theories or theoretical structures to develop
studies and interpret their findings.
Theory
A theory consists of an integrated set of defined
concepts, existence statements, and relational state-
ments that can be used to describe, explain, predict,
or control the phenomenon being discussed. Exis-
tence statements within a theory declare that a given
concept exists or that a given relationship occurs.
For example, an existence statement might claim
that a condition referred to as stress exists and that
there is a relationship between stress and health. As
discussed earlier, relational statements clarify the
relationship that exists between or among concepts.
It is the statements of a theory that are tested
through research, not the theory itself. Thus, identi-
fying statements within the theory is critical to the
research endeavor and forms the basis of the study’s
framework.
Scientific theories are those for which repeated
studies have validated the relationships among the
concepts. These theories are sometimes called laws for
this reason. Although few nursing and psychosocial
CHAPTER 7  Frameworks 117
theories have been validated to this extent, physiologi-
cal theories have this level of evidence and can provide
a strong basis for nursing studies.
Middle-Range Theories
Middle-range theories present a partial view of
nursing reality. Proposed by Merton (1968), a sociolo-
gist, middle-range theories are less abstract and
address more specific phenomena than grand theories
do (Peterson, 2009). They directly apply to practice
and focus on explanation and implementation. Middle-
range theories may emerge from grand theories or may
be developed inductively from research findings.
Grounded theory studies are a scientific source of
middle-range theory. Some middle-range theories
have been developed by combining nursing theories
with theories from other disciplines. Some middle-
range theories have been developed from clinical
practice guidelines. Whatever their source, middle-
range theories are sometimes called substantive theo-
ries, because they are more concrete than grand
theories.
Research Frameworks
One strategy for expressing the theoretical structure
guiding a study is to present a map or diagram of
the concepts and relational statements. Although
some writers call these diagrams conceptual maps
(Artinian, 1982; Fawcett, 1999; Newman, 1979,
1986), we are calling them research frameworks to
indicate the purpose for which they were developed.
A research framework summarizes and integrates
what we know about a phenomenon more succinctly
and clearly than a literary explanation and allows us
to grasp the bigger picture of a phenomenon. A
research framework should be supported by refer-
ences from the literature. These frameworks vary in
complexity and accuracy, depending on the available
body of knowledge related to the phenomena they
are describing. Mapping can also identify gaps in the
logic of the theory being used as a framework and
reveals inconsistencies, incompleteness, and errors.
Now that this basic knowledge of the roles of con-
cepts, relationships, and theories in research has
been supplied, the next sections provide a discussion
of the analysis and application of each of these
components.
Understanding Concepts
Concepts are often described as the building blocks
of theory. Abstract concepts are descriptive but may
not be as applicable to clinical practice or research
118 UNIT TWO  The Research Process
because of their abstractness. To make a concept
more concrete, you can identify how the concept can
be measured or observed. These measurable terms
are referred to as variables. A variable is more spe-
cific than a concept and implies that the term is
clearly defined and measurable. The word “variable”
implies that the numerical values associated with the
term vary from one instance to another. A variable
related to anxiety might be “palmar sweating,” which
the researcher can measure by assigning a numerical
value to the amount of sweat on the subject’s palm.
The links among constructs, concepts, and variables
are displayed in Figure 7-1. On the left of the figure
is the construct-to-variable continuum. The other two
sets of shapes are examples of a construct, concept,
and variable. Notice that a concept may have multi-
ple ways of being measured. For example, anxiety
could be measured by palmar sweating, the State-
Trait Anxiety Scale, or a checklist of behaviors such
as pacing, wringing one’s hands, and expressing
worries.
Defining concepts allows us to be consistent in the
way we use a term in practice, apply it to a theory,
and measure it in a study. A conceptual definition
differs from the denotative (or dictionary) definition
of a word. A conceptual definition (connotative
meaning) is more comprehensive than a denotative
definition because it includes associated meanings the
word may have. For example, a connotative definition
may associate the term “fireplace” with images of
hospitality and warm comfort, whereas the dictionary
definition would be a rock or brick structure in a
house designed for burning wood. A conceptual defi-
nition can be established through concept synthesis,
concept derivation, or concept analysis (Walker &
Avant, 2011).
Emotional Emotional
Construct
Responses Responses
Concept Anxiety Grief
Variable Palmar Score on
Sweating Grief Scale
Construct-Concept-Variable Continuum
Figure 7-1  Links among constructs, concepts, and variables.
Concept Synthesis
In nursing, many phenomena have not yet been identi-
fied as discrete entities. Recognizing, naming, and
describing these phenomena are often critical steps to
understanding the process and outcomes of nursing
practice. In your clinical practice, you may notice pat-
terns of behavior or find patterns in empirical data.
You may name a concept that emerges during data
analysis in a qualitative study. The process of describ-
ing and naming a previously unrecognized concept is
concept synthesis. Nursing studies often involve pre-
viously unrecognized and unnamed phenomena that
must be named and carefully defined. Lee and Coakley
(2011) conducted a concept synthesis of geropalliative
care. Conceptual and philosophical works had exam-
ined gerontology and palliative care separately but the
combined term lacked conceptual clarity. This lack of
clarity was viewed as hindering progress in this inter-
disciplinary area of study. Lee and Coakley (2011)
concluded that geropalliative care is “both a philo-
sophical stance and structured interdisciplinary model
of care delivery that guides care to patients and fami-
lies during the last 5 years of life, irrespective of
disease” (p. 247). They describe the critical attributes
of geropalliative care as “high risk for ineffective
pain management,” “geriatric syndromes,” “chronic,
comorbid conditions,” “risk for ineffective communi-
cation,” and “beneficial in the absence of disease” (p.
247). The context of the concept includes end-of-life
trajectories that are unpredictable, “shrinking social
networks, insurance limitations,” and “multiple set-
tings” (p. 247). Concept synthesis can be an important
element in the development of nursing theory (Walker
& Avant, 2011).
Concept Derivation
Concept derivation may occur when the researcher or
theorist finds no concept in nursing to explain a phe-
nomenon (Walker & Avant, 2011). Concepts identified
or defined in theories of other disciplines may provide
insight. In concept derivation, one concept is trans-
posed from one of field of knowledge to another. If a
conceptual definition is found in another discipline, it
must be examined to evaluate its fit with the new field
in which it will be used. The conceptual definition may
need to be modified so that it is meaningful within
nursing and consistent with nursing thought (Walker
& Avant, 2011). For example, the concept of weather-
ing, from physical material science, means the altera-
tion of a material exposed over time to sun, wind, and
precipitation. Health researchers, including nurses,
define weathering as the alteration of fundamental
 CHAPTER 7  Frameworks 119

components of the human organism when exposed TABLE 7-1  Methods of Concept Analysis
over time to stress, unsupportive environments, and
limited resources (Holzman et al., 2009). Concept Type of Concept
derivation is a creative process that can be fostered by Analysis
thinking deeply and having a willingness to learn (Author[s], Date) Unique Characteristics
about processes and theories in other disciplines. Principle-based Analysis guided by linguistic,
(Hupcey & epistemological, pragmatic, and
Concept Analysis Penrod, 2005)
Ordinary use (Wilson,
logical principles
Foci of analysis are exemplars
Concept analysis is a strategy that identifies a set of 1963) (cases) that are used to identify
characteristics essential to the connotative meaning of criteria, antecedents, and
a concept. The procedure will require you to explore consequences
the various ways the term is used and to identify a set Evolutionary method Contextual analysis of how the
of characteristics that clarify the range of objects or (Rodgers, 2000 ) concept has developed over time
ideas to which that concept may be applied (Walker in different settings
& Avant, 2011). These essential characteristics, called Hybrid (Schwartz- Contextual analysis and data
defining attributes or criteria, provide a means to dis- Barcott & Kim, collection in the field leading to
tinguish the concept from similar concepts. Several 2000) conclusions about how concept
has developed over time in
approaches to concept analysis have been described in
different settings
the nursing and healthcare literature. Because the Linguistic, pragmatic Analysis of explicit and implicit
approaches have varying philosophical foundations approach (Walker concept definitions in the
and products, nurse theorists and researchers must & Avant, 2011) literature to identify criteria,
select the concept analysis approach that best suits antecedents, and consequences
their purposes in a specific situation. For example, a for pragmatic purposes in practice
researcher attempting to measure the phenomenon of and research
barriers may choose Walker and Avant’s (2011) Simultaneous (Haase, Examines closely related concepts to
approach because of its product of a theoretical defini- Britt, Coward, distinguish their unique meanings
tion. Deepened understanding of the theoretical defini- Leidy, & Penn, as well as areas of overlap
1992)
tions allows the researcher to critically appraise the
construct validity of instruments or methods being
considered to measure the concept in a study. An
emerging nurse theorist may choose Rodger’s evolu-
tionary approach to understand the contextual influ-
ences on and changes in the meaning of the concept TABLE 7-2  Concept Analyses Using
over time. Several writers have compared the different Different Methods
types of concept analysis (Cronin, Ryan, & Coughlan,
Method of
2010; Duncan, Cloutier, & Bailey, 2007; Hupcey &
Concept Analyzed Concept Analysis
Penrod, 2005; Risjord, 2009; Weaver & Mitcham,
2008). Several concept analysis strategies used in Advanced nursing Principle-Based
practice (Ruel & (Hupcey &
nursing and health care are listed in Table 7-1. In addi-
Motyka, 2009) Penrod, 2005)
tion, a number of concept analyses have been pub- Dance in mental health Hybrid (Schwartz-
lished in the nursing literature, such as those listed in nursing (Ravelin, Barcott & Kim,
Table 7-2. The following section provides an example Kylma, & Korhonen, 2000)
of how one group of researchers developed a concep- 2006)
tual definition for the concept frailty through an analy- Clinical reasoning Evolutionary method
sis of published definitions. (Simmons, 2010) (Rodgers, 2000)
Acculturation in Filipino Linguistic, Pragmatic
Importance of a Conceptual Definition: immigrants (Serafica, (Walker & Avant,
An Example 2011) 2011)
Ethical demand in Simultaneous
With the growth in the proportion of the population
nursing (Mårtenson, (Haase, Britt,
that are over age 65 years, there is increased urgency Fägerskiöld, Runeson, Coward, Leidy, &
in preventing frailty among older people who live in & Berterö, 2009) Penn, 1992)
the community (Gobbens, Luijkx, Wijnen-Sponselee,
120 UNIT TWO  The Research Process

& Schols, 2010). An extensive literature review was Literary Form

done to identify conceptual and operational definitions Social support and spirituality influence coping.
of frailty. From their review of 18 conceptual defini-
tions, Gobbens et al. concluded that “frailty is a Diagram Form
Social Support
dynamic state affecting an individual who experiences
losses in one or more domains of human functioning Coping
(physical, psychological, social) that are caused by the Spirituality
influence of a range of variables and which increases
the risk of adverse outcomes” (p. 85). Key
The definition allows researchers to distinguish SS-Measurement of Social Support
frailty from disability and comorbidity. The refine- Mathematical Form SP-Measurement of Spirituality
SS(x) + SP(y) + K = C C-Measurement of Coping
ment of the conceptual definition has limited value x –Strength of relationship between
without a practical way to measure it (Gobbens et al., social support and coping
2010). Having found no operational definition that y –Strength of relationship between
spirituality and coping
was a logical fit with the conceptual definition pro-
posed, the researchers identified their next step to be Figure 7-2  Literary, diagrammatic, and mathematical forms of a simple
developing an operational definition of frailty consis- statement.
tent with the conceptual definition.

Literary Form
Examining Relational
When social support and spirituality exceed perceived stress,
Statements effective coping occurs.
Understanding relational statements is essential for
ensuring consistency between the research frame- Diagram Form Social Support
work, study design, and statistical analyses. Relational
(–)
statements in a research framework can be described Perceived Stress Coping
by their characteristics.
Spirituality
Characteristics of Relational Statements
Relational statements describe the direction, shape, Key
strength, symmetry, sequencing, probability of occur- Mathematical Form SS-Measurement of Social Support
SP-Measurement of Spirituality
rence, necessity, and sufficiency of a relationship (SS)(SP)/PS = C
PS-Measurement of Perceived Stress
(Walker & Avant, 2011). One statement may have C-Measurement of Coping
several of these characteristics; each characteristic is C = or > 1.0 indicates effective coping
not exclusive of the others. Statements may be
expressed as words in a sentence (literary form), as Figure 7-3  Literary, diagrammatic, and mathematical forms of a
complex statement.
shapes and arrows (diagram form), or as equations
(mathematical form). In nursing, the literary and dia-
grammatic forms of statements are used most fre-
quently. Figure 7-2 displays the three forms of simple when the product of spiritual perspective and social
statements of relationships among spiritual perspec- support is greater than perceived stress, coping will be
tive, social support, and coping. Notice in the diagram greater. When the product is less than perceived stress,
in Figure 7-2 that dotted arrows are used to indicate a coping is diminished.
relationship about which little is known. The mathe-
matical form indicates the same relationships. Figure Direction
7-3 provides literary, diagrammatic, and mathematical The direction of a relationship may be positive,
forms of a more complex statement among the previ- negative, or unknown (Fawcett, 1999). A positive
ous concepts with the addition of perceived stress. linear relationship implies that as one concept
Notice the change in the arrow between perceived changes (the value or amount of the concept increases
stress and coping. The arrow is darker and heavier or decreases), the second concept will also change in
until spiritual perspective and social support modify the same direction. For example, the literary state-
the relationship. The mathematical form proposes that ment, “As stress increases (A), the risk of illness (B)

increases,” expresses a positive relationship. This
positive relational statement could also be expressed
as “As stress decreases (A), the risk of illness (B)
decreases.”
A negative linear relationship implies that as one
concept changes, the other concept changes in the
opposite direction. For example, the statement “As
relaxation (A) increases, blood pressure (B) decreases”
expresses a negative linear relationship. The negative
relationship could also be expressed as “As relaxation
decreases (A), blood pressure (B) increases.”
The nature of the relationship between two con-
cepts may be unknown because it has not been studied
or because study findings have been conflicting. For
example, consider two studies that included the con-
cepts of coping and social support. Tkatch et al. (2011)
found that the number of people in the social networks
of African American patients in cardiac rehabilitation
(N = 115) and their health-related social support had
weak, but statistically significant relationships with
their coping efficacy. In contrast, Jackson et al. (2009)
found nonsignificant relationships among social
support and coping in a longitudinal study of 88
parents who had a child with a brain tumor. From the
findings, we can conclude that although there is evi-
dence that a relationship exists between these two
concepts, the studies examining that relationship have
conflicting findings. Conflicting findings may result
from differences in the researchers’ definitions and
measurements of the two concepts in various studies.
Another reason for conflicting findings may be that an
unidentified confounding variable exists that changes
the relationship between coping and social support.
Whatever the reason, conflicting findings about a rela-
tionship between concepts could be indicated by a
question mark. Figure 7-4 has diagrams representing
the directional characteristics of relationships. More
+
A B
Positive relationship
–
A B
Negative relationship
? Low
A B
Unknown relationship
Figure 7-4  Directions of relational statements.
CHAPTER 7  Frameworks 121
details on diagramming relational statements are pre-
sented in Chapter 8.
Shape
Most relationships are assumed to be linear, and sta-
tistical tests are conducted to identify linear relation-
ships. In a linear relationship, the relationship
between the two concepts remains consistent regard-
less of the values of each of the concepts. For example,
if the value of A increases by 1 point each time the
value of B increases by 1 point, then the values con-
tinue to increase at the same rate whether the value is
2 or 200. The relationship can be illustrated by a
straight line, as shown in Figure 7-5.
Relationships can also be curvilinear or form some
other shape. In a curvilinear relationship, the rela-
tionship between two concepts varies according to
the relative values of the concepts. The relationship
between anxiety and learning is a good example of a
curvilinear relationship. Very high or very low levels
of anxiety are associated with low levels of learning,
whereas moderate levels of anxiety are associated
with high levels of learning (Bierman, Comijs,
Rijmen, Jonker, & Beekman, 2008). This type of
relationship is illustrated by a curved line, as shown
in Figure 7-6.
Strength
The strength of a relationship is the amount of varia-
tion explained by the relationship. Some of the varia-
tion in a concept, but not all, is associated with
variation in another concept (Fawcett, 1999). In dis-
cussing the strength of a relationship, researchers
sometimes use the term effect size. The effect size
explains how much “effect” variation in one concept
has on variation in a second concept.
High
Concept A
High
Concept B
Figure 7-5  Linear relationship.
122 UNIT TWO  The Research Process
High
Learning
Low
Anxiety
Figure 7-6  Curvilinear relationship.
–1 0 +1
Strong
No
Negative
Relationship
Relationship
Figure 7-7  Relational statement strength.
Researchers usually determine the strength of the
relationship between concepts by correlational analy-
sis. The mathematical result of the analysis is a cor-
relation coefficient such as the following:
r = 0.35
The statistic r is the coefficient obtained by perform-
ing the statistical procedure known as Pearson’s
product-moment correlation (see Chapter 23). A value
of 0 indicates no strength, whereas a value of +1 or
−1 indicates the greatest strength (see Figure 7-7).
When the correlation is large, a greater portion of
the variation can be explained by the relationship; in
others, only a moderate or a small portion of the varia-
tion can be explained by the relationship. For example,
Tkatch et al. (2011) found a relationship of r = 0.22
(p < 0.05) between health-related social support and
coping efficacy. The strength of the relationship meant
that a small portion of the variance in coping was
explained by variations in health-related social support
(see Chapter 23 for additional information).
The + or − does not have an impact on the strength
of the relationship. For example, r = −0.35 is as strong
as r = +0.35. A weak relationship is usually considered
one with an r value of 0.1 to 0.3; a moderate relation-
ship is one with an r value of 0.31 to 0.5; and a strong
relationship is one with an r value greater than 0.5.
The greater the strength of a relationship, the easier it
A B
Asymmetrical relationship
A B
or
A B
A B
Symmetrical relationship
High
Figure 7-8  Relational statement symmetry.
is to detect relationships between the variables being
studied. This idea will be explored further in the chap-
ters on sampling, measurement, and data analysis
Strong (Chapters 15, 16, and 23, respectively).
Positive
Relationship Symmetry
Relationships may be symmetrical or asymmetrical. In
an asymmetrical relationship, if A occurs (or
changes), then B will occur (or change); but there may
be no indication that if B occurs (or changes), A will
occur (or change). An asymmetrical relationship is not
reversible (Fawcett, 1999). You can think of this rela-
tionship as a one-way street, with influence going only
in one direction. Bradford and Petrie (2008) found an
asymmetrical relationship between internalized ideal
of a thin body and body dissatisfaction in their study
of disordered eating among 236 first-year, female
college students. The internalized thin body ideal
affected body dissatisfaction, but the reverse was not
true (see Figure 7-8).
A symmetrical relationship is complex and con-
tains two statements, such as if A changes, B will
change and if B changes, A will change (Fawcett,
1999). You may think of the relationship between
these concepts as a two-way street with influence
going in both directions. These relationships may also
be called reciprocal or reversible. An example is the
symmetrical relationship between quality of life and
perceived health among patients after heart surgery
(Mathisen et al., 2007). The researchers found that
quality of life influenced perceived health and that
perceived health influenced quality of life, indicating
a symmetrical relationship between the concepts. A
second example comes from the study by Bradford
and Petrie (2008). They found a symmetrical (recipro-
cal) relationship between depression and disordered
eating in their study with adolescent females (see
Figure 7-8).

A If A, then probably B
B event of A. For example, one could state that there is
Concurrent relationship
A B
Sequential relationship
Passage of Time
Figure 7-9  Relational statement sequencing.
Sequencing
The amount of time that elapses between one concept
and another is stated as the sequential nature of a
relationship. If the two concepts occur simultaneously,
the relationship is concurrent (Fawcett, 1999). When
there is a change in one concept, there is change in the
other at the same time. If one concept changes and the
second concept changes later, the relationship is
sequential. Another example from Bradford and Pet-
rie’s study (2008) was a hypothesized concurrent rela-
tionship between internalizing the thin body ideal and
dissatisfaction with one’s body. Their findings pro-
vided mixed support for the relationship. They also
hypothesized that depression would predict later path-
ological eating. This sequential relationship was sup-
ported by their findings. These relationships are
diagrammed in Figure 7-9.
Probability of Occurrence
A relationship can be deterministic or probabilistic
depending on the degree of certainty that it will occur.
Deterministic (or causal) relationships are state-
ments of what always occurs in a particular situation.
For example, among Thai patients with heart failure,
nurse researchers found that symptom status, social
support, general health perception, and functional
status had causal relationships with health-related
quality of life (Phuangphaka, Veena, Chanokporn, &
Sloan, 2008). Scientific laws are another example of
deterministic relationships (Fawcett, 1999). A causal
relationship is expressed as follows:
If A, then always B
A probability statement expresses the probability
that something will happen in a given situation
(Fawcett, 1999). This relationship is expressed as
follows:
CHAPTER 7  Frameworks 123
Probability statements are tested statistically to deter-
mine the extent of probability that B will occur in the
greater than a 50% probability that a patient who has
an indwelling catheter for 1 week will experience a
urinary bladder infection. This probability could be
expressed mathematically as follows:
p > 0.50
The p is a symbol for probability. The > is a symbol
for “greater than.” This mathematical statement asserts
that there is more than a 50% probability that the
relationship will occur.
Necessity
In a necessary relationship, one concept must occur
for the second concept to occur (Fawcett, 1999). For
example, one could propose that if sufficient fluids are
administered (A), and only if sufficient fluids are
administered, the unconscious patient will remain
hydrated (B). This relationship is expressed as follows:
If A, and only if A, then B
In a substitutable relationship, a similar concept
can be substituted for the first concept and the second
concept will still occur. For example, a substitutable
relationship might propose that if tube feedings are
administered (A1), or if hyperalimentation is admin-
istered (A2), the unconscious patient can remain
hydrated (B). This relationship is expressed as follows:
If A1, or if A 2, then B
Sufficiency
A sufficient relationship states that when the first
concept occurs, the second concept will occur, regard-
less of the presence or absence of other factors
(Fawcett, 1999). A statement could propose that if a
patient is immobilized in bed longer than a week, he
or she will lose bone calcium, regardless of anything
else. This relationship is expressed as follows:
If A, then B, regardless of anything else
A contingent relationship will occur only if a third
concept is present. For example, a statement might
claim that if a person experiences a stressor (A), the
person will manage the stress (B), but only if she or
he uses effective coping strategies (C). The third
concept, in this case effective coping strategies, is
referred to as an intervening (or mediating) variable.
124 UNIT TWO  The Research Process

TABLE 7-3  Characteristics of Relationships Abstract General Proposition Abstract

Type of
Relationship Descriptive Statement
Positive linear As A increases, B increases.
As A decreases, B decreases. Specific Proposition
Negative linear As A increases, B decreases.
As A decreases, B increases.
Unknown linear As A changes, B may change.
Curvilinear At a specific level, as A changes, B
changes to a similar degree.
At another specific level, as A changes, B
Less Abstract Hypothesis Less Abstract
changes to a greater or lesser extent.
Asymmetrical As A changes, B changes.
Figure 7-10  Abstract to concrete—general proposition to hypothesis.
As B changes, A does not change.
Symmetrical As A changes, B changes.
As B changes, A changes. Statements at varying levels of abstraction that
Concurrent When A changes, B changes at the same
express relationships between or among the same con-
time.
Sequential After A changes, B changes.
ceptual ideas can be arranged in hierarchical form,
Causal If A occurs, B always occurs. from general to specific. This arrangement allows you
Probabilistic If A occurs, then probably B. to see (or evaluate) the logical links among the various
Necessary If A occurs, and only if A occurs, B levels of abstraction. Statement sets link the relation-
occurs. ships expressed in the framework with the hypotheses,
Sufficient If A occurs, and only if A, B occurs. research questions, or objectives that guide the meth-
Substitutable If A1 or A2 occurs, B occurs. odology of the study. The following excerpts provide
Contingent If A occurs, then B occurs, but only if C an example of the more abstract theoretical proposi-
occurs. tion that provided the basis for the hypothesis that was
tested in a study by Xu, Floyd, Westmaas, and Aron
(2010). These researchers asked 66 former and 74
Intervening variables can affect the occurrence, current smokers to select self-expanding activities,
strength, or direction of a relationship. A contingent such as a new romantic relationship or a new job, that
relationship can be expressed as follows: occurred in the time immediately preceding attempts
to stop smoking.
If A, then B, but only if C
Proposition
Being able to describe relationships among the con-
cepts is an important first step in identifying, evaluat- “Because both self-expansion and nicotine appear to
ing, and developing research frameworks. Table 7-3 trigger feelings of reward that originate from the
provides a summary of the characteristics of relational same parts of the brain, it is conceivable that the
statements. Remember that a single statement may reward from self-expanding experiences could (at
have multiple descriptive characteristics. least partly) supplant or act as a substitute for the
reward elicited by nicotine or other addictive drugs.”
Statement Hierarchy (Xu et al., 2010, p. 296)
Statements about the same two conceptual ideas can
Hypothesis
be made at various levels of abstractness. The state-
ments found in conceptual models and grand theories “We hypothesized that smokers would report greater
(general propositions) are at a high level of abstrac- success abstaining from smoking if they had engaged
tion. Statements found in middle-range theories in self-expanding activities prior to their quit attempt.”
(specific propositions) are at a moderate level of (Xu et al., 2010, p. 296)
abstraction. Hypotheses, which are a form of state- Xu and colleagues reported that, “former smokers
ment, are at a low level of abstraction and are specific. reported a mean of 4.60 self-expanding events (SD =
As statements become less abstract, they become 4.1) in the 2-month period prior to their successful
narrower in scope (Fawcett, 1999), as shown in quit attempt” (p. 298).
Figure 7-10.
 CHAPTER 7  Frameworks 125

energy fields interacting with the environmental

Grand Theories energy fields.
Most disciplines have several conceptual models, each In addition to concepts specific to the theory, nurse
with a distinctive vocabulary. Table 7-4 lists some of theorists include the metaparadigm or domain con-
the conceptual models or grand theories in nursing. cepts of nursing: person, health, environment, and
These philosophical theories of nursing vary in their nursing (Chinn & Kramer, 2011). For example, Roy,
level of abstraction and the breadth of phenomena they Orem, and Rogers all include the construct health in
explain. Each provides an overall picture, or gestalt, their models but define it in different ways (see Table
of the phenomena they explain. It is not their purpose 7-5). The definitions inherent in grand nursing theories
to provide detail or to be specific. guide nursing practice by providing the characteristics
Nurses have developed a number of conceptual of the person and environment that nurses should
models or grand nursing theories. For example, Roy’s assess. For example, Orem’s theory of self-care
(1988) model describes adaptation as the primary phe- describes factors to be assessed as universal self-care
nomenon of interest to nursing. This model identifies requisites, developmental self-care requisites, and
the constructs she considers essential to adaptation health deviation requisites. A nurse whose assessment
and how these constructs interact to produce is guided by Roy’s theory would gather assessments
adaptation (Roy & Andrews, 2008). Orem (2001) con- related to the adaptive modes of physiological, self-
sidered self-care to be the phenomenon central to concept, role function, and interdependence. Theories
nursing. Her model explains how nurses facilitate the also guide nursing practice by defining the specific
self-care of clients. Rogers (1970, 1990) regarded goal of practice (characteristics of health) and describ-
human beings as the central phenomenon of interest ing the nursing interventions (descriptions of nursing)
to nursing. In her model, human beings are viewed as to achieve the outcomes. Most are not directly testable
through research and thus cannot be used alone as the
framework for a study (Fawcett, 1999; Walker &
Avant, 2011). Application of grand nursing theories to
TABLE 7-4  Selected Grand Nursing Theories research is discussed later in this chapter.
(in Alphabetical Order by  
Author Name)
Descriptive Label Application of
Author (Year) of the Theory
Henderson, Virginia (1964) Nursing as Promoting Patient
Middle-Range Theories
Independence with 14 Middle-range theories are useful in both research and
Activities of Daily Living practice. Middle-range theories are less abstract than
Johnson, Dorothy (1974) Behavioral Systems grand theories and closer to the day-to-day substance
King, Imogene (1981) Interacting Systems Theory of clinical practice, a characteristic that explains why
of Nursing (includes they can be called substantive theories. As a result,
middle-range theory of middle-range theories guide the practitioner to under-
Goal Attainment) standing the client’s behavior, enabling interventions
Leininger, Madeline (1997) Transcultural Nursing Care, that are more effective. Because of their usefulness in
Sunrise Model of Care
practice, some writers refer to middle-range theories
Orem, Dorothea (2001) Self-Care Deficit Theory of
Nursing
as practice theories. Middle-range theories are used
Neuman, Betty (Neuman & Systems Model of Nursing more commonly than grand theories as frameworks
Fawcett, 2002) for research. Researchers need to carefully consider
Newman, Margaret (1986) Health as Expanding which middle-range theory to use as a study frame-
Consciousness work and which aspects of this theory will be tested
Nightingale, Florence (1859) Environmental Health by their study. Tables 7-6 and 7-7 identify some of the
Parse, Rosemarie (1992) Human Becoming Theory middle-range theories more frequently used as frame-
Peplau, Hildegard (1988, Interpersonal Relations works in the past 10 years of nursing research.
1991) Theory Middle-range theories have been developed from
Rogers, Martha E (1970) Unitary Human Beings
clinical insights, elements of existing theories not pre-
Roy, Calista (1988) Adaptation Model
Watson, Jean (1979) Philosophy and Science of
viously related, outcomes of qualitative studies, or
Caring conceptual models. Kolcaba (2009) noted that, although
comfort had been the goal of nursing interventions for
126 UNIT TWO  The Research Process

TABLE 7-5  Comparison of the Domain Concepts—Person, Health, Environment, and Nursing—within
Three Grand Theories
Domain Rogers: Unitary Orem: Self-Care Deficit Theory
Concept Human Beings Roy: Adaptation Model of Nursing
Person Irreducible, indivisible Biopsychosocial beings who adapt Human beings with eight universal needs,
energy fields; open to their environment two developmental needs, and the
systems that are integral potential for six health deviation
with the environment self-care needs; ability to provide
self-care is affected by the
conditioning factors of age, gender,
health, and the availability of resources
Health Symphonic interaction of Return to or maintenance of Extent to which person can meet own
the human and adaptation, through activation of universal, development, and health-
environment the regulator and cognator deviation self-care requisites
systems and of the physiological,
self-concept, role function, and
interdependence adaptive modes
Environment Irreducible, indivisible Source of focal and contextual Source of resources and other factors that
energy fields; open stimuli that may require can affect person’s ability to meet
systems that are integral responses by the person to self-care needs
with the persons maintain adaption
Nursing Promote symphonic Promote the person’s adaptation by Design and implement systems of care to
interaction in the modifying focal and contextual decrease or remove the gap between
human-environmental stimuli and facilitating use of self-care capacity and self-care needs;
energy field the adaptive modes may provide wholly compensatory
(physiological, self-concept, role care, partially compensatory care, or
function, interdependence) supportive-educative care

Compiled from Banfield (2011); Denyes, Orem, & Bekel SozWiss (2001); and Rogers (1990).

many years, the concept had never been defined pre-
cisely. Her theory of comfort began with a concept
analysis of comfort during her graduate nursing educa-
tion (Kolcaba & Kolcaba, 1991). As she continued her
doctoral studies, over the space of a year or more, she
developed a taxonomy of comfort that provides the
basis for describing types of comfort (Kolcaba, 1991)
and an operational definition (Kolcaba, 1992). Through
continued scholarly inquiry, Kolcaba’s middle-range
theory of comfort was developed (Kolcaba, 1994), as
were several instruments to measure different types of
comfort (http://www.thecomfortline.com/).
A more recent middle-range theory, the Synergy
Model for Patient Care, was developed by members
and leaders of the American Association of Critical
Care Nurses to identify the unique contributions of
nurses to patient and family outcomes (Hardin, 2009).
The basic premise of the model is that “patient char-
acteristics drive nurse competencies” (Curley, 2007,
p. 2). The patient characteristics include stability,
complexity, vulnerability, predictability, resiliency,
resource availability, participation in care, and
participation in decision making. The nurse assesses
the extent to which each of the characteristics are
present in a specific patient and family. Based on the
assessment, the nurse will use specific practice com-
petencies. The practice competencies in the model are
clinical judgment, clinical inquiry, caring practices,
response to diversity, advocacy/moral agency, facilita-
tion of learning, collaboration, and systems thinking.
When synergy exists among the eight patient charac-
teristics and eight nursing practice competencies,
optimal patient outcomes are more likely to be found.
Extensive work has been done applying the Synergy
Model in practice, nursing education, and professional
development (Curley, 2007).
A specific type of middle-range theory is interven-
tion theory. Intervention theories seek to explain the
dynamics of a patient problem and exactly how a
specific nursing intervention is expected to change
patient outcomes. Currently, these new theories are
tentative, but some will likely become substantive in
the future. Intervention theories are discussed in detail
in Chapter 14.
 CHAPTER 7  Frameworks 127

TABLE 7-6  Selected Middle-Range Nursing Theories Developed 2001 through 2011
Name of Theory
Coping of relatives of patients
in intensive care units
Flight nursing expertise
Humor in clinical nurse
specialist–patient
interactions
Music, mood, and movement
to improve health outcomes
Nursing intellectual capital
Nursing presence
Professional resilience and
career persistence
Quality oncology nursing
practice
Self-care management for
vulnerable populations
Story theory
Facilitated sense-making with
families of patients in
intensive care
Adaptation to chronic pain
Caregiver stress
Spiritual empathy
Theorist Source
Johansson, I., Hildingh, C., Fridlund, B., & Ahlstrom, G. (2006). Theoretical model of coping
among relatives of patients in intensive care units: A simultaneous concept analysis. Journal of
Advanced Nursing, 56(5), 463-471.
Reimer, A. P., & Moore, S. M. (2011). Flight nursing expertise: Toward a theory of flight nursing.
Journal of Advanced Nursing, 66(5), 1183-1192.
McCreaddie, M., & Wiggins, S. (2009). Reconciling the good patient persona with problematic and
non-problematic humour: A grounded theory. International Journal of Nursing Studies, 46(8),
1079-1091.
Murrock, C. J., & Higgins, P. A. (2009). The theory of music, mood, and movement to improve
health outcomes. Journal of Advanced Nursing, 65(10), 2249-2257.
Covell, C. L. (2008). The middle-range theory of nursing intellectual capital. Journal of Advanced
Nursing, 63(1), 94-103.
McMahon, M., & Christopher, K. (2010). Toward a middle-range theory of nursing presence.
Nursing Forum, 46(2), 71-82.
Hodges, H. F., Troyan, P. J., & Keely, A. C. (2010). Career persistence in baccalaureate-prepared
acute care nurses. Journal of Nursing Scholarship, 42(1), 83-91.
Radwin, L., Cabral, H., & Wilkes, G. (2009). Relationships between patient-centered cancer nursing
interventions and desired health outcomes in the context of the health care system. Research in
Nursing & Health, 32(1), 4-17.
Dorsey, C. J., & Murdaugh, C. L. (2003). Theory of self management for vulnerable populations.
Journal of Theory Construction and Testing, 7(2), 43-49.
Smith, M. J., & Liehr, P. (2005). Story theory: Advancing nursing scholarship. Holistic Nursing
Practice, 19(6), 272-276.
Davidson, J. E. (2010). Facilitated sensemaking: A strategy and new middle-range theory to support
families of intensive care unit patients. Critical Care Nurse, 30(6), 28-39.
Dunn, K. (2004). Toward a middle range theory of adaptation to chronic pain. Nursing Science
Quarterly, 17(1), 78-84.
Tsai, P. (2003). A middle-range theory of caregiver stress. Nursing Science Quarterly, 16(2),
137-145.
Chism, L., & Magnan, M. (2009). The relationship of nursing students’ spiritual care perspectives to
their expressions of spiritual empathy. Nursing Science Quarterly, 48(11), 597-605.

Appraising Theories and
Research Frameworks
Nurses examine and evaluate theories to determine
their maturity, applicability for practice, and useful-
ness for research. The evaluation of theories is com-
plicated by the availability of at least seven sets of
possible evaluative criteria (Fawcett, 2005). In keeping
with the purpose of this book, we focus our discussion
on critically appraising research frameworks in pub-
lished studies.
Critical Appraisal of a Research Framework
During the process of critically appraising a study,
the first task related to the research framework is to
describe it. This task is easier when the researchers
explicitly identified the framework. For example,
Benkert, Hollie, Nordstrom, Wickson, and Bins-
Emerick (2009) identified the Interactional Model of
Client Health Behavior (IMCHB; Cox, 2003) as the
theoretical framework for their study. This team of
nurse researchers conducted a descriptive correla-
tional study of patient satisfaction with the care pro-
vided by nurse practitioners (NPs) in primary care.
The study participants (N = 100) were African Ameri-
can routinely seen by NPs in three urban clinics.
Benkert et al. (2009) did not provide a diagram of the
framework but carefully defined each of the frame-
work’s concepts and described the pencil-and-paper
instruments used to measure them.
Other researchers, such as Dailey (2009), do not
identify the frameworks in their studies. In Dailey’s
(2009) study of social stressors and strengths of low-
income African American mothers and infant birth
weight, the framework can be inferred from the litera-
ture review and the study variables. Consider the fol-
lowing excerpts from the literature review of the
research report to identify the relational statements:
128 UNIT TWO  The Research Process

TABLE 7-7  Selected Middle-Range Nursing Theories Cited 2001 through 2011 but Developed Earlier
Name of Theory (Theorist[s], Year)
Caring (Swanson, 1991)
Prescriptive pain management theory
(Good & Moore, 1996)
Theory of chronic sorrow (Eakes, Burke,
& Hainsworth, 1998)
Theory of comfort (Kolcaba, 1994)
Uncertainty in illness theory (Mishel,
1988)
Self-transcendence theory (Reed, 1991)
Theory of unpleasant symptoms (Lenz,
Pugh, Milligan, Gift, & Suppe, 1997)
Need-driven, dementia-compromised
behavior (Algase et al., 1996).
Experiencing transitions (Meleis, Sawyer,
Im, Messiasis, & Schumacher, 2000)
Recent Citation
Andershed, B., & Ollson, K. (2008). Review of literature related to Kristen Swanson’s
middle-range theory of caring. Scandinavian Journal of Caring Sciences, 23(3),
598-610.
Good, M., Anderson, G., Ahn, S., & Cong, X. (2005). Relaxation and music reduced
pain following intestinal surgery. Research in Nursing & Health, 28(3), 240-251.
Gordon, J. (2009). An evidence-based approach for supporting parents experiencing
chronic sorrow. Pediatric Nursing, 35(2), 115-119.
Kolcaba, K., & DiMarco, M. (2005). Comfort theory and its application to pediatric
nursing. Pediatric Nursing, 31(3), 187-194.
Lin, L., Yeh, C., & Mishel, M. (2010). Evaluation of a conceptual model based on
Mishel’s theories of uncertainty in illness in a sample of Taiwanese parents of
children with cancer: A cross-sectional questionnaire survey. International Journal of
Nursing Studies, 47(12), 1510-1524.
Reed, P. G., & Rousseau, E. (2007). Spiritual inquiry and well-being in life-limiting
illness. Journal of Religion, Spirituality, and Aging, 19(4), 81-98.
Brant, J., Beck, S., & Miaskowski, C. (2010). Building dynamic models and theory to
advance the science of symptom management research. Journal of Advanced Nursing,
66(1), 228-240.
Kovach, C. R., Noonan, P. E., Schlidt, A. M., & Wells, T. (2005). A model of
consequences of need-driven, dementia-comprised behavior. Journal of Nursing
Scholarship, 37(2), 134-140.
Reedy, S., & Blum, K. (2010). Applying middle-range nursing theory to bariatric surgery
patients: Experiencing transitions. Bariatric Nursing and Surgical Patient Care, 5(1),
35-43.

Describing the research framework may be easier

• The link between stressful life circumstances and
birth outcomes….” (p. 340).
• “Several studies have identified associations
between maternal experiences of racial
discrimination and low-birth-weight deliveries….”
(pp. 340-341).
• “Maternal stress associated with trauma exposure
has been linked to … health outcomes” (p. 341).
• “[J]ust as important to learn how women at risk
adapt to the negative influences…. [E]ffects of
maternal stress on health outcomes should be
balanced by efforts to study personal resources
that may be influential to maternal health”
(p. 341).
• “Supportive networks contribute to the well
being of African American women during
pregnancy” (p. 341).
• “Levels of spirituality affect perinatal health”
(p. 341).
• “Numerous researchers have studied
socioeconomic status (SES), common medical
conditions during pregnancy, and health practices
in relation to birth weight outcomes” (p. 341).
if you draw a diagram of the concepts and relation-
ships among them. In the Dailey (2009) study, the
study purpose contained four constructs. The purpose
was to explore “the extent to which social stressors,
personal resources, and known perinatal risk factors
predict infant birth weight in a sample of urban low-
income African American women” (Dailey, 2009, p.
341). These constructs were the basis for a concep-
tual map of the research framework. Figure 7-11
presents our diagrams of the constructs, concepts, and
relationships among the concepts. In the figure, the
first diagram shows that the constructs of social
stressors, personal resources, and perinatal risk
factors influence birth outcomes. The constructs were
derived from the concepts discussed in the article.
The second diagram identifies the concepts that
Dailey (2009) used that were related to each con-
struct. For example, the construct of social stressors
was examined by measuring discrimination, trauma,
and socioeconomic states. Although not identified
as such, the conceptual and operational definitions
of the concepts were easily identified in the article.
Table 7-8 displays the definitions of three concepts
as examples.
 CHAPTER 7  Frameworks 129

TABLE 7-8  Conceptual and Operational Definitions of Selected Concepts from Dailey’s (2009) Study
Concept Conceptual Definition
Trauma Trauma is experiencing, either as a witness or
participant, an event that involves actual or
threatened serious injury to or death of self or
others (Dailey, 2009).
Spirituality Spirituality is defined as making meaning that
empowers the self, supports coping, and
promotes equanimity (Dailey, 2009).
Discrimination Discrimination is being excluded, exploited, or
treated with prejudice and having one’s rights
and freedoms restricted.
Operational Definition
Trauma History Questionnaire guided the assessment of
life trauma with a resulting score of 0-23, with higher
scores indicative of more types of trauma.
Spirituality was measured using the Spiritual Perspective
Scale (Reed, 1987). The operational definition is the
summed score on the SPS, ranging from 10 to 60.
Discrimination was measured using the Everyday
Discrimination Scale. The operational definition is the
summed score on the scale, ranging from 9-54.

Research Framework with Constructs 2. Do the concepts reflect the constructs identified in
the framework? Some frameworks may not iden-
Social Developed from Dailey (2009) tify constructs and may have only concepts.
Stressors
3. Do the variables reflect the concepts identified in
Personal Birth the framework?
Resources Outcomes 4. Are the conceptual definitions validated by refer-
ences to the literature?
Perinatal Risk 5. Are the propositions (relational statements) logical
Factors
and defensible?
Research Framework with Concepts The next step in critically appraising a study frame-
work is to evaluate the extent to which the framework
Discrimination guided the methodology by asking the following
Trauma question:
Socioeconomic Status
1. Do the operational definitions reflect the concep-
Spirituality Infant Birth tual definitions?
Social Support Weight 2. Do the hypotheses, questions, or objectives reflect
Substance Use the constructs and/or concepts in the propositions
Lack of Prenatal Care of the framework?
Medical Conditions 3. Is the design appropriate to test the propositions of
the framework?
Figure 7-11  Derived research framework for Dailey’s (2009) study of When a framework guides the methodology of a
stressors and strengths that predict infant birth weight. study, the answer to these questions will be “yes.”
Some researchers may describe a theory or theories
to provide context for the study but may not use the
framework to guide the methodology. Bond et al.
From your description of the framework, you are (2011) conducted a study of how nurse researchers
ready to examine the logical structure of the frame- use theory by reviewing research reports in seven
work. Meleis’s (2012) criteria for critically appraising leading journals over 5 years. In 837 of the 2184
theories include assessing the clarity and consistency research reports (38%), the researchers included a
of the logical structure. When the following questions theoretical framework, either a nursing theory or a
about clarity and consistency can be answered “yes,” theory from another discipline. Of these 837 reports,
the framework has a strong logical structure: 93% contained evidence that the theory had been inte-
1. Are the definitions of constructs consistent with the grated into the study methodology. Bond et al. have
theorist’s definitions? This question is asked only documented that, when identified, the study frame-
if the researchers link their framework to a parent work most likely will be used to guide the methodol-
theory. (The parent theory is the theory from ogy. The more prevalent problem is that nursing
which the researchers have selected the constructs researchers often do not identify the theories guiding
for their study.) their studies.
130 UNIT TWO  The Research Process
The final step in critically appraising a study frame-
work is to decide the extent to which the researcher
connected the findings to the framework by asking the
following questions:
1. Did the researcher interpret the findings in terms of
the framework?
2. Are the findings for each hypothesis, question, or
objective consistent with the relationships pro-
posed by the framework?
Even in studies clearly guided by a research frame-
work, the findings may not be discussed in terms of
what they mean in relation to the framework. Find-
ings that are consistent with the framework are evi-
dence of the framework’s validity, and this point
should be noted in the discussion. When the findings
are not consistent with the research framework,
researchers should discuss the possible reasons for
this disconnect. One reason may be a lack of construct
validity (see Chapter 16). The instruments used may
not have measured the constructs/concepts of the
study framework adequately and accurately. Other
possible reasons are that the framework was based on
assumptions that were not true for the population
being studied and that the framework does not repre-
sent the reality of the phenomena being studied in this
specific sample.
Developing a Research
Framework for Study
Developing a framework is one of the most important
steps in the research process but, perhaps, also one of
the most difficult. A research report in a journal often
contains only a brief presentation of the study frame-
work because of page limitations. As a result, you may
gain little insight into the careful, thoughtful, pro-
longed work required to develop a research frame-
work. Part of developing as a researcher is your
commitment and motivation to learn the skills needed
to perform this thoughtful work.
You have identified a research problem and are
thinking about the proposed study’s methodology. The
guideline you are using for developing the proposal
indicates that you need a research framework for a
study. Where do you start? This section presents three
basic approaches to beginning the process of con-
structing a study framework: (1) identifying an exist-
ing theory from nursing or another discipline, (2)
synthesizing a framework from research findings, and
(3) proposing a framework from clinical practice. The
final steps of constructing a research framework are
the same no matter the approach used to start the
construction and are discussed after the presentation
of the approaches.
Identifying and Adapting an Existing Theory
Take another look at the research reports you have
read related to your topic. Which theories have others
used when studying this area? In your exploration,
include studies on your topic of interest that have been
done with populations other than your own. For
example, researchers have used several health behav-
ior and psychological theories to guide studies related
to medication adherence. Kalichman et al. (2011)
tested an adherence intervention based on the conflict
theory of decision making with persons with HIV/
AIDS. For her randomized controlled trial (RCT) of a
nursing intervention to promote medication adherence
in cardiac patients, Gould (2011, p.119) used self-
regulation theory based on the “Common Sense Model
of illness perception for self regulation of care” devel-
oped by Leventhal, Diefenbach, and Leventhal (1992).
Schaffer and Tian (2004) used the protection motiva-
tion theory (Prentice-Dunn & Rogers, 1986) to
examine adherence to preventive medication use of
persons with asthma. Existing theories can provide
insights into how the topic has been studied and the
range of perspectives available on a given research
topic.
Review theory textbooks and middle-range theory
publications to examine the applicability of other
nursing theories that might provide insight to your
research problem (Tomey & Alligood, 2006). For
example, Orem’s (2001) theory of self-care could be
used as a framework for a study of the effect of acuity-
based staffing assignments on hospital length of stay
for patients after a coronary artery bypass graft surgery.
The acuity-based staff assignments could be supported
by Orem’s nursing systems—wholly compensatory
care, partially compensatory care, and supportive-
educative care. Examples of theories for specific
topics are provided in Table 7-9. Prior to making a
final decision about a theory, you should read primary
sources written by the theorists to ensure that your
topic is a conceptual and pragmatic fit with the con-
cepts, definitions of concepts, assumptions, and prop-
ositions of the theory.
Synthesis from Research Findings
Developing a theory or a framework from research
findings is the most accepted strategy of theory devel-
opment (Meleis, 2012). The research-to-theory strat-
egy, an inductive approach, begins by identifying
relevant studies. For example, to develop a framework
for a study of coping with HIV infection, Gray and

TABLE 7-9  Potential Theories for Different Research Topics
Research Topic
Light and noise as influences on patient recovery in acute
care settings
Peer support group for adolescents to decrease their use of
illegal substances
Sleep quality and dementia progression in long-term care
facilities
Medication adherence of men on antihypertensive
medications
Screening for diabetes mellitus among women with a history
of gestational diabetes
Task shifting in low-resource settings
Fear of HIV infection and sexual behavior of young gay men
New graduate nurses’ clinical competency and use of
high-fidelity simulation in baccalaureate nursing education
Children’s postoperative pain and parental anxiety
Nurse retention and environmental factors in rural hospitals
Cason (2002) conducted an integrated review of 30
studies. Studies were included if the subjects were
persons living with HIV infection, and the variables
included coping and/or psychosocial factors related to
coping. The researchers extracted study characteristics
and documented them in a literature summary table
(see Chapter 6). They carefully reviewed the findings
of each study for relationships among concepts. The
researchers became immersed in the published find-
ings by diagramming the relationships on a large piece
of paper and creating tables of each relational state-
ment with the studies supporting the statement.
Through this process, they identified concepts and
relational statements for inclusion in the proposed
study’s framework.
Proposing a Framework from Practice Experiences
As members of a practice discipline, nurses may
develop research frameworks from their clinical expe-
riences. Nurses in practice may make generalizations
about patients’ responses as they repeatedly provide
care to different types of patients. Nurses who reflect
on practice may, over time, realize underlying prin-
ciples of human behavior that guide their choices of
interventions. Meleis (2012) notes that a nurse may
have nagging questions about why certain situations
persist or how to improve patient or organizational
outcomes that lead to tentative theories. For example,
a novice researcher who worked in a newborn inten-
sive care unit was convinced from her clinical experi-
ences that a mother’s frequent visits to the hospital
were related to her infant’s weight gain. The nurse’s
CHAPTER 7  Frameworks 131
Theory (Theorist[s], Year)
Environmental theory (Nightingale, 1859, 1946, 1979)
Roy’s adaptation model (Roy & Andrews, 2008)
Cognitive brain reserve (Stern, 2009)
Information motivation behavioral skills model (Fisher, Williams,
Fisher, & Mallory, 1999)
Cardiometabolic Model (Ruhl, 2009)
Theory of self-regulating teams (Millwood, Banks, & Riga, 2010)
Theory of planned behavior (Ajzen, 1991)
NLN-Jeffries simulation framework (Jeffries et al., 2007)
Philosophical approach: Externalist perceptual view of pain (Pesut &
McDonald, 2007)
Theory of structural empowerment (Kanter, 1977)
Research Framework from Clinical Practice
Concepts Bonding Thriving
Number of Mother’s
Variables
Visits
Infant Weight Gain
Figure 7-12  Research framework from clinical experience.
ideas could be diagrammed as the lower set of rela-
tionships shown in Figure 7-12.
The relationship she identified consisted of two
concrete ideas: number of mother visits and weight
gain. From the perspective of research, these ideas are
variables. Instead of starting with a framework and
linking the concepts of the framework to possible
study variables, she was starting with variables and
needed to identify the concepts that the variables rep-
resented. She reviewed the literature and looked for
explanations for why visits by the mother were impor-
tant and what happened when a mother visited the
baby. As she reflected on what she read, she realized
that maybe the visits promoted bonding or attachment.
The researcher continued to reflect on her experiences
and remembered that when babies failed to gain
weight or lost weight, they were sometimes labeled as
“failing to thrive.” Wording that more positively, she
decided the concept related to weight gain was thriv-
ing. On the basis of her clinical experiences and her
132 UNIT TWO  The Research Process
thinking processes, the researcher began to learn more
about theories of bonding and used what she learned
to develop a framework for a study related to bonding
and thriving of newborns in the neonatal intensive care
(see Figure 7-12).
Research frameworks rarely develop from only one
source of knowledge. Nurse researchers often combine
existing theories, research findings, and insights from
their clinical experiences into a framework for a study.
For example, to study adherence to blood pressure
medications among older Chinese immigrants, Li,
Wallhagen, and Froelicher (2010) derived their model
from Becker’s Health Belief Model (1974), findings
from preliminary studies, hypertension literature, and
clinical experience. Whatever your approach to begin-
ning the process, once you have possible concepts and
relationships, you are ready to move through the
remainder of the process to develop a framework that
is explicitly expressed in the final research report. A
study framework developed in this way is considered
tentative theory until research findings provide evi-
dence to support the relationships as diagrammed.
Tentative theories are those that are developed from
other theories, research findings, and clinical practice
and that, as of yet, do not have new evidence to support
their relational statements.
Defining Relevant Concepts
Concepts are selected for a framework on the basis of
their relevance to the phenomenon you are studying.
The concepts included in the research framework
should reflect the problem statement and the literature
review of the proposal. Each concept included in a
framework must be defined conceptually. Conceptual
definitions may be found in existing theoretical works
and quoted in the proposal with sources cited. Con-
ceptual definitions may also be found in published
concept analyses, previous studies using the concept,
or the literature associated with an instrument devel-
oped to measure the concept. Although the instrument
itself is an operational definition of the concept, the
writer will often provide a conceptual definition on
which the instrument development was based. (See
Chapter 8 for more extensive discussion of conceptual
and operational definitions for study variables.) When
acceptable conceptual definitions are not available,
you should perform concept synthesis or concept anal-
ysis to develop them.
Developing Relational Statements
The next step in framework development is to link all
of the concepts through relational statements. If you
began with an existing theory, theoretical propositions
may have already been identified. If you synthesized
research findings, you have evidence that supports
relationships between or among some or all of the
concepts. This evidence supports the validity of each
relational statement. This support must include a dis-
cussion of previous quantitative or qualitative studies
(or both) that have examined the proposed relationship
and published observations from the clinical practice
perspective.
Extracting relational statements from the literary
text of an existing theory, published research, or clini-
cal literature can be a daunting task. The following
procedure describes how to do so: Select the portion
of the theory, research report, or clinical literature that
discusses the relationships among the concepts rele-
vant to your study. Write single sentences that connect
one concept to another. Change the sentence to a
diagram of the relationship, similar to those presented
earlier in this chapter (see Figures 7-3 and 7-4). Con-
tinue this process until all the relationships in the text
have been expressed as diagrams. Burns and Grove
(2011) illustrate this process in greater detail in
Chapter 7 of their book.
If statements relating the concepts of interest are
not available in the literature, statement synthesis is
necessary. Develop statements that propose specific
relationships among the concepts you are studying.
You may gain the knowledge for your statement syn-
thesis through clinical observation and integrative lit-
erature review (Walker & Avant, 2004).
Developing Hierarchical Statement Sets
A hierarchical statement set is composed of a spe-
cific proposition (relational statement) and a hypoth-
esis or research question. The specific proposition may
be preceded by a more general proposition when an
existing theory was the source of the framework. The
proposition is listed first, with the hypothesis or
research question immediately following. In some
cases, more than one hypothesis or research question
may be developed for a single proposition. This state-
ment set indicates the link between the framework and
the methodology.
Constructing a Conceptual Map
A conceptual map is a visual representation of a
research framework. With the concepts defined and
the relational statements diagrammed, you are ready
to visually represent the framework for your study.
The framework may be limited to only the concepts
that you are studying or may be inclusive of other
related concepts that are not going to be studied or
measured. When the framework includes concepts that
 CHAPTER 7  Frameworks 133

are not included in the specific study being proposed,
you must clearly identify the portion of the framework
being used.
From a practical standpoint, first arrange the rela-
tional statements you have diagrammed from left to
right with outcomes located at the far right. Concepts
that are elements of a more abstract construct can be
placed in a frame or box. Sets of closely interrelated
concepts can be linked by enclosing them in a frame
or circle. Second, using arrows, link the concepts in a
way that is consistent with the statement diagrams you
previously developed. Every concept should be linked
to at least one other concept. Third, examine the
framework diagram for completeness by asking your-
self the following questions:
1. Are all of the concepts in the study also included
on the map?
2. Are all the concepts on the map defined?
3. Does the map clearly portray the phenomenon?
4. Does the map accurately reflect all the statements?
5. Is there a statement for each of the links portrayed
by the map?
6. Is the sequence of links in the map accurate?
Developing a well-constructed conceptual map
requires repeated tries, but persistence pays off. You
may need to reexamine the statements identified. Are
there some missing links? Are some of the links inac-
curately expressed?
As the map takes shape and begins to seem right,
show it to trusted colleagues. Can they follow your
logic? Do they agree with your links? Can they iden-
tify missing elements? Can you explain the map to
them? Seek out individuals who have experienced the
phenomenon you are mapping. Does the process
depicted seem valid to them? Find someone more
experienced than you in conceptual mapping to
examine your map closely and critically.
Shanks (2010) conducted a pilot study of the
quality of life of 15 patients with coronary artery
disease who were receiving external counterpulsation
(ECP). In the literature review section of the article,
Shanks states, “According to Rector, Anand, and Cohn
(2006), the effect of heart failure pathology on quality
of life (QOL) is mediated by symptoms such as angina
and decreased exercise tolerance” (p. 1). The model
in the study by Rector et al. is shown in Figure 7-13.
Shanks (2010) used some of these concepts and pre-
sented them in a more linear fashion in her simple
conceptual framework (see Figure 7-14). She clearly
identified the concepts of cardiac pathology,

Functional Figure 7-13  Conceptual model of the effects

Limitations of heart failure on quality of life whereby
symptoms depend on pathology and mediate
Heart the effects on quality of life. (From Rector,
Quality of
Failure Symptoms T. S., Anand, I. S., & Cohn, J. H. (2006).
Life
Pathology Relationships between clinical assessments
and patients’ perceptions of the effects of
Psychological heart failure on their quality of life. Journal of
Distress Cardiac Failure, 12(2), 88.

ECP

Cardiac Functional Quality of

Pathology Limitations Life

Cardiac SAQ 6-minute SF-36 MLHF

Output walk test Questionnaire

Figure 7-14  The construct-concept-variable continuum. ECP, External counterpulsation; MLHF, Minnesota Living with Heart Failure; SAQ, Seattle
Angina Questionnaire; SF-36 = Short Form 36 Health Survey. (From Shanks, L. C. [2010]. A pilot study to examine relationships among external
counterpulsation, cardiac output, functional capacity, and quality of life [Online exclusive]. Applied Nursing Research, 2010 Dec 28. DOI: 10.1016/
j.apnr.2010.09.002.)
134 UNIT TWO  The Research Process
functional limitations, and quality of life. Her diagram
also includes the measurements for each concept, thus
providing clear logical links to the study’s methods.
She described her conceptual framework as follows:
“[Figure 7-14] presents a conceptual model based on
the model developed by Rector et al. (2006). This
model focuses on the degree to which cardiac pathol-
ogy influences functional limitations and QOL. One
or two measures were obtained for each of the three
concepts in this model.” (Shanks, 2010, p. 2)
Her research questions reflected the conceptual
model:
“1.  What are the relationships among cardiac
pathology, functional limitations, and QOL at
Week 1 (entry into ECP) and Week 7 (end of
ECP)?
2. Is there an improvement in cardiac output at the
end of the 7-week ECP program?
3. Is there an improvement in functional limitations
at the end of the 7-week ECP program?
4. Is there an improvement in perceived QOL at
the end of the 7-week ECP program?” (Shanks,
2010, p. 2).
The participants in the study ranged in age from 50
to 75 years. Most were white, married men. More than
50% were retired and had annual incomes less than
$50,000. Shanks (2010) found no significant correla-
tions in response to research question 1. She also did
not find evidence to support improvements in func-
tional status or QOL, with one exception. There was
a statistically significant improvement in the 6-minute
walk test (research question 3). On the QOL measure,
the improvement in the 6-minute walk was further
supported by similar findings related to “physical role,
physical function, and bodily pain” (p. 4). Although
the small sample precluded generalizing the findings
to other groups, the findings indicate a need for repli-
cation with larger samples.
The product of the creative and critical thinking
that you have expended in the development of your
research framework may provide a structure for one
study or become the basis for a program of research.
Continue to consider the framework as you collect
and analyze data and interpret the findings. While
you wait to hear whether your proposal has been
funded or while your data are being collected, use the
time to expand the written description of the frame-
work and the evidence supporting its relationships
into a manuscript for publication (see Chapter 27).
When disseminated, your research framework has the
potential to make a valuable contribution to nursing
knowledge.
KEY POINTS
• A concept is a term that abstractly describes and
names an object or a phenomenon, thus providing
it with a separate identity or meaning.
• A relational statement declares that a relationship
of some kind exists between two or more
concepts.
• A conceptual model or grand theory broadly
explains phenomena of interest, expresses assump-
tions, and reflects a philosophical stance.
• A theory is a set of concepts and relational
statements explaining the relationships among
them.
• Scientific theories have significant evidence and
their relationships may be considered laws.
• Substantive theories are less abstract, can easily be
applied in practice, and may be called middle-range
theories.
• Middle-range theories may be developed from
qualitative data, clinical experiences, clinical prac-
tice guidelines, or more abstract theories.
• Tentative theories are developed from research
findings and clinical experiences and have not yet
been validated.
• A framework is the abstract, logical structure of
meaning that guides the development of the study
and enables the researcher to link the findings to
the body of knowledge used in nursing.
• Relational statements are the core of the frame-
work; it is these statements that are examined
through research.
• Every study has either an implicit or explicit theo-
retical framework.
• The steps of critically appraising a research frame-
work are (1) describing the concepts and relational
statements, (2) examining its logical structure, (3)
evaluating the extent to which the framework
guided the methodology, and (4) determining the
extent to which the researcher connected the find-
ings to the framework.
• The logical adequacy of a research framework is
the extent to which the relational statements are
clear and used consistently.

• The framework should be well integrated with the
methodology, carefully structured, and clearly pre-
sented, whether the study is physiological or
psychosocial.
• Study findings should be interpreted in light of the
research framework.
• Research frameworks may start with existing
theories, research findings, and/or clinical experi-
ences.
• The remaining steps of the process are (1) selecting
and defining concepts, (2) developing statements
relating the concepts, (3) expressing the statements
in hierarchical fashion, and (4) developing a con-
ceptual map.
• Concepts and relational statements are visually rep-
resented by a diagram of the research framework,
also called a conceptual map.
• Developing a framework for a study is one of the
most important steps in the research process.
REFERENCES
Ajzen, I. (1991). The theory of planned behavior. Organizational
Behavior and Human Decisions Processes, 50(2), 179–211.
Algase, D., Beck, C., Kolanowski, A., Whall, A., Berent, S., Rich-
ards, K., et al. (1996). Need-driven, dementia-compromised
behavior: An alternative view of disruptive behavior. American
Journal of Alzheimer’s Disease, 11(6), 10, 12–19.
Alligood, M. R. (2010). Nursing theory: Utilization & application
(4th ed.). Maryland Heights, MO: Mosby Elsevier.
Andershed, B., & Ollson, K. (2008). Review of literature related to
Kristen Swanson’s middle-range theory of caring. Scandinavian
Journal of Caring Sciences, 23(3), 598–610.
Artinian, B. M. (1982) Conceptual mapping: Development of
the strategy. Western Journal of Nursing Research, 4(4),
379–393.
Banfield, B. (2011). Nursing agency: The link between practical
nursing science and nursing practice. Nursing Science Quarterly,
2(1), 42–47.
Becker, M. (1974). The Health Belief Model and sick role behavior.
Health Education Monographs, 2(4), 409–462.
Benkert, R., Hollie, B., Nordstrom, C., Wickson, B., & Bins-
Emerick, L. (2009). Trust, mistrust, racial identity and patient
satisfaction in urban African American primary care patients
of nurse practitioners. Journal of Nursing Scholarship, 4(2),
211–219.
Bierman, E., Comijs, H., Rijmen, F., Jonker, C., & Beekman, A.
(2008). Anxiety symptoms and cognitive performance in later life:
Results from the longitudinal aging study Amsterdam. Aging &
Mental Health, 12(4), 517–523.
Bond, A., Eshah, N., Bani-Khaled, M., Hamad, A., Habashneh, S.,
Kataua, H., et al. (2011). Who uses nursing theory? A univariate
descriptive analysis of five years’ research articles. Scandinavian
Journal of Caring Sciences, 25(2), 404–409.
Bradford, J., & Petrie, T. (2008). Sociocultural factors and the
development of disordered eating: A longitudinal analysis of
CHAPTER 7  Frameworks 135
competing hypotheses. Journal of Counseling Psychology, 55(2),
246–262.
Burns, N., & Grove, S. K. (2011). Understanding nursing research
(5th ed.). Maryland Heights, MO: Elsevier Saunders.
Chinn, P. L., & Kramer, M. K., (2011). Integrated theory and
knowledge development in nursing (8th ed.). St. Louis, MO: Else-
vier Mosby.
Cox, C. L. (2003). A model of health behavior to guide studies of
childhood cancer studies (Online exclusive). Oncology Nursing
Forum, 30(5), E92–E95. Retrieved from http://ons.metapress.com/
content/c82737q205725174/fulltext.pdf/.
Cronin, P., Ryan, F., & Coughlan, M. (2010). Concept analysis in
healthcare research. International Journal of Therapy and Reha-
bilitation, 17(2), 62–68.
Curley, M. (2007). Synergy: The unique relationship between nurses
and patients: The AACN Synergy Model for Patient Care. India-
napolis, IN: Sigma Theta Tau International.
Dailey, D. (2009). Social stressors and strengths as predictors of
infant birth weight in low-income African American women.
Nursing Research, 58(5), 340–347.
Denyes, M. J., Orem, D., & Bekel SozWiss, G. (2001). Self care:
A foundational science. Nursing Science Quarterly, 14(1),
48–54.
Duncan, C., Clouteir, J. D., & Bailey, P. H. (2007). Concept analy-
sis: The importance of differentiating the ontological focus.
Journal of Advanced Nursing, 58(3), 293–300.
Eakes, G. G., Burkes, M. L., & Hainsworth, M. A. (1998). Middle-
range theory of chronic sorrow. Image: Journal of Nursing Schol-
arship, 30(2), 179–184.
Fawcett, J. (1999). The relationship of theory and research (3rd ed.).
Philadelphia, PA: F. A. Davis Company.
Fawcett, J. (2005). Criteria for evaluation of theory. Nursing Science
Quarterly, 18(2), 131–135.
Fisher, W. A., Williams, S. S., Fisher, J. D., & Malloy, T. E.
(1999). Understanding AIDS risk behavior among sexually
active urban adolescents: An empirical test of the Information-
Motivation-Behavioral Skills Mode. AIDS and Behavior, 3(1),
13–23.
Fitzpatrick, J. J., & Whall, A. J. (2005). Conceptual models of
nursing: Analysis and application (4th ed.). Upper Saddle River,
NJ: Pearson Prentice Hall.
Gobbens, R. J., Luijkx, K. G., Wijnen-Sponselee, M. T., & Schols,
J. M. (2010). Toward a conceptual definition of frail community
dwelling older people. Nursing Outlook, 58(2), 76–86.
Good, M., & Moore, S. (1996). Clinical practice guidelines as a new
source of middle-range theory: Focus on pain. Nursing Outlook,
44(2), 74–79.
Gould, K. (2011). A randomized controlled trial of a discharge
nursing intervention to promote self regulation of care for early
discharge interventional cardiology patients. Dimensions of Criti-
cal Care Nursing, 30(2), 117–125.
Gray, J., & Cason, C. L. (2002). Mastery over stress among women
with HIV infection. Journal of the Association of Nurses in AIDS
Care, 13(4), 43–57.
Haase, J., Britt, T., Coward, D., Leidy, N., & Penn, P. (1992).
Simultaneous concept analysis of spiritual perspective, hope,
acceptance, and self-transcendence. Image: Journal of Nursing
Scholarship, 24(2), 141–147.
136 UNIT TWO  The Research Process
Hardin, S. (2009). The AACN Synergy Model. In S. J. Peterson &
T. S. Bredow (Eds.), Middle range theories: Application to
nursing research (2nd ed.) (pp. 99–114). Philadelphia, PA:
Wolters Kluwer/Lippincott Williams & Wilkins.
Henderson, V. (1964). The nature of nursing. American Journal of
Nursing, 64(8), 62–68.
Holzman, C., Eyster, J., Klevn, M., Messer, L. C., Kaufman, J. S.,
Laraia, B. A., et al. (2009). Maternal weathering and risk of
preterm delivery. American Journal of Public Health, 99(10),
1864–1871.
Hupcey, J., & Penrod, J. (2005). Concept analysis: Examining the
state of the science. Research for Theory and Nursing Practice,
19(2), 197–208.
Jackson, A., Enderby, K., O’Toole, M., Thomas, S., Ashley, D.,
Rosenfeld, J., et al. (2009). The role of social support in families
coping with childhood brain tumor. Journal of Psychosocial
Oncology, 27(1), 1–24.
Jeffries, P., Carter, N., Chambers, K., Childress, R. M., Childs, J.,
Decker, S., et al. (2007). Simulation in nursing education: From
conceptualization to evaluation. New York City, NY: NLN
Publications.
Johansson, I., Hildingh, C., Fridlund, B., & Ahlstrom, G. (2006).
Theoretical model of coping among relatives of patients in inten-
sive care units: A simultaneous concept analysis. Journal of
Advanced Nursing, 56(5), 463–471.
Johnson, D. E. (1974). Development of theory: A requisite for
nursing as a primary health profession. Nursing Research, 23(5),
372–377.
Kalichman, S., Cherry, C., Kalichman, M., Amaral, C., White, D.,
Pope, H., et al. (2011). Integrated behavioral intervention to
improve HIV/AIDS treatment adherence and reduce HIV trans-
mission. American Journal of Public Health, 101(3), 531–538.
Kanter, R. M. (1977). Men and women of the corporation. New
York NY: Basic Books.
King, I. M. (1981). A theory for nursing: Systems, concepts, and
processes. New York, NY: Wiley.
Kolcaba, K. (1991). A taxonomic structure for the concept comfort.
Image: The Journal of Nursing Scholarship, 23(4), 237–240.
Kolcaba, K. (1992). Holistic comfort: Operationalizing the con-
struct as a nurse-sensitive outcome. Advances in Nursing Science,
15(1), 1–10.
Kolcaba, K. (1994). A theory of holistic comfort for nursing.
Journal of Advanced Nursing, 19(6), 1176–1184.
Kolcaba, K. (2009). Comfort. In S. J. Peterson & T. S. Bredow
(Eds.), Middle-range theories: Application to nursing research
(2nd ed.) (pp. 254–272). Philadelphia, PA: Wolters Kluwer/
Lippincott Williams & Wilkins.
Kolcaba, K., & Kolcaba, R. (1991). An analysis of the concept of
comfort. Journal of Advanced Nursing, 16(11), 1301–1310.
Lee, S. M., & Coakley, E. E. (2011). Geropalliatative care: A
concept synthesis. Journal of Hospice and Palliative Care, 13(4),
242–248.
Leininger, M. M. (1997). Overview of the Theory of Culture Care
with the ethnonursing research method. Journal of Transcultural
Nursing, 8(2), 32–54.
Lenz, E., Pugh, L. C., Milligan, R. A., Gift, A. G., & Suppe, E.
(1997). The middle-range theory of unpleasant symptoms: An
update. Advances in Nursing Science, 19(3), 14–27.
Leventhal, H., Diefenbach, M., & Leventhal, E. (1992). Illness
cognition using common sense to understand treatment adherence
and affect cognition interactions. Cognitive Therapy Research,
16(2), 143–163.
Li, W.-W., Wallhagen, M., & Froelicher, E. (2010). Factors predict-
ing blood pressure control in older Chinese immigrants to the
United States of America. Journal of Advanced Nursing, 66(10),
2202–2212.
Mårtenson, E. K., Fägerskiöld, A. M., Runeson, I. V., & Berterö,
C. M. (2009). The ethical demand in nursing: The Scandinavian
perspective. Nursing Science Quarterly, 22(3), 281–288.
Mathisen, L., Andersen, M., Veenstra, M., Wahl, A., Hanestad, B.,
& Fosse, E. (2007). Quality of life can both influence and be an
outcome of general health perceptions after heart surgery. Health
& Quality of Life Outcomes, 5(27). Retrieved from http://
www.hqlo.com/content/5/1/27.
Meleis, A. I. (2012). Theoretical nursing: Development and prog-
ress (5th ed.). Philadelphia, PA: Wolters Kluwer/Lippincott Wil-
liams & Wilkins.
Meleis, A. I., Sawyer, A., Im, F., Messiasis, D., & Schumacher, K.
(2000). Experiencing transitions: An emerging middle-range
theory. Advances in Nursing Science, 23(1), 12–28.
Merton, R. K. (1968). Social theory and social structure. New York,
NY: Free Press.
Millwood, L., Banks, A., & Riga, K. (2010). Effective self regulat-
ing teams: A generative psychological approach. Team Perfor-
mance and Management, 16(1/2), 50–73.
Mishel, M. (1988). The theory of uncertainty in illness. Image:
Journal of Nursing Scholarship, 20(4), 225–232.
Neuman, B., & Fawcett, J. (2002). The Neuman Systems Model (4th
ed.). Upper Saddle River, NJ: Prentice-Hall.
Newman, M. (1979). Theory development in nursing. Philadelphia,
PA: F.A. Davis.
Newman, M. (1986). Health as expanding consciousness. St. Louis,
MO: Mosby.
Nightingale, F. (1859, 1946). Notes on nursing: What it is and what
it is not. Philadelphia, PA: Lippincott.
Orem, D. E. (2001). Nursing: Concepts for practice (6th ed.) St.
Louis, MO: Mosby Year-Book Inc.
Parse, R. (1992). Human becoming: Parse’s theory of nursing.
Nursing Science Quarterly, 5(1), 35–42.
Peplau, H. (1988, 1991). Interpersonal relations in nursing: A con-
ceptual frame of reference for psychodynamic nursing. New York,
NY: Springer Publisher.
Pesut, B., & McDonald, H. (2007). Connecting philosophy and
practice: Implications of two philosophic approaches to pain for
nurses’ expert clinical decision making. Nursing Philosophy, 8(4),
256–263.
Peterson, S. (2009). Introduction to the nature of nursing knowl-
edge. In S. J. Peterson & T. S. Bredow (Eds.), Middle-range theo-
ries: Application to nursing research (2nd ed.) (pp. 3–45).
Philadelphia, PA: Wolters Kluwer/Lippincott Williams & Wilkins.
Phuangphaka, K., Veena, J., Chanokporn, J., & Sloan, R. (2008).
Causal model of health-related quality of life in Thai patients with
heart failure. Journal of Nursing Scholarship, 40(3), 254–260.
Prentice-Dunn, S., & Rogers, R. (1986). Protection motivation
theory and preventive health: Beyond the health belief model.
Health Education Research, 1(3), 153–161.

Ravelin, T., Kylma, J., & Korhonen, T. (2006). Dance in mental
health nursing: A hybrid concept analysis. Issues in Mental Health
Nursing, 27(3), 307–317.
Rector, T. S., Anand, I. S., & Cohn, J. H. (2006). Relationships
between clinical assessments and patients’ perceptions of the
effects of heart failure on their quality of life. Journal of Cardiac
Failure, 12(2), 87–92.
Reed, P. G. (1987). Spirituality and well-being in terminally ill
hospitalized adults. Research in Nursing & Health, 10(5),
335–344.
Reed, P. G. (1991). Toward a nursing theory of self-transcendence:
Deductive reformulation using developmental theories. Advances
in Nursing Science 13(4), 64–77.
Risjord, M. (2009). Rethinking concept analysis. Journal of
Advanced Nursing, 65 (3), 684–691.
Rodgers, B. L. (2000). Concept analysis: An evolutionary view. In
B. L. Rodgers (Ed.), Concept development in nursing: Founda-
tions, techniques, and applications (2nd ed.) (pp. 77–102). Phila-
delphia, PA: W. B. Saunders.
Rogers, M. E. (1970). An introduction to the theoretical basis of
nursing. Philadelphia, PA: Davis.
Rogers, M. E. (1990). Nursing: Science of unitary, irreducible
human beings: Updated 1990. In E. A. M. Barrett (Ed.), Visions
of Rogers’ science-based nursing (pp. 5–11). New York, NY:
National League for Nursing.
Roy, C. (1988). An explication of the philosophical assumptions of
the Roy Adaptation Model. Nursing Science Quarterly, 1(1),
26–34.
Roy, C., & Andrews, H. A. (2008). Roy’s Adaptation Model for
Nursing (3rd ed.). Location: Publisher?
Ruel, J., & Motyka, C. (2009). Advanced practice nursing:
A principle-based concept analysis. Journal of the American
Academy of Nurse Practitioners, 21(17), 384–392.
Ruhl, C. (2009). Cardiometabolic health: Connecting the dots.
Nursing for Women’s Health, 13(1), 78–82.
Schaffer, S. D., & Tian, L. (2004). Promoting adherence: Effects of
theory-based asthma education. Clinical Nursing Research, 13(1),
69–89.
CHAPTER 7  Frameworks 137
Schwartz-Barcott, D., & Kim, H. S. (2000). An expansion and
elaboration of the hybrid model of concept development. In B. L.
Rogers & K. Knafl (Eds.), Concept Development in Nursing:
Foundations, Techniques, and Applications (pp. 129–159). Phila-
delphia: W. B. Saunders Company.
Serafica, R. (2011). Concept analysis of acculturation in Filipino
immigrants within health context. Nursing Forum, 46(3),
128–136.
Shanks, L. C. (2010). A pilot study to examine relationships among
external counterpulsation, cardiac output, functional capacity, and
quality of life (Online exclusive). Applied Nursing Research, 23,
1–6. Available online Dec 28. DOI: 10.1016/j.apnr.2010.09.002.
Simmons, B. (2010). Clinical reasoning: Concept analysis. Journal
of Advanced Nursing, 66 (5), 1151–1158.
Stern, Y. (2009). Cognitive reserve. Neuropsychologia, 47(10),
2015–2028.
Swanson, K. (1991). Empirical development of a middle-range
theory of caring. Nursing Research, 40(3), 161–166.
Tkatch, R., Artinian, N., Abrams, J., Mahn, J., Franks, M., Keteyian,
S., et al. (2011). Social networks and health outcomes among
African American cardiac rehabilitation patients. Heart & Lung,
40(3), 193–200.
Tomey, A. M., & Alligood, M. R. (2006). Nursing theorists and
their work (6th ed.). St. Louis, MO: Mosby Elsevier.
Walker, L. O., & Avant, K. C. (2011). Strategies for theory construc-
tion in nursing (5th ed.). Boston, MA: Prentice Hall.
Watson, J. (1979). Nursing: The philosophy and science of caring.
Boston, MA: Little Brown and Company.
Weaver, K., & Mitcham, C. (2008). Nursing concept analysis in
North America: State of the art. Nursing Philosophy, 9(1),
180–194.
Wilson, J. (1963). Thinking with concepts. Cambridge, England:
Cambridge University Press.
Xu, X., Floyd, A., Westmaas, J., & Aron, A. (2010). Self expanding
and smoking abstinence. Addictive Behaviors, 35(4), 295–301.

8
CHAPTER
Objectives, Questions, Hypotheses,
and Study Variables
R
esearchers formulate objectives, questions, and
hypotheses to bridge the gap between the more
abstractly stated research purpose and the
detailed plan for data collection and analysis. Objec-
tives, questions, and hypotheses delineate the study
variables, the relationships among the variables, and,
often, the population to be studied.
Study variables are concepts at various levels of
abstraction that are measured, manipulated, or con-
trolled in a study. Concrete concepts, such as tempera-
ture, weight, and blood pressure, are referred to as
variables in a study; abstract concepts, such as creativ-
ity, empathy, and social support, are sometimes
referred to as research concepts. Research variables
and concepts are conceptually defined, on the basis of
the study framework, and are operationally defined to
direct their measurement, manipulation, or control in
a study.
In this chapter, you will explore when objectives,
questions, or hypotheses might be developed to direct
the conduct of a study. You will also learn how to
formulate research objectives, questions, and hypoth-
eses, especially how to test different types of hypoth-
eses through research. This chapter concludes with a
discussion of different types of variables and direction
for conceptually and operationally defining variables
for a study.
Formulating Research Objectives
or Aims
Research objectives or aims are clear, concise,
declarative statements expressed in the present tense
that usually are presented following the study
purpose to specify the study focus. For clarity, an
objective usually focuses on one or two variables
(or concepts) and indicates whether the variables are
to be identified or described. Objectives can also
138
  http://evolve.elsevier.com/Grove/practice/
identify relationships or associations among vari-
ables, determine differences between groups or
compare groups on selected variables, and predict a
dependent variable on the basis of selected indepen-
dent variables.
A combination of the following formats might
be used in developing objectives to guide a study.
The focus of each objective is identified in the
parentheses at the end of each statement. The objec-
tives are placed in order from the least complex
to the most complex in generating research evi-
dence. Thus, the objectives or aims of studies
might be to:
1. Identify the elements or characteristics of variable
X in a selected population (identification).
2. Describe variable X in a selected population
(description).
3. Determine the difference between groups 1 and 2
or to compare groups 1 and 2 on variable X in a
selected population (difference).
4. Examine the relationship between variables X and
Y in a selected population (relational).
5. Determine whether certain independent variables
are predictive of a dependent variable in a selected
population (prediction).
Formulating Objectives or Aims
in Quantitative Studies
The objectives or aims in quantitative studies are
developed on the basis of the research problem and
purpose to clarify the study goals, variables, and popu-
lation. The following excerpts, from a descriptive
study of the symptom management strategies used by
elderly patients after coronary artery bypass surgery
(CABS), demonstrate the logical flow from research
problem (including the problem significance, back-
ground, and statement) and purpose to research aims
(Schulz, Zimmerman, Pozehl, Barnason, & Nieveen,
2011).
 CHAPTER 8  Objectives, Questions, Hypotheses, and Study Variables 139
Research Problem
“A major component of nursing care after coronary
artery bypass surgery (CABS) is focused on educating
patients to recognize and manage postoperative sym­
ptoms [problem significance]. The symptoms com­
monly experienced have been well documented in
the literature and include sleep disturbances (Tranmer
& Parry, 2004; Zimmerman, Barnason, Nieveen, &
Schmaderer, 2004), fatigue (Tranmer & Parry, 2004;
Zimmerman et al., 2004), swelling (Tranmer & Parry,
2004; Zimmerman et al., 2004), shortness of breath
(SOB), appetite problems, and chest and leg incision
pain (Zimmerman et al., 2004) [problem background].
Although much has been done to document postop­
erative symptoms, very little has been done to
describe the strategies that patients use to manage
these symptoms [problem statement].” (Schulz et al.,
2011, pp. 65-66)
Research Purpose
“The purpose of this study was to describe the
symptom management strategies used by older adults
after CABS.” (Schulz et al., p. 66)
Research Aims
“The specific aims of this study were to examine data
3 and 6 weeks after CABS to (a) identify categories
of symptom management strategies from patient’s
self-report data; (b) describe symptom management
strategies used for frequently reported symptoms
including SOB, fatigue, incision pain, sleep distur­
bance, swelling, and appetite problems; and (c) deter­
mine if patients used appropriate strategies to manage
symptoms by comparing reported symptom manage­
ment strategies to current evidence.” (Schulz et al.,
2011, p. 66)
In this example, the problem provides a basis for
the purpose, and the aims evolve from the purpose to
clearly focus the conduct of the study. The first aim
was focused on identification of the categories of
symptom management strategies (variable) used by
older adults (population) 3 and 6 weeks after CABS
(hospital and home settings). The study participants
were recruited from four Midwestern hospitals, but the
majority of data collection took place in the partici-
pants’ homes. The second aim was focused on descrip-
tion of the older adults’ symptom management
strategies (variable) used. The third objective was
focused on comparison of or differences in patients’
reported symptom management strategies and current
evidence-based guidelines developed by the American
Heart Association. Schulz et al. (2011, p. 65) found,
“Three weeks after surgery, the most frequently used
strategies were rest to manage shortness of breath
(53%) and fatigue (53%), medications for incision
pain (24%), and repositioning for swelling (35%) and
sleep disturbance (18%). Overall, fewer patients expe-
riencing sleep disturbances (39%), incision pain
(39%), swelling (46%), and appetite problems (17%)
reported using a strategy to manage their symptoms.”
Thus, the researchers stressed the importance of
nurses’ education of patients about symptom identifi-
cation and effective management strategies to improve
recovery following CABS.
Formulating Objectives or
Aims in Qualitative Studies
Many qualitative studies are guided by the study
purpose and do not include research objectives or
questions. However, some qualitative researchers do
develop objectives to guide selected studies. The
objectives in qualitative studies usually have a broader
focus and include more abstract and complex vari-
ables or concepts than those in quantitative studies
(Munhall, 2012). An ethnographic study by Happ,
Swigart, Tate, Hoffman, and Arnold (2007) included
objectives to direct their investigation of patients’
involvement in health-related decisions during pro-
longed critical illness, as shown by the following
excerpts.
Research Problem
“Clinicians increasingly recognize the need to involve
patients in decision making before, and, if possible,
during prolonged critical illness, but have little guid­
ance as to how and when to do this most effectively
[problem significance].… Prior reports of studies con­
taining the number of patients able to communicate
treatment preferences or to participate in decisions
during an acute or critical illness vary from none to
as high as 48% [problem background].… Moreover,
few reports of studies of treatment decision making
indicate the criteria used to make decisional capacity
assessments.… Consequently, empirical knowledge
of practice in this important area of patient care is
limited [problem statement].” (Happ et al., 2007, pp.
361-362)
Research Purpose
The purpose of this study was to “describe patterns
of communication of patients involved in health-
140 UNIT TWO  The Research Process
related decision making during prolonged mechani­
cal ventilation (PMV)” (Happ et al., 2007, p. 362).
Research Objectives
The objectives of this study were to “describe: (a)
characteristics of patients who were involved in
health-related decisions; (b) types of health-related
decisions made with patient involvement; (c) how
patient involvement occurred; and (d) the extent of
patient involvement with health-related decisions
during PMV” (Happ et al., 2007, p. 361).
In this ethnographic study, the problem statement
indicated that inadequate research had been conducted
on patient involvement in health-related decisions
during critical illness, which provided a basis for the
study purpose. All four objectives focused on detailed
descriptions of the study variables: (1) characteristics
of patients undergoing PMV, (2) health-related deci-
sion making of these patients, (3) how patient involve-
ment in decision making occurred, and (4) extent of
patient decision making. The findings from this study
indicated that families, advanced practice nurses, and
physicians were engaging critically ill patients in deci-
sion making whenever possible. However, most of the
time the patients could not make independent deci-
sions but were able to share decision making with their
families and clinicians. These findings emphasize how
important it is for families and clinicians to include
critically ill patients in health-related decisions at
whatever level possible (Happ et al., 2007).
Formulating Research Questions
A research question is a concise, interrogative state-
ment that is worded in the present tense and includes
one or more variables (or concepts). The research
questions focus on the following: (1) the identification
and/or description of the variable(s), (2) a determina-
tion of differences between two or more groups
regarding selected variables, (3) an examination of
relationships among variables (relational), and (4) the
use of independent variables to predict a dependent
variable.
You might use the following formats in developing
research questions for a study (the focus for each ques-
tion is shown in parentheses). The levels of evidence
to be generated by the following research questions
progress from simple (identification) to complex
(prediction).
1. What are the elements or characteristics of variable
X in a selected population (identification)?
2. How is variable X described by a selected popula-
tion (description)?
3. Is there a difference between groups 1 and 2 regard-
ing variable X (difference)?
4. What is the relationship between variables X and Y
in a selected population (relational)?
5. Are independent variables W, X, and Y predictive
of dependent variable Z (prediction)?
Formulating Questions in Quantitative Studies
Delaney, Apostolidis, Lachapelle, and Fortinsky (2011,
p. 285) conducted a comparative descriptive study to
examine “home care nurses’ knowledge of evidence-
based education topics for management of heart
failure.” The following excerpts from this study dem-
onstrate the flow from research problem and purpose
to research questions.
Research Problem
“Heart failure (HF), a chronic and disabling syndrome
affecting adults of all ages and particularly older adults,
is a major public health problem. An estimated 5.7
million Americans are currently affected by HF, and
this figure is expected to double over the next 25
years, primarily because of the aging of the population
and decreased mortality from other cardiovascular
conditions (Hodges, 2009).… HF is characterized by
poor posthospital discharge outcomes [problem sig-
nificance].… Home care agencies are currently being
challenged by Centers for Medicare and Medicaid
Services (CMS) to improve outcomes in HF manage­
ment.… Home care nurses play key roles in the deliv­
ery of education to patients and their families [problem
background]. However, home care nurses face unique
challenges compared with nurses at other sites
of care in providing comprehensive education on
managing HF. These challenges include a lack of
access to detailed patient information (Bowles, Pham,
O’Connor, & Horowitz, 2010), a focus on generalist
rather than specialist practice, and few opportunities
for continuing education in specialized knowledge
such as managing HF [problem statement].” (Delaney
et al., 2011, p. 286)
Research Purpose
The purpose of this study was to examine “home care
nurses’ knowledge of evidence-based HF education
related to the disease process, its management, and
patients’ self-management” (Delaney et al., 2011,
p. 286).
 CHAPTER 8  Objectives, Questions, Hypotheses, and Study Variables 141
Research Questions
Research questions included the following:
“(1) What is the level of knowledge of home care
nurses regarding evidence-based education
topics for patients with HF?…
(2) Were differences evident in nurses’ knowledge
based on education and work experience?…
(3) What are home care nurses’ self-reported
knowledge needs related to the care of patients
with HF?” (Delaney et al., 2011, p. 287)
Question 1 focused on description of the home care
nurses’ (population) knowledge about HF (variable).
Question 2 focused on determining differences in the
nurses’ knowledge on the basis of educational level
and years of work experience (demographic vari-
ables). Question 3 focused on description of the
nurses’ self-reported knowledge needs (variable) in
managing patients with HF. Delaney et al. (2011)
found that home care nurses were limited in their
evidence-based knowledge for managing HF. There
were no significance differences in the nurses’ knowl-
edge and their educational level and years of experi-
ence. The researchers concluded that the home care
nurses needed educational programs focused on HF
patient management to improve the quality of patient
education they could provide.
Formulating Questions in Qualitative Studies
The questions in qualitative studies are often limited in
number, have a broad focus, and include concepts that
are more complex and abstract than those in quantita-
tive studies. Marshall and Rossman (2011) noted that
the questions in qualitative research either might be
theoretical ones, which can be studied with different
populations or in a variety of sites, or could be focused
on a particular population or setting. Hudson et al.
(2010) conducted an exploratory-descriptive qualita-
tive study to examine the health-seeking challenges
perceived by homeless young adults. The problem,
purpose, and research questions used to direct this
study are presented in the following excerpts.
Research Problem
“Adolescent homelessness is a distressing social
problem. Approximately 1.5 to 2 million homeless
adolescent persons live on the streets in the United
States (Bucher, 2008); homelessness among young
persons is more common than homelessness among
older adults.… Homeless young adults are highly vul­
nerable to negative health consequences because of
the realities of street life [problem significance].…
Homeless young persons are at risk for sexual and
physical abuse.… Other negative health conse­
quences experienced by homeless young adults
include sexually transmitted infections, poorly con­
trolled chronic mental illness, and lack of immuniza­
tion for conditions, such as hepatitis A and hepatitis
B (Hudson, Nyamathi, & Sweat, 2008).… Homeless
persons are more likely to be admitted to the hospital
and have increased durations of hospitalization than
those of nonhomeless persons due to negative health
consequences associated with street living.… Nearly
half of all homeless young persons have no regular
source of health care (Sneller et al., 2008) [problem
background].… The perceptions of how homeless
persons view the health care system have not been
well studied.… There is a growing assertion that
improvements should be made with respect to the
provision of quality care for the homeless young
adults living in the United States.… One way to
achieve high-quality programs designed to improve
health care for homeless adults is to solicit these
adults’ input in program development [problem state-
ment].” (Hudson et al., 2010, pp. 212-213)
Research Purpose
“The purpose of this article was to gain a further
understanding of the perceptions of homeless youth
regarding their healthcare-seeking behaviors”
(Hudson et al., 2010, p. 213).
Research Questions
This study included the following research questions:
“1.  What are homeless young adults’ perspectives on
facilitators and barriers to receiving health care?
2.  How can existing homeless youth and young-
adult-centered healthcare programs be
improved?” (Hudson et al., 2010, 213)
The first study question focused on developing a
description of the homeless young adults’ (popula-
tion) perspectives on facilitators and barriers to receiv-
ing health care (research concepts). The second
question focused on identifying and describing how
young-adult-centered healthcare programs can be
improved (research concept). The study’s “identified
themes were failing access to care based on perceived
structural barriers (limited clinic sites, limited hours
of operation, priority health conditions, and long wait
142 UNIT TWO  The Research Process
times) and social barriers (perception of discrimina-
tion by uncaring professionals, law enforcement, and
society in general…)” (Hudson et al., 2010, p. 212).
The researchers also gained insights into the program-
matic and agency resources that are needed to promote
health-seeking behaviors by homeless young adults.
Formulating Hypotheses
A hypothesis is the formal statement of the expected
relationship or relationships between two or more
variables in a selected population. The hypothesis
translates the problem and purpose into a clear expla-
nation or prediction of the expected results or out-
comes of the study (Shadish, Cook, & Campbell,
2002). This section describes the purpose, sources,
and types of hypotheses that are commonly developed
by researchers. In addition, the process for developing
and testing hypotheses in nursing studies is described.
Purpose of Hypotheses
The purpose of a hypothesis is similar to that of
research objectives and questions. A hypothesis (1)
specifies the variables you will manipulate or measure,
(2) identifies the population you will examine, (3)
indicates the type of research, and (4) directs the
conduct of your study. Hypotheses direct the conduct
of a study by influencing the study design, sampling
technique, data collection and analysis methods, and
interpretation of findings. Hypotheses differ from
objectives and questions by predicting the outcomes
of a study. Study hypotheses are used to organize the
results section of a study, and the results indicate
support or nonsupport of each hypothesis. Hypothesis
testing allows us to generate knowledge by testing
theoretical statements or relationships that were iden-
tified in previous research, proposed by theorists, or
observed in practice (Chinn & Kramer, 2008; Fawcett
& Garity, 2009).
Sources of Hypotheses
Research hypotheses can be generated by observing
phenomena or problems in nursing practice, analyzing
theory, and reviewing the research literature. Many
hypotheses originate from real-life experiences. Clini-
cians and researchers observe events in practice and
identify relationships among these events (theorizing),
which are the bases for formulating hypotheses. For
example, you may notice that the hospitalized patient
who complains the most about pain receives the most
pain medicine and other pain management strategies.
The relationship identified is a prediction about events
in clinical practice that has potential for empirical
testing, because certain patients might not be receiving
adequate pain management.
You could conduct a literature review to identify a
theory that supports this relationship. For example,
Fagerhaugh and Strauss (1977) developed a theory of
pain management and identified the following rela-
tionship or proposition: As expressions of pain
increase, pain management increases. The researchers
developed this proposition through the use of grounded
theory research. Additional testing is necessary to
determine its usefulness in describing how patients
express pain and how that pain is managed in a variety
of practice situations. On the basis of theory and clini-
cal observation, the following hypothesis might be
formulated: The more frequently a hospitalized patient
verbalizes perceptions of pain, the greater the admin-
istration of analgesic medications by healthcare
providers.
Some hypotheses are initially generated from rela-
tionships expressed in a theory, when the intent of the
researcher is to test a theory. Usually, middle-range
theories are tested in research, and a proposition or
relationship from the theory provides the basis for the
generation of one or more study hypotheses (Fawcett
& Garity, 2009; Smith & Liehr, 2008). For example,
Rungruangsiripan, Sitthimongkol, Maneesriwongul,
Talley, and Vorapongsathorn (2011) tested the rela-
tionships in the Common-Sense Model of Illness Rep-
resentation (Diefenbach & Leventhal, 1996) to
examine the factors affecting medication adherence in
individuals with schizophrenia. Figure 8-1 contains
the framework model for this study based on the
Common-Sense Model of Illness Representation,
which has three stages: sources of information, illness
representation, and coping. “Sources of information
included social support variable, therapeutic alliance
variable, and experience of medication side effects
variable. Coping consisted of intention to change
adherence behavior and adherence behavior” (Run-
guangsiripan et al., 2011, p. 272). This model shows
the direct and indirect relationships among the con-
cepts that provide a basis for the study hypotheses. A
direct relationship is when one concept links to another
concept without an intervening concept. For example,
the concept of social support is linked directly to
illness representation. In an indirect relationship, one
concept is linked to another concept through an inter-
vening third concept. For example, the concept experi-
ence with medication side effects is indirectly linked
to the concept intention to change adherence behavior
through the concept of illness representation (see
Figure 8-1). The Rungruangsiripan et al. (2011) study
set the following hypotheses:
 CHAPTER 8  Objectives, Questions, Hypotheses, and Study Variables 143

Social Support

+ Illness
Therapeutic Representation
Alliance

+
+
+
– Intention to Change
Adherence Behavior

Experience of Medication
– –
Side Effects

Adherence Behavior

Figure 8-1  Hypothesized medication adherence model. (From Rungruangsiripan, M., Sitthimongkol, Y., Maneesriwongul, W., Talley, S., &
Vorapongsathorn, T. [2011]. Mediating role of illness representation among social support, therapeutic alliance, experience of medication side effects,
and medication adherence in persons with schizophrenia. Archives of Psychiatric Nursing, 25[4], 273.)

These hypotheses were formulated to test the prop-

“(a)  Social support and therapeutic alliance would
have a positive direct effect on illness
representation and adherence behavior in
individuals with schizophrenia;
(b)  Experience of medication side effects would
have a negative direct effect on illness
representation and adherence behavior in
individuals with schizophrenia;
(c)  Intention to change adherence behavior would
have a negative direct effect on adherence
behavior;
(d)  Social support, therapeutic alliance, and
experience of medication side effects would
have an indirect effect on intention to change
adherence behavior and adherence behavior
through illness representation; and
(e)  Illness representation would have a positive
direct effect on intention to change adherence
behavior and adherence behavior as well as an
indirect effect on adherence behavior via
intention to change adherence behavior.”
(Rungruangsiripan et al., 2011, p. 272)
ositions or relationships from the Common-Sense
Model of Illness Representation (see Figure 8-1). The
researchers found that “therapeutic alliance and the
experience of medication side-effects enhanced illness
representation, which in turn led to an intention to
change adherence behavior. Social support did not
alter illness representation or adherence behavior”
(Rungruangsiripan et al., 2011, p. 269). Illness repre-
sentation is the patients’ perception of their schizo-
phrenia and their ability to cope with the illness.
Patients with a clear perception of their schizophrenia
have strong intentions to change their adherence
behaviors. Thus, mental health nurses need to promote
the patients’ understanding of their schizophrenia
illness to enhance their adherence to their
medications.
Reviewing the research literature and synthesizing
findings from different studies can also be used to
generate hypotheses. For example, Ross, Sawatphanit,
Mizuno, and Takeo (2011) synthesized the findings
from studies to identify the factors that predict depres-
sive symptoms in postpartum women who are HIV-
positive. They developed a conceptual framework for
144 UNIT TWO  The Research Process
Self-esteem
– might have one, three, or more hypotheses, depending
Emotional
Support –
Depressive
Symptoms
+ tional versus nondirectional, and (4) null versus
Physical
Symptoms
–
Infant
Health Status
Figure 8-2  Conceptual framework of the study. (From Ross, R.,
Sawatphanit, W., Mizuno, M., & Takeo, K. [2011]. Depressive symptoms
among HIV-positive postpartum women in Thailand. Archives of
Psychiatric Nursing, 25[1], 37.)
their study that is presented in Figure 8-2. The
researchers “hypothesized that depressive symptoms
are negatively related to self-esteem, emotional
support, and infant health status but positively associ-
ated with physical symptoms” (Ross et al., 2011,
p. 37).
Ross et al. (2011) found that self-esteem and infant
health status were significant predictors of postpartum
women’s depressive symptoms but physical symp-
toms and emotional support were not. The study
results indicated that 74.1% of the HIV-positive post-
partum women had symptoms of depression, and the
researchers encouraged nurses to examine the self-
esteem and infant health status of such women to
increase identification of episodes of depression. The
researchers also recommended further research to
identify additional factors that might be predictive of
depression in HIV-positive postpartum women. Thus,
two relationships, those of self-esteem and infant
health status to depressive symptoms, were supported
in the framework model (see Figure 8-2). However,
the other relationships, those of emotional support and
physical symptoms to depressive symptoms, were not
supported in this study. Additional research is needed
to increase understanding of the factors that might be
predictive of depression in postpartum women who
are HIV-positive.
Types of Hypotheses
Hypotheses identify different types of relationships
and include different numbers of variables. Studies
on the complexity and scope of the study. The type of
hypothesis developed is based on the problem and
purpose of a study. The following four categories are
used to describe types of hypotheses: (1) associative
versus causal, (2) simple versus complex, (3) direc-
research.
Associative versus Causal Hypotheses
The relationships in hypotheses are identified as asso-
ciative or causal. An associative relationship identi-
fies variables that occur or exist together in practice,
and as one variable changes so does the other. For
example, research indicates there is an associative
relationship between anxiety and depression, and as a
person’s depression changes so does the anxiety level.
Thus, associative hypotheses are developed to
examine relationships among variables in a study. The
formats used for expressing associative hypotheses
follow:
1. Variable X is related to or associated with variable
Y in a selected population. (Predicts a relationship
between two variables but does not indicate the
type of relationship.)
2. An increase in variable X is related to an increase
in variable Y, or variable X is positively related to
variable Y in a selected population. (Predicts a posi-
tive relationship.)
3. A decrease in variable X is related to a decrease in
variable Y in a selected population. (Predicts a posi-
tive relationship.)
4. An increase in variable X is related to a decrease
in variable Y, or variable X is negatively related to
variable Y in a selected population. (Predicts a
negative or inverse relationship.)
5. Variables X and Y are predictive of variable Z in a
study. (The independent variables X and Y are used
to predict the dependent variable Z in a predictive
correlational study.)
Associative hypotheses identify relationships among
variables in a study but do not indicate that one variable
causes an effect on another variable. Researchers state
associative hypotheses when the focus of their study is
to examine relationships and not to determine cause
and effect. For example, Reishtein (2005) conducted
a predictive correlational study to examine the relation-
ships between symptoms and functional performance in
patients with chronic obstructed pulmonary disease
 CHAPTER 8  Objectives, Questions, Hypotheses, and Study Variables 145
(COPD). Reishtein developed the following associative
hypotheses to guide the study:
“1.  Positive relationships exist among dyspnea,
fatigue, and sleep difficulty in people with
COPD;
2.  Dyspnea, fatigue, and sleep difficulty are related
to functional performance; and
3.  Dyspnea, fatigue, and sleep difficulty, taken
together, will explain more of the variance in
functional performance in people with COPD
than any of these symptoms alone.” (Reishtein,
2005, p. 40)
Hypothesis 1 predicts positive relationships or
associations among the variables of dyspnea, fatigue,
and sleep difficulty for patients with COPD. A positive
relationship means that the variables change together;
thus, they will all increase together in value or all
decrease together. These relationships are depicted in
the following diagram:
+
Dyspnea Fatigue
+ +
Sleep Difficulty
Hypothesis 2 predicts relationships between three
variables—dyspnea, fatigue, and sleep difficulty—and
the variable functional performance, but it does not
identify the type of relationship. These relationships
are shown in the following diagram:
Dyspnea
Fatigue
Functional performance
Sleep difficulty
Hypothesis 3 uses the independent variables
dyspnea, fatigue, and sleep difficulty to predict the
dependent or outcome variable functional perfor-
mance in COPD patients. The predictive relationship
is shown in the following diagram:
Dyspnea + Fatigue + Sleep Difficulty →
Functional Performance
The results from Reishtein’s (2005) study partially
supported hypothesis 1, in that dyspnea had positive,
significant relationships with fatigue (r = 0.43, p <
0.001) and sleep difficulty (r = 0.39, p < 0.001), but
fatigue and sleep difficulty were positively but not
significantly related (r = 0.19). Hypothesis 2 was also
partially supported in that dyspnea (r = −0.54, p <
0.001) and fatigue (r = −0.24, p < 0.01) were signifi-
cantly, negatively related to functional performance,
but sleep difficulty (r = −0.17) was not. Thus, in
hypothesis 3, dyspnea was the most predictive of func-
tional performance, with fatigue and sleep difficulty
providing limited prediction. Thus, managing dyspnea
may be the best way to improve the symptoms and
functional performance in patients with COPD. Addi-
tional research may distinguish other symptoms that
might predict functional performance in COPD
patients and thereby help them and healthcare provid-
ers in managing this chronic disease.
Causal relationships identify a cause-and-effect
interaction between two or more variables, which are
referred to as independent and dependent variables.
The independent variable (intervention, treatment,
or experimental variable) is manipulated or varied
by the researcher to have an effect on the dependent
variable. The dependent variable (outcome or
response variable) is measured to examine the effect
created by the independent variable. A format for
stating a causal hypothesis is as follows: Subjects
experiencing the independent variable X demonstrate
greater change in dependent variable Y than do the
subjects in the control or comparison group. For
example, cancer patients receiving a relaxation music
intervention have less perceived pain than those
receiving usual care.
Scott, Hofmeister, Rogness, and Rogers’s (2010)
review of the literature indicated that nurses working
12-hour shifts had difficulties staying awake on duty,
reduced time for sleep, and significantly higher risk
for errors. Thus, these researchers conducted a quasi-
experimental study to determine the effect of a fatigue
countermeasures program for nurses (FCMPN) on
nurses’ alertness and number of near and actual patient
errors. A causal hypothesis was tested in the conduct
of this study. The researchers hypothesized that imple-
menting the FCMPN had the potential for improving
sleep quality and sleep duration and decreasing
daytime sleepiness, drowsiness episodes, risk for auto
accidents, and actual and near reported work errors
(Scott et al., 2010).
The independent variable or study intervention was
the FCMPN that was implemented to determine its
146 UNIT TWO  The Research Process
impact on the dependent or outcome variables sleep
quality, sleep duration, daytime sleepiness, drowsiness
episodes, auto accident risk, and work errors. The
study population was hospital staff nurses, and the
settings were medical-surgical units in three major
acute care Michigan hospitals. Scott et al. (2010)
found that the FCMPN significantly improved the hos-
pital nurses’ sleep duration, sleep quality, alertness
(fewer episodes of drowsiness), and accident and error
risk. However, there was not significant improvement
in daytime sleepiness. Thus, the hypothesis was sup-
ported for five of the six dependent variables and not
Independent Variable
+
+ Sleep duration (supported)
Fatigue Countermeasures
Program for Nurse –
(FCMPN)
–
–
–
Actual and near reported work errors (supported)
Simple versus Complex Hypotheses
A simple hypothesis predicts the relationship (asso-
ciative or causal) between two variables. One format
for stating a simple associative hypothesis is as
follows: Variable X is related to variable Y. A simple
causal hypothesis identifies the relationship between
one independent variable and one dependent variable,
for example, independent variable X causes a change
in dependent variable Y. Vasan et al. (2003) studied
the relationship of elevated plasma homocysteine
levels with the risk for congestive heart failure (CHF)
in adults without prior myocardial infarction (MI). A
simple, associative hypothesis was developed to direct
this study: “We hypothesized that elevated plasma
homocysteine levels are associated with an increased
risk for CHF” (Vasan et al., 2003, p. 1251). The fol-
lowing diagram demonstrates the positive relationship
that was identified between the two study variables of
plasma homocysteine level and risk for CHF:
↑ Plasma homocysteine level
supported for the outcome daytime sleepiness. The
interpretation of the results might have been facilitated
by the statement of additional hypotheses with fewer
dependent variables in each hypothesis. A diagram of
this hypothesis follows, with causal arrows (→) indi-
cating the cause-and-effect relationship between the
independent variable (IV) and dependent variables
(DVs). Causal arrows indicate the effect of IV on the
DV (see Chapter 7 for discussion of types of relation-
ships). After each dependent variable is indicated
whether this part of the hypothesis was supported or
not supported.
Dependent or Outcome Variables
Sleep quality (supported)
Daytime sleepiness (not supported)
Drowsiness episodes (supported)
Risk of auto accidents (supported)
The results of this 8-year study indicated that an ele-
vated plasma homocysteine concentration was posi-
tively and strongly related to risk of CHF in both men
and women who did not have a prior history of MI.
The hypothesis was supported in this study, indicating
that nurse practitioners and clinical nurse specialists
need to examine plasma homocysteine levels in indi-
viduals with a family history of CHF and treat those
levels as needed.
A complex hypothesis predicts the relationship
(associative or causal) among three or more variables.
A complex associative hypothesis predicts the rela-
tionships among three or more variables, such as the
relationships among the variables X, Y, and Z. Complex
causal hypotheses also include three or more variables
but predict the effects of one independent variable on
two (or more) dependent variables or predict the
effects of two or more independent variables on one
or more dependent variables. For example, Scott et al.
+
↑ Risk for CHF
 CHAPTER 8  Objectives, Questions, Hypotheses, and Study Variables 147
(2010) tested a complex causal hypothesis in their
examination of the effects of the nurse fatigue inter-
vention program (FCMPN) on the sleep duration,
sleep quality, daytime sleepiness, drowsiness episodes,
risk of auto accident, and actual and near reported
work errors. The figure presented earlier demonstrat-
ing this causal hypothesis includes one independent
variable, FCMPN, and six dependent or outcome vari-
ables. This intervention offers potential benefits for
managing nurses’ fatigue but requires additional
investigation to determine its effectiveness in manag-
ing daytime sleepiness. Often, in practice situations,
multiple variables cause an event, or an intervention
results in multiple outcomes. Therefore, complex
rather than simple associative or causal hypotheses are
often more representative of nursing practice.
Nondirectional versus Directional Hypotheses
A nondirectional hypothesis states that a relationship
exists but does not predict the nature of the relation-
ship. If the direction of the relationship being studied
is not clear in clinical practice or the theoretical or
empirical literature, the researcher has no clear indica-
tion of the nature of the relationship and states a non-
directional hypothesis (Fawcett & Garity, 2009). For
example, Reishtein’s (2005, p. 4) second hypothesis
(introduced earlier in this chapter) is nondirectional:
“Dyspnea, fatigue, and sleep difficulty are related to
functional performance.” This hypothesis indicates
that dyspnea, fatigue, and sleep difficulty are related
to functional performance, but it does not indicate the
direction or nature (positive or negative) of the rela-
tionship. This hypothesis is nondirectional, complex
(four variables), and associative (indicating that a rela-
tionship exists).
A directional hypothesis states the nature or direc-
tion of the relationship between two or more variables.
These hypotheses are developed from theoretical
statements, findings of previous studies, and clinical
experience. As the knowledge on which a study is
based increases, the researcher is able to predict the
direction of a relationship between the variables being
studied. Terms such as less, more, increase, decrease,
positive, negative, greater, and smaller indicate the
directions of relationships in hypotheses. Directional
hypotheses can be associative or causal and simple or
complex. In a study introduced earlier, Ross et al.
(2011) “hypothesized that depressive symptoms are
negatively related to self-esteem, emotional support,
and infant health status but positively associated
with physical symptoms” based on their framework
model [see Figure 8-2] (Ross et al., 2011, p. 37). This
hypothesis is associative (examines relationships or
associations among variables), complex (includes five
variables), and directional (identifies positive and
negative associations among variables).
A causal hypothesis predicts the effect of an inde-
pendent variable on a dependent variable, specifying
the direction of the relationship. Thus, all causal
hypotheses are directional. Efe and Özer (2007) exam-
ined the pain-relieving effect of breast-feeding during
immunization injections in healthy neonates and used
a causal hypothesis to direct their study. “The hypoth-
esis tested was that breast-feeding would decrease the
length of crying time, prevent an increase in heart rate,
and prevent a decrease in oxygen saturation during
vaccination as compared with the control condition
(i.e., no breast-feeding)” (Efe & Özer, 2007, p. 11).
This causal hypothesis predicted the effect of an inter-
vention or independent variable, breast-feeding,
during immunization injections on the dependent or
outcome variables length of crying time, heart rate,
and oxygen saturation. Thus, this is a complex (four
variables), directional (decrease crying time and
prevent increased heart rate and decreased oxygen
saturation), causal hypothesis. The breast-feeding sig-
nificantly decreased crying time but did not signifi-
cantly affect the neonates’ heart rate or oxygen
saturation values. Because breast-feeding did decrease
neonate crying time during immunizations, nurses
might encourage mothers to implement this safe, easy,
effective intervention in practice.
Null versus Research Hypotheses
The null hypothesis (H0), also referred to as a statisti-
cal hypothesis, is used for statistical testing and inter-
pretation of these results. Even if the null hypothesis
is not stated, it is implied, because it is the converse
of the research hypothesis (Kerlinger & Lee, 2000;
Shadish et al., 2002). A null hypothesis can be simple
or complex and associative or causal. An associative
null hypothesis states that there is no relationship
between the variables studied. A causal null hypoth-
esis might be stated in one of the following formats:
1. The independent variable has no effect on the
dependent variable.
2. The experimental group, who received the inde-
pendent variable, is no different from the control
or comparison group for the dependent variable.
Youngkin and Lester (2010, p. 5) conducted a study
to “(a) determine if nonpregnant, childbearing-age
women could accurately self-screen to predict BV
[bacterial vaginosis] using a researcher-developed
home self-test system comprised of three integrated
components—education, application of self-test
method, and scoring of self-test findings using a
148 UNIT TWO  The Research Process
unique scoring system—and (b) determine if the
women would seek early professional diagnosis and
treatment.” These researchers developed two hypoth-
eses to direct their study:
“The researchers expected to find a significant posi­
tive correlation between the self-screening scores
from the women and the follow-up evaluation scores
from the nurse researchers using the self-test system.
It was further expected that there would be no sig­
nificant differences between the women’s scores on
the criteria using the scoring method and the related
researchers’ scores.” (Youngkin & Lester, 2010, p. 5)
The first hypothesis is associative (identifies a rela-
tionship between women’s scores and researchers’
scores on the self-test system), simple (includes two
variables), and directional (examines positive correla-
tion or relationship). The second hypothesis is a null
hypothesis that states there is no difference between
the women’s and the researchers’ scores on the basis
of the scoring method. Null hypothesis is nondirec-
tional since it states that no relationship exists between
variables or that no difference exists between groups.
Research hypotheses are supported or not supported
and null hypotheses are accepted or rejected on the
basis of statistical results. Youngkin and Lester (2010)
stated a simple null hypothesis about no difference
between the women’s and researchers’ scores on the
basis of previous research and their clinical expertise.
The researchers found that the women accurately
tested themselves for BV using the self-test system
and appropriately sought definitive professional diag-
nosis and treatment. Thus, the research hypothesis was
supported and the null hypothesis was accepted on the
basis of study results.
A research hypothesis is the alternative hypothesis
(H1 or Ha) to the null. The research hypothesis states
that there is a relationship between two or more vari-
ables, and it can be simple or complex, nondirectional
or directional, and associative or causal. The predic-
tion in a research hypothesis might be based on theo-
retical statements, previous research findings, and/or
clinical experience. All the previous examples of
hypotheses presented in this chapter are research
hypotheses except for the one null hypothesis formu-
lated by Youngkin and Lester (2010).
Researchers have different beliefs about when to
state a research hypothesis versus a null hypothesis.
Some researchers state the null hypothesis because it
is more easily interpreted on the basis of the results of
statistical analyses. A researcher will also use the null
hypothesis when she or he believes there is no rela-
tionship between two or more variables and when
there is inadequate theoretical or empirical informa-
tion to state a research hypothesis. Otherwise it is best
to state a research hypothesis that clearly predicts the
outcome of a study. As previously discussed, Young-
kin and Lester (2010) stated both a research hypoth-
esis and a null hypothesis based on previous research
and clinical expertise. They expected both hypotheses
to be supported by their study results, and they were.
Developing Hypotheses
Developing hypotheses requires inductive and deduc-
tive thinking. Most people have a predominant way of
thinking and will use that thinking pattern in develop-
ing hypotheses. Inductive thinkers have a tendency to
focus on the relationships they observe in clinical prac-
tice, and they synthesize these observations to formu-
late a general statement about the relationships. For
example, inductive thinkers might note that elderly
patients who are not instructed in the importance of
early postoperative ambulation are slow to get out of
bed. Deductive thinkers examine more abstract state-
ments from theories or previous research and then
formulate a hypothesis for study (Smith & Liehr, 2008).
Deductive thinkers might translate a statement or prop-
osition, such as “People who receive education about
self-care are more capable in caring for themselves,”
from Orem’s (2001) theory into a hypothesis.
The inductive thinker must link the relational state-
ment or hypothesis that was developed from clinical
observations with a theoretical framework. Making
this connection with the framework requires deductive
thinking and improves the usefulness of the study
findings. The deductive thinker must use inductive
thinking to determine whether the proposition from a
theory accurately predicts the relationship of events in
clinical practice. Without this real-world experience,
the selection of subjects and the identification of ways
to measure the variables would be unclear. An example
hypothesis is, “Elderly patients receiving an activity
educational program before surgery ambulate earlier
and have a shorter hospital stay after surgery than
elderly patients receiving standard care.”
In formulating a hypothesis, you as a researcher
will have several decisions to make. These decisions
are directed by the problem studied and by your own
expertise and preference. You must decide whether the
problem is best investigated with the use of simple or
complex hypotheses. Complex hypotheses frequently
require complex methodology, and the outcomes may
be difficult to interpret. Some beginning researchers
prefer the clarity of simple hypotheses.
 CHAPTER 8  Objectives, Questions, Hypotheses, and Study Variables 149
The research problem and purpose determine
whether you will study an associative or a causal rela-
tionship. Testing a hypothesis that states a causal rela-
tionship requires expertise in implementing a treatment
and controlling extraneous variables. Another decision
you must make involves the formulation of a research
or a null hypothesis. You must make this decision
according to what you believe is the most accurate
prediction of the relationship between the study
variables.
A hypothesis that is clearly and concisely stated
gives the greatest direction for conducting a study. For
clarity, hypotheses are expressed as declarative state-
ments written in the present tense. Thus, hypotheses
are best developed without the phrase “There will/will
not be a relationship…,” because the future tense
refers to the sample being studied. Hypotheses are
statements of relationships about populations, not
about study samples. According to mathematical
theory regarding generalization, one cannot generalize
to the future (Kerlinger & Lee, 2000).
Hypotheses are clearer without the phrase “There
is no significant difference…,” because the level of
significance is only a statistical technique applied to
sample data. In addition, hypotheses should not iden-
tify methodological points, such as techniques of
sampling, measurement, and data analysis (Kerlinger
& Lee, 2000). Therefore, a statement such as “mea-
sured by,” “in a random sample of,” or “using
ANOVA (analysis of variance)” are not appropriate.
Such a phrase limits the hypothesis to measurement
methods, sampling methods, or data analysis tech-
niques in a single study. A well-formulated hypothe-
sis clearly identifies the relationship between the
variables and the study population. There is no set
number for how many hypotheses are needed to
direct a study, but the number formulated usually
reflects the researcher’s expertise and the complexity
of the problem and purpose being studied. However,
most studies contain one to three hypotheses, and the
relationships identified in these hypotheses set the
limits for a study (Fawcett & Garity, 2009; Shadish
et al., 2002).
Testing Hypotheses
A hypothesis’s value is ultimately derived from
whether or not it can be tested in the real world. A
testable hypothesis contains variables that can be
measured or manipulated in practice. For example,
Efe and Özer (2007) manipulated the breast-feeding
intervention in their study using set protocol so that
the treatment was consistently manipulated for each
study situation. They measured crying time in seconds
and measured the heart rate and oxygen saturation
with a pulse oximeter (Nellcor N180).
Hypotheses are evaluated with statistical analyses.
If the hypothesis states an associative relationship,
correlational analyses are usually conducted on the
data. Spearman’s rank order correlation coefficient is
often used to analyze ordinal level data, and Pearson’s
product-moment correlation coefficient is used for
interval and ratio level data (see Chapter 23). These
correlational analyses determine the existence, type,
and degree of the relationship between the variables
studied.
A hypothesis that states a causal relationship is
analyzed through the use of statistics that examine
differences, such as the Mann-Whitney U test, the t
test, and analysis of variance (ANOVA) (see Chapter
25). It is the null hypothesis (stated or implied) that is
tested through statistical analysis. The intent is to
determine whether the independent variable had a sig-
nificant effect on the dependent variable. The level of
significance, alpha (α) = 0.05, 0.01, 0.001, is set after
the generation of causal hypotheses and before the
conduct of the study. To learn more about selecting
statistical tests and a level of significance for testing
hypotheses, see Chapter 21.
The results obtained from testing a hypothesis are
described with the use of certain terminology. Research
findings do not prove hypotheses true or false; instead,
hypotheses are statements of relationships or differ-
ences in populations. Even after a series of studies, the
word proven is not used in scientific language because
of the tentative nature of science. Research hypotheses
are described as being supported or not supported in
a study. When a null hypothesis is tested, it is either
rejected or accepted. Accepting the null hypothesis
indicates that no relationship or effect was found
among the variables. Rejecting the null hypothesis
indicates the possibility that a relationship or differ-
ence exists. A study might partially support a complex
hypothesis. Efe and Özer’s (2007, p. 11) hypothesis
stated that “breast-feeding decreases the length of
crying time, prevents an increase in heart rate, and
prevents a decrease in oxygen saturation during vac-
cination as compared with the control condition (i.e.,
no breast-feeding).” Their study supported the
decreased crying time part of the hypothesis, but it did
not support the part of the hypothesis that focused on
the prevention of increased heart rate and decreased
oxygen saturation. However, the study did provide
valuable evidence about the effectiveness of breast-
feeding in reducing the pain of immunization injec-
tions in infants. In addition, this study provides
direction for future research.
150 UNIT TWO  The Research Process
Selecting Objectives, Questions,
or Hypotheses for Quantitative
or Qualitative Research
Selecting objectives, questions, or hypotheses for a
study is often based on (1) the number and quality of
relevant studies conducted on a selected problem
(existing knowledge base), (2) the framework of the
study, (3) the expertise and preference of the researcher,
and (4) the type of study to be conducted (quantitative
or qualitative). Commonly, if minimal or no research
has been conducted on a problem, investigators state
objectives or questions because they do not have the
knowledge necessary to formulate hypotheses. The
framework for a study indicates whether the intent is
to develop or to test theory. Objectives and questions
are usually stated to guide theory development, and
the focus of a hypothesis is to test theory.
Researcher expertise and preference can also influ-
ence the selection of objectives, questions, or hypoth-
eses to direct a study. The number of nursing studies
containing hypotheses continues to grow, and there
appears to be a trend away from descriptive quantita-
tive studies toward studies focused on examining rela-
tionships between variables and testing hypotheses.
The greater use of hypotheses to direct quantitative,
outcomes, and intervention research indicates growth
of knowledge in selected problem areas and the
increasing sophistication of nurse researchers (Doran,
2011; Forbes, 2009). However, it is important that
researchers state hypotheses to direct their studies
explicitly versus implicitly or through implication. An
explicit statement of hypotheses is important to
provide clear direction for the conduct of a study, the
interpretation of the findings, and the use of the find-
ings in practice (Brown, 2009; Craig & Smyth, 2012;
Fawcett & Garity, 2009).
The objectives, questions, or hypotheses desig-
nated for study frequently indicate a pattern that the
researcher uses in conducting investigations. Prob-
lems can be investigated in a variety of ways. Some
researchers start at the core of a problem and work
their way outward. Other investigators study a problem
from the outside edge and work to the core (Kaplan,
1964). Each study must logically build on the other,
as the researcher establishes a pattern for studying a
problem area that will affect the quality and quantity
of the knowledge generated in that area.
Researchers select objectives, questions, or hypoth-
eses according to the type of study they plan to
conduct. Objectives and questions are typically stated
when the intent of the study is to identify or describe
TABLE 8-1  Selecting Objectives, Questions,
or Hypotheses for Different Types  
of Research
Objectives, Questions, or
Hypotheses Commonly
Type of Research Developed
Qualitative research Objectives, questions, or none
Quantitative research:
  Descriptive studies Objectives, questions, or none
  Correlational studies Objectives, questions,
hypotheses, or none
  Quasi-experimental studies Usually hypotheses
  Experimental studies Hypotheses
Outcomes research Hypotheses or questions
Intervention research Hypotheses
characteristics of variables, to examine relationships
among variables, or both. Thus, objectives or ques-
tions are formulated to direct qualitative and selected
quantitative (descriptive and correlational) studies
(see Table 8-1). However, some experienced research-
ers can clearly focus and develop a study without
using objectives or questions. In these studies, the
research purpose directs the research process.
In some qualitative research, the investigator uses
only a problem and purpose to direct the study. The
specification of objectives or questions might limit the
scope of the study and the methods of data collection
and analysis (Munhall, 2012). Discovery is important
in qualitative research, and hypotheses are neither nec-
essary nor desirable in qualitative studies.
Researchers often develop hypotheses when the
relationships or results of a study can be anticipated
or predicted. Hypotheses are typically used in quanti-
tative research to direct predictive correlational, quasi-
experimental, and experimental studies and are also
important to guide outcomes and intervention studies
(Doran, 2011; Forbes, 2009; Shadish et al., 2002).
Identifying and Defining
Study Variables
The research purpose and objectives, questions, and
hypotheses identify the variables or concepts to
be examined in a study. Variables are qualities, prop-
erties, or characteristics of persons, things, or situa-
tions that change or vary in a study. Variables are
characterized by degrees, amounts, and differences
within a study. Variables are also concepts of various
levels of abstraction that are concisely defined so that
 CHAPTER 8  Objectives, Questions, Hypotheses, and Study Variables 151
they can be measured or manipulated within a study
(Waltz, Strickland, & Lenz, 2010).
The concepts examined in research can be concrete
and directly measurable in practice, such as heart rate,
hemoglobin value, and tidal volume of the lung. These
concrete concepts are usually referred to as variables
in a study. Other concepts, such as anxiety, coping,
and pain, are more abstract and are indirectly observ-
able in the real world. Thus, the properties of these
concepts are inferred from a combination of measure-
ments. For example, one can infer the properties of
anxiety by combining information obtained from (1)
observing the signs and symptoms of anxiety (fre-
quent movements, sweating, rapid eye movement,
lack of eye contact, and verbalization of anxiety), (2)
examining completed questionnaires or scales (state
and trait anxiety scales), and (3) measuring physiolog-
ical responses (galvanic skin response). The concept
of anxiety might be represented by the variable
“reported anxiety” or “perceived level of anxiety”
(Waltz et al., 2010).
In many qualitative studies and in some quantita-
tive studies (descriptive and correlational), the focus
is abstract concepts, such as grieving, caring, and pro-
moting health (Creswell, 2009; Munhall, 2012).
Researchers identify the elements of the study as con-
cepts, not variables. In the ethnographic study previ-
ously described, Happ et al. (2007) investigated the
concept of health-related decision making by critically
ill patients during prolonged mechanical ventilation.
The concept health-related decision making was
defined as “choices about initiating, continuing, or
discontinuing treatment, diagnostics, or therapeutic
care activities” (p. 363). Qualitative studies are often
conducted to clarify the definitions of concepts so
these definitions are usually presented in the results
section of the study. In the results of this study, health-
related decision making was described as including
the following: “choices about mechanical ventilation
and other therapies, such as invasive diagnostic pro-
cedures and placement of central lines and nutritional
access devices that may or may not require written
informed consent, and about discharge placement.
Financial or legal decisions, such as appointment of a
power of attorney or signing financial documents to
enable insurance payment for health care, were con-
sidered health-related in the context of prolonged criti-
cal illness” (Happ et al., 2007, p. 363).
Types of Variables
Variables have been classified into a variety of types
to explain their use in research. Some variables are
manipulated; others are controlled. Some variables are
identified but not measured; others are measured with
refined measurement devices. The different types of
variables presented in this section include the follow-
ing: independent, dependent, research, extraneous,
demographic, moderator, and mediator.
Independent and Dependent Variables
The relationship between independent and dependent
variables is the basis for formulating hypotheses
for predictive correlational, quasi-experimental, and
experimental studies. As introduced earlier in this
chapter, an independent variable is an intervention
or treatment manipulated by the researcher to create
an effect on the dependent variable. A dependent
variable is the outcome the researcher wants to predict
or explain. Changes measured in the dependent vari-
able are presumed to be caused by the independent
variable (Kerlinger & Lee, 2000; Shadish et al., 2002).
Recall the quasi-experimental study by Scott et al.
(2011) introduced earlier in this chapter that included
independent and dependent variables. These variables
were identified in the study’s complex, causal, direc-
tional research hypothesis that focused on the effect
of the FCMPN on sleep duration, sleep quality, day­
time sleepiness, drowsiness episodes, risk of auto
accident, and actual and near reported work errors.
The independent variable that was manipulated in this
study was the FCMPN, and the effect of this program
was determined by measuring the dependent variables
sleep duration, sleep quality, daytime sleepiness,
drowsiness episodes, risk of auto accident, and actual
and near errors (see the previous figure showing these
variables).
Research Variables or Concepts
Qualitative studies and some quantitative (descriptive
and correlational) studies involve the investigation of
research variables or concepts. Research variables or
concepts are the qualities, properties, or characteris-
tics identified in the research purpose and objectives
or questions that are measured in a study. They are
used when the intent of the study is to measure vari-
ables as they exist in a natural setting without the
implementation of a treatment. Thus, no independent
variables are manipulated, and no cause-and-effect
relationships are examined.
Qualitative studies often focus on abstract con-
cepts. For example, Hudson et al. (2010) conducted
an exploratory-descriptive qualitative study to describe
the concept of “health-seeking challenges” among
homeless youth. This study, also introduced earlier
in the chapter, focused on gaining an understanding
of the perceptions of homeless youth regarding
152 UNIT TWO  The Research Process
their healthcare-seeking behaviors. The study results
described the concept of “health-seeking challenges”
with a discussion of the young adults’ perspectives on
the facilitators and barriers to receiving health care.
Extraneous Variables
Extraneous variables exist in all studies and can
affect the selection of study participants, implementa-
tion of the study intervention, measurement of study
variables, and implementation of study procedures.
Extraneous variables are of primary concern in quan-
titative and intervention studies, because they can
obscure one’s understanding of the relational or causal
dynamics within the studies. Extraneous variables are
classified as (1) recognized or unrecognized and (2)
controlled or uncontrolled.
The extraneous variables that are not recognized
until the study is in process or that are recognized
before the study is initiated but cannot be controlled
are referred to as confounding variables. Sometimes
these variables can be measured during the study and
controlled statistically during analysis. In other cases,
it is not possible to measure a confounding variable,
and the variable thus hinders the interpretation of
findings. Such extraneous variables must be identified
as limitations or areas of study weakness in the
discussion section of a research report. As control
decreases in quasi-experimental and experimental
studies, the potential influence of confounding vari-
ables increases.
Researchers attempt to recognize and control as
many extraneous variables as possible in quasi-
experimental and experimental studies, and specific
designs have been developed to control the influence
of such variables (see Chapter 11). Youngkin and
Lester (2010) conducted a quasi-experimental study
to determine the effectiveness of a home self-test
system in screening for bacterial vaginosis (BV). The
research and null hypotheses developed to direct this
study were introduced earlier, in the discussion of the
null hypothesis. These researchers controlled some of
the extraneous variables in their study by selecting
inclusion and exclusion sample criteria to ensure that
study participants were similar or homogeneous. The
home self-test system for BV was implemented using
a detailed protocol to ensure that all participants
received consistent, complete implementation of the
intervention (Fawcett & Garity, 2009). The interven-
tion was implemented on the basis of the Centers for
Disease Control (CDC) 2002 guidelines, which are
the same as the Centers for Disease Control and Pre-
vention 2006 guidelines. The researchers used a struc-
tured scoring process for determining the incidence of
BV that reduced the potential for measurement error
(see Figure 8-3) (Waltz et al., 2010). The design of
the study ensured the blinding of researchers to the
participants’ scores on their home self-test for BV
before they determined their own score. The follow-
ing study excerpts identify the controls the research-
ers used in their study to decrease the effect of
extraneous variables and increase the likelihood that
the findings are an accurate reflection of reality and
not due to error:
Sample Criteria
“Study eligibility was female participants who were
(a) volunteers, (b) aged 18 to 45 years, (c) not preg­
nant, and (d) able to read and understand English. A
prior history of vaginitis or vaginosis was not required
for eligibility. Participants came from three clinical
sites: a university college of nursing community well­
ness center, a private family practice, and a university
student health service site.” (Youngkin & Lester,
2010, p. 5)
Intervention
“The researchers provided the educational compo­
nent about BV, the BV home self-test system and safe
and correct use of the system, interpretation of self-
test results and scoring findings, and how to follow
up for provider evaluation after self-testing. Visual
and verbal instruction methods were employed.…
The BV self-test kit was provided free, and all com­
ponents were reviewed with the specific regimen for
use. Researchers selected the kit contents with con­
sideration for cost, future use, and ease of use. Ten
applications of self-testing materials and scoring
sheets were provided in the kit along with educa­
tional materials, a copy of the signed consent form,
laminated instructions for kit use, and a listing of con­
tents. Directions for follow-up with the researcher
clinicians were on the outside of each kit.” (Youngkin
& Lester, 2010, p. 5)
Measurement with BV Scoring Sheet
“The women were taught how to safely obtain a
vaginal specimen using directions outlined on a card
in the kit and to test and score the specimen using
the criteria of (a) discharge color and consistency, (b)
pH level, and (c) amine odor. The tests and scoring
were specific to these three of the four criteria that
CDC guidelines (2002) advised for providers to use
for clinical diagnosis. Figure 8-3 provides an example
of the scoring sheet. The specifics of the scoring
 CHAPTER 8  Objectives, Questions, Hypotheses, and Study Variables 153

BACTERIAL VAGINOSIS SELF-DIAGNOSIS

SCORING SHEET

Please mark the correct item by putting in the points in the blank space beside the item. For instance, if the
color of the strip of paper with your discharge is the number 4.5 or a smaller number, please put a “1” in
the first column, but if it is a color indicating a higher number, put a 2 in the second column. If the color of
the discharge is milky or creamy white or gray, put a 2 in the second colomn. If it is any other color, put a 1
in the first column. If there is no fishy odor, put a 1 in the first column. If there is a fishy odor, put a 2 in the
second column.

Remember: don’t worry about what you think you have; just put the number that “fits” the characteristic
the best in either Column 1 or 2. You should have only one number for the characteristic.

Add the numbers in the two columns up, then, add these two numbers together for the final score. For
example, if the total for Column 1 is a 2 and the total for Column 2 is 2, then the total score is 4.

Characterisitic Column 1 (non BV) Column 2 (BV)

Put a 1 in the column Put a 2 in the column
beside the correct finding beside the correct finding
Color of paper from
dispenser after it is wet
from vaginal discharge 4.5 or under 5 or higher

Color of discharge on
cotton of applicator yellow, green, white or gray
clumpy white, smooth or
brown, red creamy or frothy

Fishy smell when

discharge is mixed
with the KOH drop No fishy smell Fishy smell

Scores Column 1 Column 2

Add the scores from Column 1 and 2 together. The Total Score is

If your score is 4, 5, or 6, please make an appointment to see the clinician listed on the front of your
BV Self-Diagnosis Kit.

Figure 8-3  Scoring sheet for BV [bacterial vaginosis] home self-test. (From Youngkin, E. Q., & Lester, P. B. [2010]. Promoting self-care and secondary
prevention in women’s health: A study to test the accuracy of a home self-test system for bacterial vaginosis. Applied Nursing Research, 23[1], 6.)

Environmental variables are types of extraneous

method developed by the researchers and the form
for recording the self-test findings were taught as the
last component of the BV home self-test system.…
Thus, higher total scores would be associated with a
greater chance of BV if the woman was accurately
testing herself. The provider’s scoring would either
support or not support the self-test scoring.… The
women brought their scored results sheets with
them; the researchers did not review these results
until after the professional examination was com­
pleted and the findings had been scored on separate
scoring form.” (Youngkin & Lester, 2010, pp. 5-6)
variables that make up the setting in which the study
is conducted. Examples are climate, home, healthcare
system, community setting, and governmental organi-
zations. If a researcher is studying humans in an
uncontrolled or natural setting, it is impossible and
sometimes undesirable to control most of the environ-
mental variables. In qualitative and some quantitative
(descriptive and correlational) studies, researchers
make little or no attempt to control environmental
variables. Their intent is to study subjects in their
natural environment without controlling or altering it.
The environmental variables in quasi-experimental
and experimental research can be controlled through
154 UNIT TWO  The Research Process

the use of a study protocol and a laboratory setting or TABLE 8-2  Demographic Summary
a specially constructed research unit in a hospital. of Study Participants
Demographic Variables Initial Intervention
Demographic variables are attributes of the subjects Respondents Respondents
that are measured during the study and used to describe Attribute (n = 62) (n = 47)
the sample. Some common demographic variables Age (years) 37.74 ± 11.70 38.98 ± 12.21
examined in nursing research are age, gender, ethnicity, Time as a 9.82 ± 10.95 10.95 ± 11.87
educational level, income, job classification, length of registered
nurse (years)
hospital stay, and medical diagnosis. Researchers select
Shift length (hours):
demographic variables according to the focus of their   8 9 7
study, the demographic variables included in previous   12 52 40
studies, and clinical experience. However, age, gender, Shift pattern:
and ethnicity are essential demographic variables to   Days 25 17
examine in all types of research. These demographics   Evenings 4 2
describe the sample and determine the population for   Nights 29 25
generalization of the findings. More research is needed   Rotating 0 0
to improve health care for elderly, women, children, Relationship/marital status:
and minorities, and funding agencies often give priority   Single 15 12
  Spouse or 47 35
to studies that focus on these individuals.
significant
To obtain data on demographic variables, research- other
ers ask subjects to complete a demographic or infor- Additional 4 (6.5%) 3 (6.4%)
mation sheet. When the study is completed, the employment
demographic information is analyzed to provide a
picture of the sample, which is called the sample Note: Frequencies may not match sample size due to missing data.
characteristics. Sample characteristics are presented Adapted from Scott, L. D., Hofmeister, N., Rogness, N., & Rogers, A. E.
(2010). An interventional approach for patient and nurse safety: A fatigue
in a table and/or discussed in the narrative of the countermeasures feasibility study. Nursing Research, 59(4), 252.
research report. As previously discussed, Scott et al.
(2010) implemented the FCMPN intervention to
improve staff nurses’ sleep duration, sleep quality,
daytime sleepiness, drowsiness episodes, risk for auto The demographic variables in this study were eth-
accident, and actual and near work errors. These nicity, gender, age, time as a registered nurse in years,
researchers summarized their sample characteristics in shift length in hours, shift pattern, relationship/marital
a table (see Table 8-2) and discussed them in the nar- status, and additional employment (see description of
rative of their article, as follows: sample and Table 8-2). The participants from three
different sites were compared and found to have no
differences in demographic variables at the start of the
“Most participants were White (96.8%) and women study. These demographic variables seem appropriate
(96.8%), with an average age of 37.74 years (range = to describe the sample of staff nurses for this study. It
22-63 years, SD = 11.70 years). Fifty-two (83.9%) of was important to ensure that the participants from the
the nurses reported being employed in 12-hour day three different settings were similar at the start of
shifts positions. Although all three work shifts were the study so significant differences in the dependent
represented, the nurses worked predominantly variables might be assumed to be caused by the
12-hour day shifts (e.g., 7 : 00 a.m.-7:30 p.m.; n = 25) FCMPN intervention and not due to demographical
or night shift (e.g., 7:00 p.m.-7:30 a.m.; n = 29). Only differences.
4 nurses (6.5%) reported working evening shifts. The
average length of experience as hospital staff nurses Moderator and Mediator Variables
was 9.82 years (range = 1-43 years, SD = 11.83 years). Moderator and mediator variables are examined in
Although the nurses were employed in full-time posi­ intervention effectiveness research to improve our
tions, 4 nurses (6.5%) reported working a second understanding of the effect of the intervention on
job. No significant differences were noted between practice-related outcomes. A moderator variable
participants across the three sites before data aggre­ occurs with the intervention (independent variable)
gation.” (Scott et al., 2010, p. 252) and alters the causal relationship between the inter-
vention and the outcomes. Moderator variables include
 CHAPTER 8  Objectives, Questions, Hypotheses, and Study Variables 155

TABLE 8-3  Conceptual and Study Variables, Instruments, and Scores or Measurement
Possible Scores and
Concepts Variables Instruments Measurement
Sleep loss (deprivation or Sleep duration Logbook (self-report sleep Total sleep duration
disruption) times)
Sleep quality Pittsburgh Sleep Quality Index Global score = 0-21
Cognitive-behavioral Daytime sleepiness Epworth Sleepiness Scale Summative score = 0-24
outcomes
Vigilance (inability to remain Drowsiness and unplanned Frequencies
alert) sleep episodes at work and
while driving
Short-term memory Logbook (accident or error Frequencies
data)
Problem solving and coping Logbook (error description) Frequencies

From Scott, L. D., Hofmeister, N., Rogness, N., & Rogers, A. E. (2010). An interventional approach for patient and nurse safety: A fatigue countermeasures
feasibility study. Nursing Research, 59(4), 253.

characteristics of the subjects and of the person imple-
menting the intervention. Mediator variables bring
about the effects of the intervention after it has
occurred and thus influence the outcomes of the study.
The theoretical model that provides the framework for
the study usually identifies the relevant moderator and
mediator variables to be examined in the study. The
design is developed to examine not only the indepen-
dent (intervention) and dependent (outcomes) vari-
ables but also the moderator and mediator variables
(see Chapter 14 for a detailed discussion of interven-
tion research).
Operationalizing Variables
or Concepts for a Study
Conceptual and Operational Definitions of
Variables in Quantitative Studies
Operationalizing a variable or concept in quantita-
tive studies involves developing conceptual and oper-
ational definitions. A conceptual definition provides
the theoretical meaning of a concept or variable and
is derived from a theorist’s definition of that concept
or is developed through concept analysis. The study
framework, which includes concepts and their defini-
tions, provides a basis for conceptually defining the
variables. The Scott et al. (2010) study described pre-
viously in the sections on causal hypotheses, indepen-
dent and dependent variables, and demographic
variables is presented here as an example of how
to operationalize study variables. These researchers
implemented the FCMPN intervention with staff
nurses to improve the following outcomes: sleep
duration, sleep quality, daytime sleepiness, episodes
of drowsiness, risk of auto accident, and actual and
near work errors. These researchers linked their frame-
work concepts, variables, instruments of measure-
ment, and possible scores and measurement in Table
8-3. Conceptual and operational definitions of the
independent variable and three dependent variables in
this study are as follows:
Independent Variable: Fatigue
Countermeasures Program for
Nurses (FCMPN)
Conceptual definition: The FCMPN evolved from the
model of impaired sleep by Lee and colleagues (2004).
The FCMPN was thought to have the potential to
decrease sleep loss and improve sleep quality and
their associated cognitive-behavioral responses (Scott
et al., 2010).
Operational definition: “Comprehensive program to
manage fatigue in work settings usually includes the
following six elements: (a) education and training, (b)
compliance with hours of service regulations, (c)
appropriate scheduling practices, (d) countermea­
sures that can be instituted in the work setting, (e)
design, and (f) research.… The FCMPN used in this
study is modeled after the National Aeronautics and
Space Administration Ames Research Center’s
Fatigue Countermeasures Program . . . with additional
content obtained from the Sleep, Alertness, and
Fatigue Education in Residency Program (American
Academy of Sleep Medicine, 2006).” (Scott et al.,
2010, pp. 252-253)
156 UNIT TWO  The Research Process
Dependent Variables
Sleep Duration
Conceptual definition: Sleep duration is an indication
of the sleep loss (deprivation and/or disruption) and
fatigue that is experienced by an employee (see Table
8-3 for the link of the model concepts to this study
variable).
Operational definition: Sleep duration was measured
by self-report in a Logbook that involved the partici­
pants completing 15 items about their sleep patterns
(Scott et al., 2010).
Daytime Sleepiness
Conceptual definition: Daytime sleepiness is a cogni­
tive-behavior outcome resulting from employees
who experience sleep loss or poor quality sleep.
Operational definition: Daytime sleepiness was mea­
sured with the Epworth Sleepiness Scale with a sum­
mative score equal to 0 through 24.
Episodes of Drowsiness
Conceptual definition: Vigilance is a cognitive-behavioral
outcome that focuses on the ability and inability to
remain alert at work without drowsiness.
Operational definition: Vigilance was measured by
the drowsiness and unplanned sleep episodes at work
and while driving. These were part of the self-report
items in the participants’ Logbook.
The variables in quasi-experimental and experi-
mental quantitative research, outcomes research, and
intervention research are narrow and specific in focus
and are capable of being quantified (converted to
numbers) or manipulated through the use of specified
steps. In addition, the variables are objectively defined
to reduce researcher bias.
Conceptual Definitions in Qualitative Studies
The concepts in qualitative studies are usually more
abstract and broadly defined than the variables in quan-
titative, outcomes, and intervention studies. Many
researchers believe that the concepts in qualitative
studies do not have operational definitions because
sensitizing or experiencing the real situation rather than
operationalizing the concepts is most important. Oper-
ational definitions are not appropriate because they are
thought to limit the investigation so that a phenome-
non, such as pain, or a characteristic of a culture, such
as health practices, is not completely experienced or
understood. In qualitative research, often the phenom-
enon being studied is not clearly identified and/or
defined until the results of the study are determined
(Marshall & Ross, 2011; Munhall, 2012). Thus, some
concepts may not be clearly conceptually defined until
late in the conduct of the study. For example, Hudson
et al. (2010) developed the following description of
the health-seeking challenges among homeless youth
during their discussion of study results:
Conceptual Definition of Health-Seeking
Challenges among Homeless Youth
“Participants were quite verbal in expressing their
perspectives and experiences in order that positive
changes could be made. The major themes expressed
related to the issues they experienced in accessing
health care, followed by barriers that were experi­
enced, as well as a pervasive sense of stigma and
discrimination, which was quite telling.…
Failing Access to Health Care
“Homeless young adults revealed that accessing
health care was challenging due to scarcity of service
sites and generally long waiting times for services, yet
homeless young adults experienced a number of
health problems ranging from chronic conditions such
as migraine headaches to more serious conditions of
asthma and meningitis. Mental health issues were
reported most commonly.…
Needing More Help
“Homeless young persons experienced a number of
unmet needs.… Despite their hardships, young adults
craved for support from family, friends, and homeless
peers.…
Perceiving Stigma
“The youth were most frustrated by the discrimina­
tion they experienced from passersby and law
enforcement. Rather than provide resources for
homeless youth, youth frequently were confronted
with unforgettable comments that were full of
judgment.…
Making it Work
“Regardless of the challenges experienced, homeless
young adults were able to deal with life in the only
ways they knew how and were grateful for the help
and support of others. For several homeless young
adults, they were grateful for the information they
received from the more experienced homeless young
persons in terms of when services were available and
where to go.” (Hudson et al., 2010, pp. 215-216)
 CHAPTER 8  Objectives, Questions, Hypotheses, and Study Variables 157
KEY POINTS
• Research objectives, questions, and hypotheses are
formulated to bridge the gap between the more
abstractly stated research problem and purpose and
the detailed design and plan for data collection and
analysis.
• Research objectives are clear, concise, declarative
statements that are expressed in the present tense
to provide specific focus to the conduct of a
study.
• A research question is a concise, interrogative
statement that is worded in the present tense and
consists of one or more variables (or concepts).
• A hypothesis is the formal statement of the expected
relationships between two or more variables in a
selected population.
• Hypotheses can be described in terms of four cat-
egories: (1) associative versus causal, (2) simple
versus complex, (3) nondirectional versus direc-
tional, and (4) null versus research.
• Selecting objectives, questions, or hypotheses for a
study is based on (1) the number and quality of
relevant studies conducted on a selected problem
(existing knowledge base), (2) the framework of
the study, (3) the expertise and preference of the
researcher, and (4) the type of study to be con-
ducted (quantitative, qualitative, outcomes, and
intervention).
• Variables are qualities, properties, or characteristics
of persons, things, or situations that change or vary
in a study.
• The types of variables discussed in this chapter are
independent, dependent, research, extraneous,
demographic, moderator, and mediator.
• An independent variable is an intervention or
experimental treatment that the researcher manipu-
lates or varies to create an effect on the dependent
variable.
• A dependent variable is the outcome, response, or
behavior that the researcher wants to predict or
explain and is measured in the study.
• Research variables or concepts are the qualities,
properties, or characteristics that are measured
in selected quantitative studies (descriptive and
correlational).
• Demographic variables are attributes of the research
participants that are measured in a study to describe
the sample. The summary of the data collected on
demographic variables is called sample character-
istics and is used to describe the sample.
• The variables in quantitative studies require
conceptual and operational definitions, and a
conceptual definition provides the theoretical
meaning of a concept or variable and is derived
from a theorist’s definition of the concept or is
developed through concept analysis.
• An operational definition is derived from a set of
procedures or progressive acts that a researcher
performs either to manipulate an independent
variable or to measure the existence or degree of
existence of the dependent variable or research
variable.
• Concepts in qualitative research are usually con-
ceptually defined when the results or findings of the
study are presented.
REFERENCES
American Academy of Sleep Medicine. (2006). Sleep Alertness &
Fatigue Education in Residency (SAFER). Retrieved from http://
www.aAmsment.org/.
Bowles, K. H., Pham, J., O’Connor, M., & Horowitz, D. A. 2010.
Information deficits in home care: A barrier to evidence-based
disease management. Home Health Care Management and Prac-
tice, 22(4), 278–285.
Brown, S. J. (2009). Evidence-based nursing: The research-practice
connection. Sudbury, MA: Jones and Bartlett Publishers.
Bucher, C. E. (2008). Toward a needs-based typology of homeless
youth. Journal of Adolescent Health, 42(6), 549–554.
Centers for Disease Control. (2002). Sexually transmitted diseases
treatment guidelines 2002. Morbidity and Mortality Weekly
Report, 52(RR-6), 42–44.
Centers for Disease Control and Prevention. (2006). Sexually trans-
mitted disease treatment guidelines, 2006. Morbidity and Mortal-
ity Weekly Report, 53(3), 1–94.
Chinn, P. L., & Kramer, M. K. (2008). Integrated theory and knowl-
edge development (7th ed.). St. Louis: Mosby.
Craig, J. V., & Smyth, R. L. (2012). The evidence-based practice
manual for nurses (3rd ed.). Edinburgh, Scotland: Churchill
Livingstone.
Creswell, J. W. (2009). Research design: Qualitative, quantitative,
and mixed methods approaches (3rd ed.). Los Angeles, CA: Sage.
Delaney, C., Apostolidis, B., Lachapelle, L., & Fortinsky, R. (2011).
Home care nurses’ knowledge of evidence-based education topics
for management of heart failure. Heart & Lung, 40(4),
285–292.
Diefenbach, M. A., & Leventhal, H. (1996). The Common-Sense
Model of Illness Representation: Theoretical and practical consid-
erations. Journal of Social Distress and the Homeless, 5(1),
11–38.
Doran, D. (2011). Nursing-sensitive outcomes: The state of the
science (2nd ed.). Sudbury, MA: Jones & Bartlett Learning.
Efe, E., & Özer, Z. C. (2007). The use of breast-feeding for pain
relief during neonatal immunization injections. Applied Nursing
Research, 20(1), 10–16.
Fagerhaugh, S. Y., & Strauss, A. (1977). Politics of pain manage-
ment. Menlo Park, CA: Addison-Wesley.
Fawcett, J., & Garity, J. (2009). Evaluating research for evidence-
based nursing practice. Philadelphia, PA: F. A. Davis.
158 UNIT TWO  The Research Process
Forbes, A. (2009). Clinical intervention research in nursing. Inter-
national Journal of Nursing Studies, 46(4), 557–568.
Happ, M. B., Swigart, V. A., Tate, J. A., Hoffman, L. A., & Arnold,
R. M. (2007). Patient involvement in health-related decisions
during prolonged critical illness. Research in Nursing & Health,
30(4), 361–372.
Hodges, P. (2009). Heart failure: Epidemiologic update. Critical
Care Nursing Quarterly, 32(1), 24–32.
Hudson, A. L., Nyamathi, A., Greengold, B., Slagle, A., Koniak-
Griffin, D., Khalilifard, F., et al. (2010). Health-seeking chal-
lenges among homeless youth. Nursing Research, 59(3),
212–218.
Hudson, A. L., Nyamathi, A., & Sweat, J. (2008). Homeless youths’
interpersonal perspectives of health care providers. Issues in
Mental Health Nursing, 29(12), 1277–1289.
Kaplan, A. (1964). The conduct of inquiry: Methodology for behav-
ioral science. New York, NY: Harper & Row.
Kerlinger, F. N., & Lee, H. B. (2000). Foundations of behavioral
research (4th ed.). Fort Worth, TX: Harcourt College
Publishers.
Lee, K. A., Landis, C., Chasens, E. R., Dowling, G., Merritt, S.,
Parker, K. P., et al. (2004). Sleep and chronobiology: Recom-
mendations for nursing education. Nursing Outlook, 52(3),
126–133.
Marshall, C., & Rossman, G. B. (2011). Designing qualitative
research (5th ed.). Los Angeles, CA: Sage.
Munhall, P. L. (2012). Nursing research: A qualitative perspective
(5th ed.). Sudbury, MA: Jones & Bartlett Learning.
Orem, D. E. (2001). Nursing concepts of practice. St. Louis, MO:
Mosby.
Reishtein, J. L. (2005). Relationship between symptoms and func-
tional performance in COPD. Research in Nursing & Health,
28(1), 39–47.
Ross, R., Sawatphanit, W., Mizuno, M., & Takeo, K. (2011).
Depressive symptoms among HIV-positive postpartum women in
Thailand. Achieves of Psychiatric Nursing, 25(1), 36–42.
Rungruangsiripan, M., Sitthimongkol, Y., Maneesriwongul, W.,
Talley, S., & Vorapongsathorn, T. (2011). Mediating role of illness
representation among social support, therapeutic alliance,
experience of medication side effects, and medication adherence
in persons with schizophrenia. Archives of Psychiatric Nursing,
25(4), 269–283.
Schulz, P. S., Zimmerman, L., Pozehl, B., Barnason, S., & Nieveen,
J. (2011). Symptom management strategies used by elderly
patients after coronary artery bypass surgery. Applied Nursing
Research, 24(2), 65–73.
Scott, L. D., Hofmeister, N., Rogness, N., & Rogers, A. E. (2010).
An interventional approach for patient and nurse safety: A fatigue
countermeasures feasibility study. Nursing Research, 59(4),
250–258.
Shadish, W. R., Cook, T. D., & Campbell, D. T. (2002). Experimen-
tal and quasi-experimental designs for generalized causal infer-
ence. Chicago, IL: Rand McNally.
Smith, M. J., & Liehr, P. R. (2008). Middle range theory for nursing
(2nd ed.). New York, NY: Springer Publishing Company.
Sneller, V. P., Fischbein, D. B., Weinbaum, C. M., Lombard, A.,
Murray, P., McLaurin, J. A., et al. (2008). Vaccinating adolescents
in high-risk settings: Lessons learned from experiences with hepa-
titis B vaccine. Pediatrics, 121(Suppl. 1), S55–S62.
Tranmer, J. E., & Parry, M. J. (2004). Enhancing postoperative
recovery of cardiac surgery patients: A randomized clinical trial
of an advanced practice nursing intervention. Western Journal of
Nursing Research, 26(5), 515–532.
Vasan, R. S., Beiser, A., D’Agostino, R. B., Levy, D., Selhub, J.,
Jacques, P. F., et al. (2003). Plasma homocysteine and risk for
congestive heart failure in adults without prior myocardial infarc-
tion. Journal of the American Medical Association, 289(10),
1251–1257.
Waltz, C. F., Strickland, O. L., & Lenz, E. R. (2010). Measurement
in nursing and health research (4th ed.). New York, NY: Springer.
Youngkin, E. Q., & Lester, P. B. (2010). Promoting self-care and
secondary prevention in women’s health: A study to test the accu-
racy of a home self-test system for bacterial vaginosis. Applied
Nursing Research, 23(1), 2–10.
Zimmerman, L., Barnason, S., Nieveen, J., & Schmaderer, M.
(2004). Symptom management intervention in elderly coronary
artery bypass graft patients. Outcomes Management, 8(1), 5–12.
   http://evolve.elsevier.com/Grove/practice/
N
ursing research requires not only expertise and
diligence but also honesty and integrity. Con-
ducting research ethically starts with the iden-
tification of the study topic and continues through the
publication of the study. Over the years, ethical codes
and regulations have been developed to provide guide-
lines for (1) selecting a study purpose, design, and
subjects; (2) collecting and analyzing data; (3) inter-
preting study results; and (4) presenting and publish-
ing a study. One of the more recent regulations, the
Health Insurance Portability and Accountability Act
(HIPAA), was enacted in 2003 to protect the privacy
of an individual’s health information. HIPAA has had
an important impact on researchers and institutional
review boards (IRBs) in universities and healthcare
agencies. This chapter provides an overview of this
act and the other United States and international regu-
lations that have been developed to promote the ethical
conduct of research.
An ethical problem that has received increasing
attention since the 1980s is research misconduct. Mis-
conduct has occurred during the conduct, reporting,
and publication of studies, and the Office of Research
Integrity (ORI, 2012) was developed to manage this
problem. Many disciplines, including nursing, have
experienced episodes of research misconduct that
have affected the quality of research evidence gener-
ated and disseminated.
Ethical research is essential to generate a sound
evidence-based practice for nursing, but what does the
ethical conduct of research involve? This question has
been debated for many years by researchers, politi-
cians, philosophers, lawyers, and even research sub-
jects. The debate continues, probably because of the
complexity of human rights issues; the focus of
research in new, challenging arenas of technology and
genetics; the complex ethical codes and regulations
governing research; and the various interpretations of
these codes and regulations. Even though the
9  
CHAPTER
Ethics in Research
phenomenon of the ethical conduct in research defies
precise delineation, the historical events, ethical codes,
and regulations presented in this chapter provide guid-
ance for nurse researchers. The chapter also discusses
the actions essential for conducting research ethically:
(1) protecting the rights of human subjects; (2) balanc-
ing benefits and risks in a study; (3) obtaining informed
consent from study participants; and (4) submitting a
research proposal for institutional review. A discus-
sion of current ethical issues related to research mis-
conduct and animals used in research concludes the
chapter.
Historical Events Affecting the
Development of Ethical Codes
and Regulations
The ethical conduct of research has been a focus since
the 1940s because of the mistreatment of human sub-
jects in selected studies. Four experimental projects
have been highly publicized for their unethical treat-
ment of subjects: (1) Nazi medical experiments; (2)
the Tuskegee syphilis study; (3) the Willowbrook
study; and (4) the Jewish Chronic Disease Hospital
study. Although these were biomedical studies and the
primary investigators were physicians, there is evi-
dence that nurses were aware of the research, identi-
fied potential subjects, delivered treatments to the
subjects, and served as data collectors in these proj-
ects. The four projects demonstrate the importance of
ethical conduct for anyone reviewing, participating in,
and conducting nursing or biomedical research. These
four projects and other incidences of unethical treat-
ment of subjects and research misconduct in the devel-
opment, implementation, and reporting of research
have influenced the formulation of ethical codes and
regulations that direct research today. In addition, the
concern for patient privacy with the electronic storage
159
160 UNIT TWO  The Research Process
and exchange of health information has resulted in
HIPAA privacy regulations (Olsen, 2003).
Nazi Medical Experiments
From 1933 to 1945, the Third Reich in Europe imple-
mented atrocious, unethical activities (Steinfels &
Levine, 1976). The programs of the Nazi regime con-
sisted of sterilization, euthanasia, and numerous
medical experiments to produce a population of
racially pure Germans, or Aryans, who the Nazis
maintained were destined to rule the world. The Nazis
encouraged population growth among the Aryans
(“good Nazis”) and sterilized people they regarded as
racial enemies, such as the Jews. They also practiced
what they called “euthanasia,” which involved killing
various groups of people whom they considered
racially impure, such as the insane, deformed, and
senile. In addition, researchers conducted numerous
medical experiments on prisoners of war as well as on
racially “valueless” persons who had been confined to
concentration camps.
The medical experiments involved exposing sub-
jects to high altitudes, freezing temperatures, malaria,
poisons, spotted fever (typhus), and untested drugs and
operations, usually without any anesthesia (Steinfels
& Levine, 1976). These medical experiments were
conducted to generate knowledge about human beings,
but the goal often was to destroy certain groups of
people. Extensive examination of the records from
some of these studies showed that they were poorly
designed and conducted. Thus, they generated little if
any useful scientific knowledge.
The Nazi experiments violated numerous rights of
the research participants. Researchers selected sub-
jects on the basis of their race, demonstrating an unfair
selection process. The subjects also had no opportu-
nity to refuse participation; they were prisoners who
were coerced or forced to participate. The study par-
ticipants were frequently killed during the experi-
ments or sustained permanent physical, mental, and
social damage (Levine, 1986; Steinfels & Levine,
1976). The mistreatment of human subjects in these
Nazi studies led to the development of the Nuremberg
Code in 1949.
Nuremberg Code
The people involved in the Nazi experiments were
brought to trial before the Nuremberg Tribunals,
which publicized their unethical activities. These
unethical studies resulted in the Nuremberg Code
(1949), which was developed with guidelines for (1)
subjects’ voluntary consent to participate in research;
(2) the right of subjects to withdraw from studies; (3)
protection of subjects from physical and mental suf-
fering, injury, disability, and death during studies; and
(4) the balance of benefits and risks in a study. Box
9-1 reproduces the Nuremberg Code, which was for-
mulated mainly to direct the conduct of biomedical
research worldwide; however, the rules it contains are
essential to research in other sciences, such as nursing,
psychology, and sociology.
Declaration of Helsinki
The Nuremberg Code provided the basis for the devel-
opment of the Declaration of Helsinki in 1964 by the
World Medical Association (WMA) General Assem-
bly. Over the years, the Declaration of Helsinki has
been amended as needed with the final amendment in
2008 (WMA General Assembly). This document dif-
ferentiates therapeutic research from nontherapeutic
research. Therapeutic research gives the patient
an opportunity to receive an experimental treatment
that might have beneficial results. Nontherapeutic
research is conducted to generate knowledge for a
discipline, and the results from the study might benefit
future patients but will probably not benefit those
acting as research subjects.
The Declaration of Helsinki includes ethical prin-
ciples for medical research involving human subjects,
such as the following: (1) well-being of the individual
research subject must take precedence over all other
interests; (2) a strong, independent justification must
be documented prior to exposing healthy volunteers
to risk of harm just to gain new scientific information;
(3) investigators must protect the life, health, privacy,
and dignity of research subjects; and (4) extreme care
must be taken in making use of placebo-controlled
trials, which should be used only in the absence of
existing proven therapy (WMA General Assembly,
2008). Clinical trials must focus on improving diag-
nostic, therapeutic, and prophylactic procedures for
patients with selected diseases without exposing sub-
jects to any additional risk of serious or irreversible
harm. Most institutions worldwide in which clinical
research is conducted have adopted the Declaration of
Helsinki. However, neither this document nor the
Nuremberg Code has prevented some investigators
from conducting unethical research (Beecher, 1966;
ORI, 2012).
Tuskegee Syphilis Study
In 1932, the U.S. Public Health Service (U.S. PHS)
initiated a study of syphilis in black men in the small,
rural town of Tuskegee, Alabama (Brandt, 1978;
Rothman, 1982). The study, which continued for 40
years, was conducted to determine the natural course

Box 9-1 The Nuremberg Code
1. The voluntary consent of the human subject
is absolutely essential.
2. The experiment should be such as to yield
fruitful results for the good of society,
unprocurable by other methods or means of
study, and not random and unnecessary in
nature.
3. The experiment should be so designed and
based on the results of animal
experimentation and knowledge of the
natural history of the disease or other
problem under study that the anticipated
results will justify the performance of the
experiment.
4. The experiment should be so conducted as
to avoid all unnecessary physical and mental
suffering and injury.
5. No experiment should be conducted where
there is a prior reason to believe that death
or disabling injury will occur, except,
perhaps, in those experiments where the
experimental physicians also serve as
subjects.
6. The degree of risk to be taken should never
exceed that determined by the humanitarian
importance of the problem to be solved by
the experiment.
From Nuremberg Code. (1949). Trials of War Criminals before the Nuremberg Military Tribunals under Control
Council Law No. 10, Vol. 2, pp. 181–182. Washington, D.C.: U.S. Government Printing Office, 1949. Retrieved from
http://ohsr.od.nih.gov/guidelines/nuremberg.html/.
of syphilis in the adult black male. The research sub-
jects were organized into two groups: one group con-
sisted of 400 men who had untreated syphilis, and the
other was a control group of 200 men without syphilis.
Many of the subjects who consented to participate in
the study were not informed about the purpose and
procedures of the research. Some individuals were
unaware that they were subjects in a study.
By 1936, study results indicated that the men with
syphilis experienced more health complications than
the control group. Ten years later, the death rate of the
group with syphilis was twice as high as that of the
control group. The subjects were examined periodi-
cally but were not treated for syphilis, even after peni-
cillin was determined to be an effective treatment for
the disease in the 1940s (Brandt, 1978). Published
reports of the Tuskegee syphilis study first started
CHAPTER 9  Ethics in Research 161
7. Proper preparations should be made and
adequate facilities provided to protect the
experimental subject against even remote
possibilities of injury, disability, or death.
8. The experiment should be conducted only by
scientifically qualified persons. The highest
degree of skill and care should be required
through all stages of the experiment of
those who conduct or engage in the
experiment.
9. During the course of the experiment the
human subject should be at liberty to bring
the experiment to an end if he has reached
the physical or mental state where
continuation of the experiment seems to him
to be impossible.
10. During the course of the experiment the
scientist in charge must be prepared to
terminate the experiment at any stage, if he
has probable cause to believe, in the exercise
of the good faith, superior skill, and careful
judgment required of him, that a
continuation of the experiment is likely to
result in injury, disability, or death to the
experimental subject.
appearing in 1936, and additional papers were pub-
lished every 4 to 6 years. In 1969, the U.S. Centers
for Disease Control (CDC) reviewed the study and
decided that it should continue. In 1972, a story
describing the study published in the Washington Star
sparked public outrage. Only then did the U.S. Depart-
ment of Health, Education, and Welfare (DHEW) stop
the study. An investigation of the Tuskegee syphilis
study found it to be ethically unjustified.
Willowbrook Study
From the mid-1950s to the early 1970s, Dr. Saul
Krugman at Willowbrook, an institution for the men-
tally retarded, conducted research on hepatitis
(Rothman, 1982). The subjects, all children, were
deliberately infected with the hepatitis virus. During
the 20-year study, Willowbrook closed its doors to
162 UNIT TWO  The Research Process
new inmates because of overcrowded conditions.
However, the research ward continued to admit new
inmates. To gain their child’s admission to the institu-
tion, parents were forced to give permission for the
child to be a subject in the study.
From the late 1950s to early 1970s, Krugman’s
research team published several articles describing the
study protocol and findings. Beecher (1966) cited
the Willowbrook study as an example of unethical
research. The investigators defended injecting the
children with the virus by citing their own belief that
most of the children would have acquired the infection
after admission to the institution. The investigators
also stressed the benefits that the subjects received,
which were a cleaner environment, better supervision,
and a higher nurse-patient ratio on the research ward
(Rothman, 1982). Despite the controversy, this unethi-
cal study continued until the early 1970s.
Jewish Chronic Disease Hospital Study
Another highly publicized example of unethical
research was a study conducted at the Jewish Chronic
Disease Hospital in the 1960s. Its purpose was to
determine the patients’ rejection responses to live
cancer cells. Twenty-two patients were injected with
a suspension containing live cancer cells that had been
generated from human cancer tissue (Levine, 1986).
An extensive investigation of this study revealed
that the patients were not informed that they were
taking part in research or that the injections they
received were live cancer cells. In addition, the Jewish
Chronic Disease Hospital’s institutional review board
never reviewed the study; even the physicians caring
for the patients were unaware that the study was
being conducted. The physician directing the research
was an employee of the Sloan-Kettering Institute for
Cancer Research, and there was no indication that
this institution had reviewed the research project
(Hershey & Miller, 1976). The study was considered
unethical and was terminated, with the researcher
being in violation of the Nuremberg Code (1949) and
the Declaration of Helsinki (WMA General Assem-
bly, 1964). This research had the potential to cause
the study participants serious or irreversible harm and
possibly death and reinforced the importance of con-
scientious institutional review and ethical researcher
conduct.
U.S. Department of Health, Education, and
Welfare Regulations
The continued conduct of harmful, unethical research
made additional controls necessary. In 1973, the
DHEW published its first set of regulations intended
to protect human subjects. Clinical researchers were
presented with strict regulations for research involving
humans, with additional regulations to protect persons
having limited capacities to consent, such as the ill,
mentally impaired, and dying (Levine, 1986). All
research involving human subjects had to undergo full
institutional review, even nursing studies that involved
minimal or no risks to study participants. Institutional
review improved the protection of subjects’ rights;
however, reviewing all studies, without regard for the
degree of risk involved, overwhelmed the review
process and greatly prolonged the time required for
study approval. The government recognized the need
for additional strategies to manage the problems
related to the DHEW regulations.
National Commission for the Protection of
Human Subjects of Biomedical and
Behavioral Research
Because of the problems related to the DHEW regula-
tions, the National Commission for the Protection of
Human Subjects of Biomedical and Behavioral
Research (1978) was formed. The goals of the com-
mission were (1) to identify the basic ethical principles
that should underlie the conduct of biomedical and
behavioral research involving human subjects and (2)
to develop guidelines based on these principles. The
commission developed The Belmont Report (available
online at http://www.fda.gov/). This report identified
three ethical principles as relevant to research involv-
ing human subjects: the principles of respect for
persons, beneficence, and justice. The principle of
respect for persons holds that persons have the right
to self-determination and the freedom to participate or
not participate in research. The principle of benefi-
cence requires the researcher to do good and “above
all, do no harm.” The principle of justice holds that
human subjects should be treated fairly. Currently,
these ethical principles must be followed when
researchers in the United States and internationally
conduct studies. The commission developed ethical
research guidelines based on these three principles,
made recommendations to the U.S. Department of
Health and Human Services (U.S. DHHS), and was
dissolved in 1978.
In response to the commission’s recommendations,
the U.S. DHHS developed federal regulations in 1981
to protect human research subjects, which have been
revised as needed over the last 30 years. The most
current U.S. DHHS (2009) regulations are part of the
Code of Federal Regulations (CFR), Title 45, Part 46,
Protection of Human Subjects (available online at

http://www.hhs.gov/ohrp/policy/ohrpregulations
.pdf/). These regulations are interpreted by the Office
for Human Research Protection (OHRP), an agency
within U.S. DHHS (2012), whose functions include:
(1) providing guidance and clarification of regula-
tions; (2) developing educational programs and mate-
rials; (3) maintaining regulatory oversight of research;
and (4) providing advice on ethical and regulatory
issues related to biomedical and social-behavior
research.
The U.S. DHHS (2009) regulations provide direc-
tion for (1) the protection of human subjects in
research, with additional protection for pregnant
women, human fetuses, neonates, children, and pris-
oners; (2) the documentation of informed consent; and
(3) the implementation of the institutional review
board process. These regulations apply to all research
involving human subjects in the following areas: (1)
studies conducted, supported, or otherwise subject to
regulations by any federal department or agency; (2)
research conducted in educational and healthcare set-
tings; (3) research involving the use of biophysical
measures, educational tests, survey procedures, scales,
interview procedures, or observation; and (4) research
involving the collection or study of existing data,
documents, records, pathological specimens, or diag-
nostic specimens.
Essentially all the biomedical and behavioral
studies conducted in the United States are governed
by the U.S. DHHS (2009) Protection of Human Sub-
jects Regulations or the U.S. Food and Drug Admin-
istration (U.S. FDA). The FDA, within the U.S.
DHHS, manages the CFR Title 21, Food and Drugs,
Part 50, Protection of Human Subjects (U.S. FDA,
2010a), and Part 56, Institutional Review Boards
(IRBs) (U.S. FDA, 2010b). These regulations apply to
studies of drugs for humans, medical devices for
human use, biological products for human use, human
dietary supplements, and electronic products. The role
of the FDA was expanded by the Food and Drug
Administration Amendments Act (FDAAA) of 2007
to include increased responsibility for the manage-
ment of new drugs and medical devices. Physicians
and nurses conducting clinical trials to generate new
drugs and refine existing drug treatments must comply
with these FDA regulations. In summary, these regula-
tions focus on the protection of human subjects’ rights,
informed consent (U.S, FDA, 2010a), and IRBs (U.S.
FDA, 2010b), with content that is consistent with the
U.S. DHHS (2009) regulations.
The U.S. DHHS and FDA regulations provide
guidelines to protect subjects in federally and pri-
vately funded research by ensuring privacy and
CHAPTER 9  Ethics in Research 163
confidentiality of information obtained through
research. However, with the advent of electronic
access and transfer, the public was concerned about
the potential abuses of the health information of indi-
viduals in all circumstances, including research proj-
ects. Thus HIPAA was implemented in 2003 to protect
an individual’s health information. The U.S. DHHS
developed regulations titled the Standards for Privacy
of Individually Identifiable Health Information, and
compliance with these regulations is known as the
Privacy Rule (U.S. DHHS, 2003). The HIPAA Privacy
Rule established the category of protected health
information (PHI), which allows covered entities,
such as health plans, healthcare clearinghouses, and
healthcare providers that transmit health information,
to use or disclose PHI to others only in certain situa-
tions. These situations are discussed later in this
chapter.
The HIPAA Privacy Rule affects not only the health-
care environment but also the research conducted in
this environment (U.S. DHHS, 2010). An individual
must provide his or her signed permission, or authori-
zation, before his or her PHI can be used or disclosed
for research purposes. To determine how the HIPAA
Privacy Rule might impact the informed consent
and IRB processes for your study, go to the website
at http://privacyruleandresearch.nih.gov/, which was
developed to address researchers’ questions.
Table 9-1 was developed to clarify the overall
objectives and applicability of the HIPAA Privacy
Rule, U.S. DHHS Protection of Human Subjects Reg-
ulations, and U.S. FDA Protection of Human Subjects
Regulations (U.S. DHHS, 2007a). Any study you
propose with human subjects must comply with these
regulations. Thus, this chapter covers these regula-
tions in the sections on protecting human rights,
obtaining informed consent, and institutional review
of research.
Protection of Human Rights
Human rights are claims and demands that have been
justified in the eyes of an individual or by the consen-
sus of a group of individuals. Having rights is neces-
sary for the self-respect, dignity, and health of an
individual (Fry, Veatch, & Taylor, 2011). The Ameri-
can Nurses Association (ANA, 2001) Code of Ethics
for Nurses and the American Psychological Associa-
tion (APA, 2010) Principles of Psychologists and
Code of Conduct provide guidelines for protecting the
rights of human subjects in biological and behavioral
research. Researchers and reviewers of research
have an ethical responsibility to protect the rights of
164 UNIT TWO  The Research Process

TABLE 9-1  Clarification of the Focus of Federal Regulations and Impact on Research
Area of
Distinction HIPAA Privacy Rule
Overall objective Establishes a federal floor of
privacy protections for most
individually identifiable health
information by establishing
conditions for its use and
disclosure by certain healthcare
providers, health plans, and
healthcare clearinghouses.
Applicability Applies to HIPAA-defined
covered entities, regardless of
the source of funding.
U.S. DHHS Protection of
Human Subjects Regulations
Title 45 CFR Part 46
To protect the rights and welfare
of human subjects involved in
research conducted or
supported by U.S. DHHS.
Not specifically a privacy
regulation.
Applies to human subject
research conducted or
supported by U.S. DHHS and
research with private funding.
U.S. FDA Protection of Human
Subjects Regulations Title 21
CFR Parts 50 and 56
To protect the rights, safety, and
welfare of subjects involved in
clinical investigations regulated by
the FDA. Not specifically a privacy
regulation.
Applies to research involving
products regulated by the FDA.
Federal support is not necessary for
FDA regulations to be applicable.
When research subject to FDA
jurisdiction is federally funded,
both the U.S. DHHS Protection of
Human Subjects Regulations and
FDA Protection of Human Subjects
Regulations apply.

CFR, Code of Federal Regulations; DHHS, U.S. Department of Health and Human Services; U.S. FDA, U.S. Food and Drug Administration; HIPAA, Health
Insurance Portability and Accountability Act.
From U.S. Department of Health and Human Services. (2007a). How do other privacy protections interact with the privacy rule? Retrieved from http://
privacyruleandresearch.nih.gov/pr_05.asp/.

human research participants. The human rights that
require protection in research are (1) the right to self-
determination;, (2) the right to privacy; (3) the right
to anonymity and confidentiality; (4) the right to fair
treatment or justice; and (5) the right to protection
from discomfort and harm (ANA, 2001; APA, 2010;
Fry et al., 2011).
Right to Self-Determination
The right to self-determination is based on the ethical
principle of respect for persons. This principle holds
that because humans are capable of self-determination,
or controlling their own destinies, they should be
treated as autonomous agents who have the freedom
to conduct their lives as they choose without external
controls. As a researcher, you treat prospective sub-
jects as autonomous agents by informing them about
a proposed study and allowing them to voluntarily
choose to participate or not. In addition, subjects have
the right to withdraw from a study at any time without
a penalty (Fry et al., 2011). Conducting research ethi-
cally requires that research subjects’ right to self-
determination not be violated and that persons with
diminished autonomy have additional protection
during the conduct of studies (U.S. DHHS, 2009).
Preventing Violation of Research Subjects’ Right to
Self-Determination
A subject’s right to self-determination can be violated
through the use of (1) coercion; (2) covert data collec-
tion; and (3) deception. Coercion occurs when one
person intentionally presents another with an overt
threat of harm or the lure of excessive reward to obtain
his or her compliance. Some subjects are coerced to
participate in research because they fear that they will
suffer harm or discomfort if they do not participate. For
example, some patients believe that their medical or
nursing care will be negatively affected if they do not
agree to be research subjects. Sometimes students feel
forced to participate in research to protect their grades
or prevent negative relationships with the faculty con-
ducting the research. Other subjects are coerced to
participate in studies because they believe that they
cannot refuse the excessive rewards offered, such as
large sums of money, specialized health care, special
privileges, and jobs. Most nursing studies do not offer
excessive rewards to subjects for participating. Some-
times nursing studies have included a small financial
reward of $10 to $30 or support for transportation to
increase participation, but this would not be considered
coercive (Fawcett & Garity, 2009; Fry et al., 2011).

An individual’s right to self-determination can also
be violated if he or she becomes a research subject
without realizing it. Some researchers have exposed
persons to experimental treatments without their
knowledge, a prime example being the Jewish Chronic
Disease Hospital study. Most of the patients and their
physicians were unaware of the study. The subjects
were informed that they were receiving an injection
of cells, but the word cancer was omitted (Beecher,
1966). With covert data collection, subjects are
unaware that research data are being collected because
the investigator develops a description of the study
indicating that it is normal activity or part of health
care (Reynolds, 1979). This type of data collection has
more commonly been used by psychologists to
describe human behavior in a variety of situations, but
it has also been used by nursing and other disciplines
(APA, 2010). Qualitative researchers have debated
this issue, and some believe that certain group and
individual behaviors are unobservable within the
normal ethical range of research activities, such as the
actions of cults or the aggressive or violent behaviors
of individuals. Thus, these types of behaviors require
study with covert data collection processes. However,
covert data collection is considered unethical when
research deals with sensitive aspects of an individual’s
behavior, such as illegal conduct, sexual behavior, and
drug use (U.S. DHHS, 2009). With the HIPAA Privacy
Rule (U.S. DHHS, 2003), the use of any type of covert
data collection would be questionable and illegal if
PHI data were being used or disclosed.
The use of deception in research can also violate
a subject’s right to self-determination. Deception is the
actual misinforming of subjects for research purposes
(Kelman, 1967). A classic example of deception is the
Milgram (1963) study, in which the subjects thought
they were administering electric shocks to another
person. The subjects were unaware that the person was
really a professional actor who pretended to feel the
shocks. Some subjects experienced severe mental
tension, almost to the point of collapse, because of
their participation in this study. The use of deception
still occurs in some healthcare, social, and psychologi-
cal investigations, but it is a controversial research
activity. If deception is to be used in a study, research-
ers must determine that there is no other way to gain
the essential research data needed and that the subjects
will not be harmed. In addition, the subjects must be
informed of the deception once the study is completed,
provided full disclosure of the study activities that
were conducted, (APA, 2010; Fry, 2011; U.S. DHHS,
2009) and given the opportunity to withdraw their data
from the study.
CHAPTER 9  Ethics in Research 165
Protecting Persons with Diminished Autonomy
Some persons have diminished autonomy or are vul-
nerable and less advantaged because of legal or mental
incompetence, terminal illness, or confinement to an
institution (Fry et al., 2011). These persons require
additional protection of their right to self-determination,
because they have a decreased ability, or an inability,
to give informed consent. In addition, these persons
are vulnerable to coercion and deception. The U.S.
DHHS (2009) has identified certain vulnerable groups
of individuals, including pregnant women, human
fetuses, neonates, children, mentally incompetent
persons, and prisoners, who require additional protec-
tion in the conduct of research. Researchers need to
justify their use of subjects with diminished autonomy
in a study, and the need for justification increases as
the subjects’ risk and vulnerability increase. However,
in many situations, the knowledge needed to provide
evidence-based care to these vulnerable populations
can be gained only by studying them.
Legally and Mentally Incompetent Subjects
Neonates and children (minors), the mentally impaired,
and unconscious patients are legally or mentally
incompetent to give informed consent. These indi-
viduals lack the ability to comprehend information
about a study and to make decisions regarding partici-
pation in or withdrawal from the study. Their vulner-
ability ranges from minimal to absolute. The use of
persons with diminished autonomy as research sub-
jects is more acceptable if (1) the research is therapeu-
tic, so that the subjects have the potential to benefit
directly from the experimental process; (2) the
researcher is willing to use both vulnerable and non-
vulnerable individuals as subjects; (3) preclinical and
clinical studies have been conducted and provide data
for assessing potential risks to subjects; and (4) the
risk is minimized and the consent process is strictly
followed to secure the rights of the prospective sub-
jects (U.S. DHHS, 2009).
Neonates
A neonate is defined as a newborn and is identified
as either viable or nonviable on delivery. Viable neo-
nates are able to survive after delivery, if given the
benefit of available medical therapy, and can indepen-
dently maintain a heartbeat and respiration. A nonvi-
able neonate is a newborn who after delivery, although
living, is not able to survive (U.S. DHHS, 2009). Neo-
nates are extremely vulnerable and require extra pro-
tection to determine their involvement in research.
However, research may involve viable neonates, neo-
nates of uncertain viability, and nonviable neonates if
the following five conditions are met:
166 UNIT TWO  The Research Process

1. The study is scientifically appropriate and the pre-
clinical and clinical studies have been conducted
and provided data for assessing the potential risks
to the neonates.
2. The study provides important biomedical knowl-
edge that cannot be obtained by other means and
will not add risk to the neonate.
3. The research has the potential to enhance the prob-
ability of survival of the neonate.
4. Both parents are fully informed about the research
during the consent process.
5. The research team will have no part in determining
the viability of the neonate.
In addition, for the nonviable neonate, the vital
functions of the neonate should not be artificially
maintained because of the research, and the research
should not terminate the heartbeat or respiration of the
neonate (U.S. DHHS, 2009).
Children
The unique vulnerability of children makes the
decision to use them as research subjects particularly
important. To safeguard their interests and protect
them from harm, special ethical and regulatory con-
siderations have been put in place for research involv-
ing children (U.S. DHHS, 2009). However, the laws
defining the minor status of a child are statutory and
vary from state to state. Often a child’s competency to
consent is governed by age, with incompetence being
nonrefutable up to age 7 years (Broome, 1999; Fry et
al., 2011). Thus, a child younger than 7 years is not
believed to be mature enough to assent or consent to
research. Developmentally by age 7, a child is capable
of concrete operations of thought and can give
meaningful assent to participate as a subject in studies
(Thompson, 1987). With advancing age and maturity,
a child should have a stronger role in the consent
process.
To obtain informed consent, federal regulations
require both the assent of the children (when capable)
and the permission of their parents or guardians (U.S.
DHHS, 2009). Assent means a child’s affirmative
agreement to participate in research. Permission to
participate in a study means the agreement of parents
or guardian to the participation of their child or ward
in research (U.S. DHHS, 2009). If a child does not
assent to participate in the study, he or she should not
be included as a subject even if parental permission is
obtained.
Using children as research subjects is also influ-
enced by the therapeutic nature of the research and the
risks versus the benefits. Thompson (1987) developed
a guide for obtaining informed consent that is based
on the child’s level of competence, the therapeutic
nature of the research, and the risks versus the benefits
(Table 9-2). Children who are experiencing a devel-
opmental delay, cognitive deficit, emotional disorder,
or physical illness must be considered individually
(Broome, 1999; Broome & Stieglitz, 1992).
A child 7 years or older with normal cognitive
development can provide assent or dissent to partici-
pation in a study, and the process for obtaining
the assent should be included in the research proposal.
In the assenting process, the child must be given
developmentally appropriate information on the study
purpose, expectations, and benefit-risk ratio (discussed
later). DVDs, written materials, demonstrations,

TABLE 9-2  Guide to Obtaining Informed Consent Based on the Relationship between a Child’s Level
of Competence, the Therapeutic Nature of the Research, and Risk versus Benefit
Nontherapeutic Research Therapeutic Research
MMR-LB MR-LB MR-HB MMR-HB
Child, Incompetent (generally, 0-6 yr)
Parents’ consent Necessary Necessary Sufficient* Sufficient
Child’s assent Optional† Optional† Optional
Child, Relatively Competent (7 yr and older)
Parents’ consent Necessary Necessary Sufficient‡ Recommended
Child’s assent Necessary Necessary Sufficient§ Sufficient

*A parent’s refusal can be superseded by the principle that a parent has no power to forbid the saving of a child’s life.
†
Children making “deliberate objection” would be precluded from participation by most researchers.
‡
In cases not involving the privacy rights of a “mature minor.”
§
In cases involving the privacy rights of a “mature minor.”
HB, high benefit; LB, low benefit; MMR, more than minimal risk; MR, minimal risk.
From Thompson, P. J. (1987). Protection of the rights of children as subjects for research. Journal of Pediatric Nursing, 2(6), 397.

diagrams, role-modeling, and peer discussions are
possible methods for communicating study informa-
tion. The child also needs an opportunity to sign an
assent form and to have a copy of this form. An
example assent form is presented in Box 9-2. During
the study, the researcher must give the child the oppor-
tunity to ask questions and to withdraw from the study
if he or she desires (Broome, 1999). Assent becomes
more complex if the child is bilingual, because the
researchers must determine the most appropriate lan-
guage to use for the consent process for the child and
the parents. Holaday, Gonzales, and Mills (2007) offer
a list of seven questions in their article to assist
researchers in determining the language for commu-
nication during a study.
Rew, Horner, and Fouladi (2010) conducted a study
of school-aged children’s health behaviors to deter-
mine whether they were precursors of adolescents’
health-risk behaviors. The sample included Hispanic
and non-Hispanic children and their parents. The
ethical aspects of the study are described in the fol-
lowing quotation:
“The study took place in a rural setting in central
Texas, a state with a rapidly expanding population of
Hispanics, primarily of Mexican descent. The non­
probability sample was composed of 1,934 children
in Grades 4 (n = 781), 5 (n = 621), and 6 (n = 532)
who were enrolled in three rural school districts in
central Texas and one of their parents….
After study approval was obtained from the uni­
versity’s institutional review board [IRB] and each of
the school administrators, a packet was mailed to
parents of all the children in Grades 4 through 6 in
three rural school districts. The packets included a
cover letter from the child’s school, an explanatory
letter from the researchers, and consent forms. All
materials were written in English and Spanish, with
forward and backward translations by independent
speakers, and were reviewed by bilingual members
of the community for translation clarity and accuracy
before mailing. Informational meetings were held at
the schools after parent-teacher meetings. At those
school meetings, the study was explained to the chil­
dren, questions were answered, and signed permis­
sions were obtained from parents. Data were later
collected during school hours using audio (optional)
computer-assisted self-interviewing (A-CASI) tech­
nology using laptop computers after the children who
agreed to participate provided written assent.…
CHAPTER 9  Ethics in Research 167
The children were oriented to the A-CASI format
and were directed to select either the English- or
Spanish-language version to complete. For children
who had difficulty with reading, audio support was
engaged on the laptop computer and the children
listened with an earpiece as the items were read to
them in their preferred language.… As each child
completed the questionnaire, the research assistants
saved the data record to a secure Web site.” (Rew et
al., 2010, pp. 158, 160)
Rew et al. (2010) provided a detailed description
of the protection of the children and their parents’
rights. The study was described in a language of
choice with an offer to answer questions. The parents
agreed to their children’s participation in the study
through signed permissions. The children gave written
assent to participating in the study. Other ethical
aspects of the study were the IRB approvals from the
university and school administrators and the storage
of study data in a secure location. All of these activi-
ties promoted the ethical conduct of this study accord-
ing to the U.S. DHHS (2009) regulations. The
researchers found that girls have more health-focused
behaviors than boys, health behaviors decreased from
grades 4 to 6, and the school environment was impor-
tant for promoting health behaviors.
Pregnant Women
Pregnant women require additional protection in
research because of their fetuses. Federal regulations
define pregnancy as encompassing the period of time
from implantation until delivery. “A woman is assumed
to be pregnant if she exhibits any of the pertinent
presumptive signs of pregnancy, such as missed
menses, until the results of a pregnancy test are nega-
tive or until delivery” (U.S. DHHS, 2009, 45 CFR
Section 46.202). Research conducted with pregnant
women should have the potential to directly benefit
the woman or the fetus. If your investigation is thought
to provide a direct benefit only to the fetus, you must
obtain the consent of the pregnant woman and father.
In addition, studies with pregnant women should
include no inducements to terminate the pregnancy
(U.S. DHHS, 2009).
Adults with Diminished Capacity
Certain adults have a diminished capacity for, or
are incapable of, giving informed consent because of
mental illness (Beebe & Smith, 2010), cognitive
impairment, or a comatose state (Simpson, 2010).
Persons are said to be incompetent if a qualified clini-
cian judges them to have attributes that designate them
168 UNIT TWO  The Research Process
Box 9-2 Sample Assent Form for
Children Ages 6 To 12 Years:
Pain Interventions for
Children with Cancer
Oral Explanation
I am a nurse who would like to know whether
relaxation, special ways of breathing, and
using your mind to think pleasant things help
children like you to feel less afraid and feel
less hurt when the doctor has to do a bone
marrow aspiration or spinal tap. Today, and
the next five times you and your mom and/or
dad come to the clinic, I would like for you have the ability to understand what they are being
to answer some questions about the things asked to do, to make reasonably free choices, and to
in the clinic that scare you. I would also like
you to tell me about how much pain you (U.S. DHHS, 2009).
felt during the bone marrow or spinal tap. A number of people in intensive care units and
In addition, I would like to videotape (take
pictures of) you and your mom and/or dad
during the tests. The second time you visit the
clinic I would like to meet with you and teach
you special ways to relax, breathe, and use
your mind to imagine pleasant things. You can
use the special imagining and breathing
during your visits to the clinic. I would ask you
and your mom and/or dad to practice the
things I teach you at home between your visits
to the clinic. At any time you could change
your mind and not be in the study anymore.
To Child
1. I want to learn special ways to relax,
breathe, and imagine.
2. I want to answer questions about things
children may be afraid of when they come
to the clinic.
3. I want to tell you how much pain I feel
during the tests I have.
4. I will let you videotape me while the doctor
does the tests (bone marrow and spinal
taps).
If the child says YES, have him/her put an
“X” here: _______________________
If the child says NO, have him/her put an
“X” here: ________________________
Date: ______________________
Child’s signature: ________________________
From Broome, M. E. (1999). Consent (assent) for
research with pediatric patients. Seminars in
Oncology Nursing, 15(2), 101.
as incompetent (U.S. DHHS, 2009). Incompetence
can be temporary (e.g., inebriation), permanent (e.g.,
advanced senile dementia), or subjective or transitory
(e.g., behavior or symptoms of psychosis).
If an individual is judged incompetent and inca-
pable of consent, you must seek approval from the
prospective subject and his or her legally authorized
representative. A legally authorized representative
means an individual or other body authorized under
law to consent on behalf of a prospective subject to
his or her participation in research. However, indi-
viduals can be judged incompetent and can still assent
to participate in certain minimal-risk research if they
communicate their choices clearly and unambiguously
nursing homes are experiencing some level of cogni-
tive impairment. These individuals must be assessed
for their capacity to give consent to participate in
research. The assessment needs to include the follow-
ing elements: understanding of the study information,
developing a belief about the information, reasoning
ability, and understanding of a choice. Simpson (2010)
reviewed the literature and found that the MacArthur
Competency Assessment Tool for Clinical Research
(MacCAT-CR) is one of the strongest instruments
available for assessing an individual’s capacity to give
informed consent. Using this instrument or others dis-
cussed by Simpson (2010), researchers can make a
more sound decision about a subject’s ability to
consent to research or about whether the legal guard-
ian must be contacted for permission.
Some individuals have become permanently
incompetent from the advanced stages of senile
dementia of the Alzheimer type (SDAT), and their
legal guardians must give permission for their partici-
pation in research. Often families or guardians of these
patients are reluctant to give consent for their partici-
pation in research. However, nursing research is
needed to establish evidence-based interventions for
comforting and caring for these individuals. Levine
(1986) identified two approaches that families, guard-
ians, researchers, or IRBs might use when making
decisions on behalf of these incompetent individuals:
(1) best interest standard and (2) substituted judgment
standard. The best interest standard involves doing
what is best for the individual on the basis of balanc-
ing risks and benefits. The substituted judgment
standard is concerned with determining the course of
action that incompetent individuals would take if they
were capable of making a choice (Beattie, 2009).

Jones, Munro, Grap, Kitten, and Edmond (2010)
conducted a quasi-experimental study to determine the
effect of toothbrushing on bacteremia risk in mechani-
cally ventilated adults. These researchers described
their process for obtaining consent from their study
participants in the following study excerpt:
“The study was reviewed and approved by the uni­
versity’s institutional review board. All subjects who
met inclusion criteria were assessed for ability to
provide informed consent through gesturing or
writing. If subjects had medications that impaired cog­
nition or were unable to provide informed consent
due to their illness, the legally authorized representa­
tive provided informed consent.” (Jones et al., 2010,
p. S58)
Jones et al. (2010) developed a process for deter-
mining the cognitive competence of their potential
research participants and obtained appropriate consent
on the basis of their assessments. The competent sub-
jects were given the right to self-determination regard-
ing study participation. The researchers found that the
toothbrushing intervention did not cause transient bac-
teremia in this population of ventilated patients.
Terminally Ill Subjects
When conducting research on terminally ill sub-
jects, you should determine (1) who will benefit from
the research and (2) whether it is ethical to conduct
research on individuals who might not benefit from
the study (U.S. DHHS, 2009). Participating in research
could have greater risks and minimal or no benefits
for these subjects. In addition, the dying subject’s con-
dition could affect the study results and lead you to
misinterpret the results. However, Hinds, Burghen,
and Pritchard (2007) stressed the importance of con-
ducting end-of-life studies in pediatric oncology to
generate evidence that will improve the care for ter-
minally ill children and adolescents.
Some terminally ill individuals are willing subjects
because they believe that participating in research is a
way to contribute to society before they die. Others
want to take part in research because they believe that
the experimental process will benefit them. For
example, individuals with AIDS might want to partici-
pate in AIDS research to gain access to experimental
drugs and hospitalized care. Researchers studying
populations with serious or terminal illnesses are
faced with ethical dilemmas as they consider the rights
of the subjects and their responsibilities in conducting
quality research (Fry et al., 2011; U.S. DHHS, 2009).
CHAPTER 9  Ethics in Research 169
Subjects Confined to Institutions
Hospitalized patients have diminished autonomy
because they are ill and are confined in settings that
are controlled by healthcare personnel (Levine, 1986).
Some hospitalized patients feel obliged to be research
subjects because they want to assist a particular prac-
titioner (nurse or physician) with his or her research.
Others feel coerced to participate because they fear
that their care will be adversely affected if they refuse.
Some of these hospitalized patients are survivors of
trauma (such as auto accidents, gunshot wounds, or
physical and sexual abuse) who are very vulnerable
and often have decreased decision-making capacities
(McClain, Laughon, Steeves, & Parker, 2007). When
conducting research with these types of patients, you
must pay careful attention to the informed consent
process and make every effort to protect these subjects
from feelings of coercion and harm (U.S. DHHS,
2009).
In the past, prisoners have experienced diminished
autonomy in research projects because of their con-
finement. They might feel coerced to participate in
research because they fear harm if they refuse or
because they desire the benefits of early release,
special treatment, or monetary gain. Prisoners have
been used for drug studies in which there were no
health-related benefits and there was possible harm for
the prisoners. Current regulations regarding research
involving prisoners require that “the risks involved in
the research are commensurate with risks that would
be accepted by nonprisoner volunteers and procedures
for the selection of subjects within the prison are fair
to all prisoners and immune from arbitrary interven-
tion by prison authorities or prisoners” (U.S. DHHS,
2009, Section 46.305).
Protecting the rights of subjects with diminished
autonomy in research is regulated internationally by
the Council for International Organizations of Medical
Sciences (CIOMS). CIOMS (2010) developed inter-
national ethical guidelines for biomedical research
involving human subjects, and the guidelines require
protection of vulnerable individuals, groups, commu-
nities, and populations during research. Researchers
must evaluate each prospective subject’s capacity for
self-determination and must protect subjects with
diminished autonomy during the research process
(ANA, 2001, APA, 2010; U.S. DHHS, 2009).
Right to Privacy
Privacy is an individual’s right to determine the time,
extent, and general circumstances under which per-
sonal information will be shared with or withheld from
others. This information consists of one’s attitudes,
170 UNIT TWO  The Research Process
beliefs, behaviors, opinions, and records. The Privacy
Act of 1974 provided the initial protection of an indi-
vidual’s privacy. Because of this act, data collection
methods were to be scrutinized to protect subjects’
privacy, and data cannot be gathered from subjects
without their knowledge. Individuals also have the
right to access their records and to prevent access by
others (U.S. DHHS, 2009). The intent of this act was
to prevent the invasion of privacy that occurs when
private information is shared without an individual’s
knowledge or against his or her will. Invading an
individual’s privacy might cause loss of dignity,
friendships, or employment or create feelings of
anxiety, guilt, embarrassment, or shame.
The HIPAA Privacy Rule expanded the protection
of an individual’s privacy, specifically his or her pro-
tected individually identifiable health information, and
described the ways in which covered entities can use
or disclose this information. “Individually identifi-
able health information (IIHI) is information that is
a subset of health information, including demographic
information collected from an individual, and: (1) is
created or received by healthcare provider, health
plan, or healthcare clearinghouse; and (2) [is] related
to past, present, or future physical or mental health or
condition of an individual, the provision of health care
to an individual, or the past, present, or future payment
for the provision of health care to an individual, and
that identifies the individual; or with respect to which
there is a reasonable basis to believe that the informa-
tion can be used to identify the individual” (U.S.
DHHS, 2003, 45 CFR, Section 160.103).
According to the HIPAA Privacy Rule, the IIHI is
protected health information (PHI) that is transmitted
by electronic media, maintained in electronic media,
or transmitted or maintained in any other form or
medium. Thus, the HIPAA privacy regulations affect
nursing research in the following ways: (1) accessing
data from a covered entity, such as reviewing a
patient’s medical record in clinics or hospitals; (2)
developing health information, such as the data devel-
oped when an intervention is implemented in a study
to improve a subject’s health; and (3) disclosing data
from a study to a colleague in another institution, such
as sharing data from a study to facilitate development
of an instrument or scale (Olsen, 2003).
The U.S. DHHS developed the following guide-
lines to help researchers, healthcare organizations, and
healthcare providers determine when they can use and
disclose IIHI. IIHI can be used or disclosed to a
researcher in the following situations:
• The protected health information has been
“de-identified” under the HIPAA Privacy Rule.
(De-identifying PHI is defined in the following
section.)
• The data are part of a limited data set, and a
data use agreement with the researcher(s) is in
place.
• The individual who is a potential subject for a study
authorizes the researcher to use and disclose his or
her PHI.
• A waiver or alteration of the authorization require-
ment is obtained from an IRB or a privacy board
(U.S. DHHS, 2007b) (see http://privacyruleand
research.nih.gov/pr_08.asp/).
The first two items are discussed in this section of
the chapter. The authorization process is discussed in
the section on obtaining informed consent, and the
waiver or alteration of authorization requirement is
covered in the section on institutional review of
research.
De-identifying Protected Health Information under
the Privacy Rule
Covered entities, such as healthcare providers and
agencies, can allow researchers access to health infor-
mation if the information has been de-identified.
De-identifying health data involves removing the 18
elements that could be used to identify an individual
or his or her relatives, employer, or household
members. The 18 identifying elements are listed in
Box 9-3.
An individual’s health information can also be
de-identified through the use of statistical methods.
However, the covered entity and you as the researcher
must ensure that the individual subject cannot be iden-
tified or that there is a very small risk that the subject
could be identified from the information used. The
statistical method used for de-identification of the
health data must be documented, and you must certify
that the 18 elements for identification have been
removed or revised to ensure the individual is not
identified. You must retain this certification informa-
tion for 6 years.
Limited Data Set and Data Use Agreement
Covered entities (healthcare provider, health plan, and
healthcare clearinghouse) may use and disclose a
limited data set to a researcher for a study without an
individual subject’s authorization or an IRB waiver.
However, a limited data set is considered PHI, and the
covered entity and the researcher must have a data use
agreement. The data use agreement limits how the
data set may be used and how it will be protected. The
HIPAA Privacy Rule requires the data use agreement
to do the following (U.S. DHHS, 2003):

1. Specifies the permitted uses and disclosures of the
limited data set.
2. Identifies the researcher who is permitted to use or
receive the limited data set.
3. Stipulates that the recipient (researcher) will:
a. Not use or disclose the information other than
permitted by the agreement.
b. Use appropriate safeguards to prevent the use or
disclosure of the information, except as provided
for in the agreement.
c. Hold any other person (co-researchers, stat­
isticians, or data collectors) to the standards,
restrictions, and conditions stated in the data
use agreement with respect to the health
information.
d. Not identify the information or contact the
individuals whose data are in the limited
data set.
Riegel et al. (2011) conducted a secondary analysis
of data from the Heart Failure (HF) Quality of Life
Registry database. The purpose of the study was to
establish whether confidence and activity status deter-
mined the HF patients’ self-care performance. The
researchers found three levels of self-care perfor-
mance: (1) novice in self-care with limited confidence
and few activity restrictions; (2) inconsistent in self-
care abilities; and (3) expert with confidence in self-
care abilities. The researchers ensured the PHI of the
individuals in the database was ethically managed, as
described in the following excerpt:
“By prior consensus of investigators in the HF Quality
of Life Registry, study samples are enrolled using com­
parable inclusion and exclusion criteria, as well as the
same variables and measures whenever possible. All
data are stored at one site, where one of the inves­
tigators has volunteered to integrate newly acquired
data. The only identifiers in the data set are site (e.g.,
Cleveland Clinic) and the specific study name, as
more than one study is common at each site. No
protected health information [PHI] is included in the
database. All requests to use the full database are
viewed and approved by the lead investigators. For
this analysis, five samples enrolled at three different
sites in the United States between 2003 and 2008
were used.” (Riegel et al., 2011, p. 133)
Right to Autonomy and Confidentiality
On the basis of the right to privacy, the research
subject has the right to anonymity and the right
to assume that the data collected will be
CHAPTER 9  Ethics in Research 171
Box 9-3 18 Elements that Could Be
Used to Identify an Individual
to Relatives, Employer, or
Household Members
1. Names.
2. All geographical subdivisions smaller than
a state, including street address, city,
county, precinct, zip code, and their
equivalent geographical codes, depending
on size of population according to the
current publicly available data from the
Bureau of the Census, as follows:
a. If the geographical unit formed by
combining all zip codes with the same
three initial digits contains more than
20000 people, the initial three digits of
the zip code may be retained.
b. For all such geographical units
containing 20000 or fewer people, the
initial three digits of the zip code are
changed to 000.
3. All elements of dates (except year) for
dates directly related to an individual,
including birth date, admission date,
discharge date, date of death; and all ages
over 89 and all elements of dates (including
year) indicative of such age, except that
such ages and elements may be aggregated
into a single category of age 90 or older.
4. Telephone numbers.
5. Facsimile numbers.
6. Electronic mail (email) addresses.
7. Social security numbers.
8. Medical record numbers.
9. Health plan beneficiary numbers.
10. Account numbers.
11. Certificate/license numbers.
12. Vehicle identifiers and serial numbers,
including license plate numbers.
13. Device identifiers and serial numbers.
14. Web universal resource locators (URLs).
15. Internet protocol (IP) address numbers.
16. Biometric identifiers, including fingerprints
and voiceprints.
17. Full-face photographic images and any
comparable images.
18. Any other unique identifying number,
characteristic, or code, unless otherwise
permitted by the Privacy Rule for
Re-identification (see http://privacyruleand
research.nih.gov/pr_08.asp/) (U.S. DHHS,
2007b).
172 UNIT TWO  The Research Process
kept confidential. Anonymity exists if the subject’s
identity cannot be linked, even by the researcher, with
his or her individual responses (APA, 2010; Fry et al.,
2011). For studies that use de-identified health infor-
mation or data from a limited data set, the subjects are
anonymous to the researchers. The researchers are
unable to contact these subjects for additional infor-
mation without special approval, as described in the
study by Riegel et al. (2011).
In most studies, researchers desire to know the
identity of their subjects and promise that their identity
will be kept confidential. Confidentiality is the
researcher’s management of private information
shared by a subject that must not be shared with others
without the authorization of the subject. Confidential-
ity is grounded in the following premises: (1) indi-
viduals can share personal information to the extent
they wish and are entitled to have secrets; (2) one can
choose with whom to share personal information; (3)
people who accept information in confidence have an
obligation to maintain confidentiality; and (4) profes-
sionals, such as researchers, have a duty to maintain
confidentiality that goes beyond ordinary loyalty
(Levine, 1986; U.S. DHHS, 2009).
Breach of Confidentiality
A breach of confidentiality can occur when a
researcher, by accident or direct action, allows an
unauthorized person to gain access to the study raw
data. Confidentiality can also be breached in the
reporting or publication of a study when a subject’s
identity is accidentally revealed, violating the sub-
ject’s right to anonymity (Munhall, 2012a). Breaches
of confidentiality can harm subjects psychologically
and socially as well as destroy the trust they had in
the researchers. Breaches of confidentiality can be
especially harmful to a research participant if they
involve (1) religious preferences; (2) sexual prac-
tices; (3) employment; (4) racial prejudices; (5) drug
use; (6) child abuse; or (7) personal attributes, such
as intelligence, honesty, and courage. For example, a
university researcher conducted a study of nurses’
stressful life events and work-related burnout in an
acute care hospital. One of the two male participants
in the study was a nurse who is being treated for an
anxiety disorder. Reporting that one of the male
nurses in the study was being treated for an anxiety
disorder would violate his confidentiality and poten-
tially cause harm. Nurse administrators might be less
likely to promote a nurse who is receiving mental
health treatment.
Some nurse researchers have encountered health-
care professionals who believe that they should have
access to information about the patients in the hospital
and will request to see the data the researchers have
collected. Sometimes, family members or close friends
would like to see the data collected on specific sub-
jects. Sharing research data in these circumstances is
a breach of confidentiality. When requesting permis-
sion to conduct a study, you should tell healthcare
professionals, family members, and others in the
setting that you will not share the raw data. However,
you may elect to share the research report, including
a summary of the data and findings from the study,
with healthcare providers, family members, and other
interested parties.
Maintaining Confidentiality
Researchers have a responsibility to protect the ano-
nymity of subjects and to maintain the confidentiality
of data collected during a study. You can protect ano-
nymity by giving each subject a code number. Keep a
master list of the subjects’ names and their code
numbers in a locked place; for example, subject Mary
Jones might be assigned the code number “001.” All
of the instruments and forms that Mary completes and
the data you collect about her during the study will be
identified with the “001” code number, not her name.
The master list of subjects’ names and code numbers
is best kept separate from the data collected to protect
subjects’ anonymity. You should not staple signed
consent forms and authorization documents to instru-
ments or other data collection tools, as this would
make it easy for unauthorized persons to readily iden-
tify the subjects and their responses. Consent forms
are often stored with the master list of subjects’ names
and code numbers. When entering the data collected
into the computer, use the code numbers for identifica-
tion and ensure that the data are stored in a secure
place on a flash drive, in the computer, or on a website.
In the study by Rew et al. (2010) that was introduced
earlier in this chapter, the school-aged children par-
ticipating in the study of their health behaviors com-
pleted a questionnaire on the computer and their data
were saved by the research assistants to a secure
website. These actions ensured that each subject’s data
were kept confidential during and after completion of
the study.
Another way to protect your subjects’ anonymity is
to have subjects or study participants generate their
own identification codes (Damrosch, 1986). With this
approach, each subject generates an individual code
from personal information, such as the first letter of a
mother’s name, the first letter of a father’s name, the
number of brothers, the number of sisters, and middle
initial. Thus, the code would be composed of three

letters and two numbers, such as “BD21M.” This code
would be used on each form that the subject com-
pletes. Using a subject-generated code is especially
helpful when the researcher needs to link the subject’s
data over time. Because of the specific components of
the ID number, the subject does not have to remember
the code number from one data collection point until
the next. Even you as the researcher would not know
the subject’s identity, only the subject’s code. If a
qualitative study is being conducted in which exten-
sive, often sensitive data are collected on just a few
study participants, researchers might want to use this
type of coding system. Qualitative researchers also use
the approach of allowing participants to provide
pseudonyms by which they want to be known during
the study.
Maintaining confidentiality of participants’ data in
qualitative studies often requires more effort than in
quantitative research. The nature of qualitative
research requires that the “investigator must be close
enough to understand the depth of the question under
study, and must present enough direct quotes and
detailed description to answer the question” (Ramos,
1989, p. 60). The small number of participants used
in a qualitative study and the depth of detail gathered
on each participant requires planning to ensure confi-
dentiality. Ford and Reutter (1990) have recommended
that to maintain confidentiality, the researcher should
(1) use pseudonyms instead of the participants’ names
and (2) distort certain details in the participants’ stories
while leaving the contents unchanged. Researchers
must respect participants’ privacy as they decide how
much detail and editing of private information are
necessary to publish a study (Munhall, 2012a; Orb,
Eisenhauer, & Wynaden, 2001).
Researchers should also take precautions during
data collection and analysis to maintain confidentiality
in qualitative studies. The interviews conducted with
participants are frequently recorded and later tran-
scribed, so the participants’ names should not be men-
tioned during the recording. They have the right to
know whether anyone other than you will be transcrib-
ing information from the interviews. In addition, par-
ticipants should be informed on an ongoing basis that
they have the right to withhold information. By allow-
ing other researchers to critically appraise the rigor
and credibility of a qualitative study, an audit trail is
produced. Allowing others to examine the data to
confirm the study findings may create a dilemma
regarding the confidentiality of participants’ data, so
you must inform them if other researchers will be
examining their data to ensure the credibility of the
study findings (Munhall, 2012a; Orb et al., 2001).
CHAPTER 9  Ethics in Research 173
Confidentiality of subjects’ information can also be
ensured during the data analysis process in quantita-
tive research. The data collected should undergo group
analysis so that an individual cannot be identified by
his or her responses. If the subjects are divided into
groups for data analysis and there is only one subject
in a group, combine that subject’s data with that of
another group or delete the data. In writing the research
report, you should describe the findings in such a way
that an individual or a group of individuals cannot be
identified from their responses.
Right to Fair Treatment
The right to fair treatment is based on the ethical
principle of justice. This principle holds that each
person should be treated fairly and should receive
what he or she is due or owed. In research, the selec-
tion of subjects and their treatment during the course
of a study should be fair.
Fair Selection of Subjects
In the past, injustices in subject selection have resulted
from social, cultural, racial, and sexual biases in
society. For many years, research was conducted on
categories of individuals who were thought to be espe-
cially suitable as research subjects, such as the poor,
charity patients, prisoners, slaves, peasants, dying
persons, and others who were considered undesirable
(Reynolds, 1979). Researchers often treated these sub-
jects carelessly and had little regard for the harm and
discomfort they experienced. The Nazi medical exper-
iments, Tuskegee syphilis study, and Willowbrook
study all exemplify unfair subject selection and
treatment.
The selection of a population and the specific sub-
jects to study should be fair, and the risks and benefits
of a study should be fairly distributed on the basis of
the subject’s efforts, needs, and rights. Subjects should
be selected for reasons directly related to the problem
being studied and not for “their easy availability, their
compromised position, or their manipulability”
(National Commission for the Protection of Human
Subjects of Biomedical and Behavioral Research,
1978, p. 10).
Another concern with subject selection is that some
researchers select certain people as subjects because
they like them and want them to receive the specific
benefits of a study. Other researchers have been
swayed by power or money to make certain individu-
als subjects so that they can receive potentially benefi-
cial treatments. Random selection of subjects can
eliminate some of the researcher bias that might influ-
ence subject selection.
174 UNIT TWO  The Research Process
A current concern in the conduct of research is
finding an adequate number of appropriate subjects to
take part in certain studies. As a solution to this
problem in the past, some biomedical researchers have
offered physicians finder’s fees for identifying research
subjects. For example, investigators studying patients
with lung cancer would give a physician a fee for
every patient with lung cancer the physician referred
to them. However, the HIPAA Privacy Rule requires
that individuals give their authorization before PHI
can be shared with others. Thus, healthcare providers
cannot recommend individuals for studies without
their permission. Researchers can obtain a partial
waiver from the IRB or privacy board so that they can
obtain PHI necessary to recruit potential subjects
(U.S. DHHS, 2003). This makes it more difficult for
researchers to find subjects for their studies; however,
researchers are encouraged to work closely with their
IRBs and the personnel in the settings of their studies
to enlarge their sample sizes.
Fair Treatment of Subjects
Researchers and subjects should have a specific agree-
ment about what a subject’s participation involves and
what the role of the researcher will be (APA, 2010).
While conducting a study, you should treat the sub-
jects fairly and respect that agreement. If the data
collection requires appointments with the subjects, be
on time for each appointment and terminate the data
collection process at the agreed-on time. You should
not change the activities or procedures that a subject
is to perform unless you obtain the subject’s consent.
The benefits promised the subjects should be pro-
vided. For example, if you promise a subject a copy
of the study findings, you should deliver on your
promise when the study is completed. In addition,
subjects who participate in studies should receive
equal benefits, regardless of age, race, and socioeco-
nomic status. When possible, the sample should be
representative of the study population and should
include subjects of various ages, ethnic backgrounds,
and socioeconomic status. Treating subjects fairly
often facilitates the data collection process and
decreases the chances of subjects’ withdrawal from a
study (Fry et al., 2011; Orb et al., 2001).
Right to Protection from Discomfort
and Harm
The right to protection from discomfort and harm
is based on the ethical principle of beneficence, which
holds that one should do good and, above all, do no
harm. Therefore, researchers should conduct their
studies to protect subjects from discomfort and harm
and try to bring about the greatest possible balance of
benefits in comparison with harm. Discomfort and
harm can be physiological, emotional, social, and eco-
nomic in nature. Reynolds (1979) identified the fol-
lowing five categories of studies, which are based on
levels of discomfort and harm: (1) no anticipated
effects; (2) temporary discomfort; (3) unusual levels
of temporary discomfort; (4) risk of permanent
damage; and (5) certainty of permanent damage. Each
level is defined in the following discussion.
No Anticipated Effects
In some studies, the subjects expect neither positive
nor negative effects. For example, studies that involve
reviewing patients’ records, students’ files, pathology
reports, or other documents have no anticipated effects
on the subjects. In these types of studies, the researcher
does not interact directly with the research subjects.
Even in these situations, however, there is a potential
risk of invading a subject’s privacy. The HIPAA
Privacy Rule requires that the agency providing the
health information de-identify the 18 essential ele-
ments (see Box 9-3), which could be used to identify
an individual, to promote subjects’ privacy during
a study.
Temporary Discomfort
Studies that cause temporary discomfort are described
as minimal-risk studies, in which the discomfort
encountered is similar to what the subject would expe-
rience in his or her daily life and ceases with the ter-
mination of the study. Many nursing studies require
the subjects to complete questionnaires or participate
in interviews, which usually involve minimal risk. The
physical discomforts might be fatigue, headache, or
muscle tension. The emotional and social risks might
entail the anxiety or embarrassment associated with
responding to certain questions. The economic risks
might consist of the time spent participating in the
study or travel costs to the study site. Participation
in many nursing studies is considered a mere incon-
venience for the subject, with no foreseeable risks
of harm.
Most clinical nursing studies examining the impact
of a treatment involve minimal risk. For example, your
study might involve examining the effects of exercise
on the blood glucose levels of type 2 diabetics. During
the study, you ask the subjects to test their blood
glucose level one extra time per day. There is discom-
fort when the blood is drawn and a risk of physical
changes that might occur with exercise. The subjects
might also experience anxiety and fear in association
with the additional blood testing, and the testing is an

added expense. The diabetic subjects in this study
would experience similar discomforts in their daily
lives, and the discomforts would cease with the termi-
nation of the study.
Unusual Levels of Temporary Discomfort
In studies that involve unusual levels of temporary
discomfort, the subjects commonly experience dis-
comfort both during the study and after its termina-
tion. For example, subjects might experience a deep
vein thrombosis (DVT), prolonged muscle weakness,
joint pain, and dizziness after participating in a study
that required them to be confined to bed for 7 days to
determine the effects of immobility. Studies that
require subjects to experience failure, extreme fear, or
threats to their identity or to act in unnatural ways
involve unusual levels of temporary discomfort. In
some qualitative studies, participants are asked ques-
tions that reopen old emotional wounds or involve
reliving traumatic events (Ford & Reutter, 1990;
Munhall, 2012a). For example, asking participants to
describe their rape experience could precipitate feel-
ings of extreme fear, anger, and sadness. In these types
of studies, you should be vigilant about assessing the
participants’ discomfort and refer them for appropriate
professional intervention as necessary.
Risk of Permanent Damage
In some studies, subjects have the potential to suffer
permanent damage; this potential is more common in
biomedical research than in nursing research. For
example, medical studies of new drugs and surgical
procedures have the potential to cause subjects perma-
nent physical damage. However, nurses have investi-
gated topics that have the potential to damage subjects
permanently, both emotionally and socially. Studies
examining sensitive information, such as sexual behav-
ior, child abuse, or drug use, can be risky for subjects.
These types of studies have the potential to cause per-
manent damage to a subject’s personality or reputation.
There are also potential economic risks, such as
reduced job performance or loss of employment.
Certainty of Permanent Damage
In some research, such as the Nazi medical experi-
ments and the Tuskegee syphilis study, the subjects
experience permanent damage. Conducting research
that will permanently damage subjects is highly ques-
tionable, regardless of the benefits gained. Frequently,
the benefits are for other people but not for the sub-
jects. Studies causing permanent damage to subjects
violate the fifth principle of the Nuremberg Code
(1949) (see Box 9-1).
CHAPTER 9  Ethics in Research 175
Balancing Benefits and Risks for
a Study
Researchers and reviewers of research must examine
the balance of benefits and risks in a study. To deter-
mine this balance or benefit-risk ratio, you must (1)
predict the outcome of your study; (2) assess the actual
and potential benefits and risks on the basis of this
outcome; and then (3) maximize the benefits and mini-
mize the risks (see Figure 9-1). The outcome of a
study is predicted on the basis of previous research,
clinical experience, and theory.
Assessment of Benefits
The probability and magnitude of a study’s potential
benefits must be assessed. A research benefit is
defined as something of health-related, psychosocial,
or other value to a subject, or something that will
contribute to the acquisition of knowledge for
evidence-based practice. Money and other compensa-
tions for participation in research are not benefits but,
rather, are remuneration for research-related inconve-
niences (U.S. DHHS, 2009). In most proposals, the
Developing a Study Benefit-Risk Ratio
Predict the outcomes of the study
Assess Assess
benefits risks
Benefit-risk ratio
Maximize benefits
and minimize risks
Benefits are greater Risks
than or equal outweigh
to risks benefits
Approve Reject
study study
Figure 9-1  Balancing benefits and risks for a study.
176 UNIT TWO  The Research Process
research benefits are described for the individual sub-
jects, subjects’ families, and society.
The type of research conducted, whether therapeu-
tic or nontherapeutic, affects the potential benefits for
the subjects. In therapeutic nursing research, the indi-
vidual subject has the potential to benefit from the
procedures, such as skin care, range of motion, touch,
and other nursing interventions, that are implemented
in the study. The benefits might include improvement
in the subject’s physical condition, which could facili-
tate emotional and social benefits. In addition, the
knowledge generated from the research might expand
the subjects’ and their families’ understanding of
health. The conduct of nontherapeutic nursing research
does not benefit the subject directly but is important
to generate and refine nursing knowledge for practice.
By participating in research, subjects have the poten-
tial to increase their understanding of the research
process and an opportunity to know the findings from
a particular study (Fry et al., 2011).
Assessment of Risks
You must assess the type, severity, and number of risks
that subjects might experience by participating in your
study. The risks involved depend on the purpose of the
study and the procedures used to conduct it. Research
risks can be physical, emotional, social, and economic
in nature and can range from no risk or mere incon-
venience to the risk of permanent damage (Reynolds,
1979). Studies can have actual (known) risks and
potential risks for subjects. In a study of the effects of
prolonged bed rest, for example, an actual risk would
be muscle weakness and the potential risk would be a
DVT. Some studies have actual or potential risks for
the subjects’ families and society. You must determine
the likelihood of the risks and take precautions to
protect the rights of subjects when implementing your
study.
Benefit-Risk Ratio
The benefit-risk ratio is determined on the basis of the
maximized benefits and the minimized risks. The
researcher attempts to maximize the benefits and mini-
mize the risks by making changes in the study purpose
or procedures or both. If the risks entailed by your
study cannot be eliminated or further minimized, you
need to justify their existence. If the risks outweigh
the benefits, you probably need to revise the study or
develop a new one. If the benefits equal or outweigh
the risks, you can usually justify conducting the study,
and an IRB will probably approve it (see Figure 9-1).
Say, for example, that you want to balance the
benefits and risks of a study that would examine the
effect of an exercise and diet program on the partici-
pants’ serum lipid values (serum cholesterol, low-
density lipoprotein [LDL], and high-density lipoprotein
[HDL]) and cardiovascular (CV) risk level. The ben-
efits to the participants are instruction about exercise
and diet and information about their serum lipid values
and CV risk level at the start of the program and 1
year later. The potential benefits are improved serum
lipid values, lowered CV risk level, and better exercise
and dietary habits. The risks consist of the discomfort
of having blood specimens drawn twice for serum
lipid measurements and the time spent participating in
the study (Bruce & Grove, 1994). These discomforts
are temporary, are no more than what the subject
would experience in his or her daily life, and would
cease with the termination of the study. The subjects’
time participating in the study can be minimized
through organization and precise scheduling of
research activities. When you examine the ratio of
benefits to risks, you find that (1) the benefits are
greater in number and importance than the risks and
(2) the risks are temporary and can be minimized.
Thus, you could justify conducting this study, and it
would probably receive approval from the IRB.
The obligation to balance the benefits and risks
of studies is the responsibility of the researcher,
health professionals, and society. The researcher must
balance the benefits and risks of a particular study and
protect the subjects from harm during it. Health pro-
fessionals participating on IRBs must evaluate the
benefit-risk ratio of studies to ensure the conduct of
ethical research in their agencies. Society must be
concerned with the benefits and risks of the entire
enterprise of research and with the protection of all
human research subjects from harm.
Obtaining Informed Consent
Obtaining informed consent from human subjects is
essential for the conduct of ethical research in the
United States (U.S. FDA, 2010a; U.S. DHHS, 2009)
and internationally (CIOMS, 2010). Informing is the
transmission of essential ideas and content from the
investigator to the prospective subject. Consent is the
prospective subject’s agreement to participate in a
study as a subject, which the subject reaches after
assimilating essential information. The phenomenon
of informed consent was formally defined in the first
principle of the Nuremberg Code as follows: “The
voluntary consent of the human subject is absolutely
essential.… This means that the person involved
should have legal capacity to give consent; should be
so situated as to be able to exercise free power of

choice, without the intervention of any element of
force, fraud, deceit, duress, over-reaching or other
ulterior form of constraint or coercion; and should
have sufficient knowledge and comprehension of the
elements of the subject matter involved as to enable
him to make an understanding and enlightened deci-
sion” (Levine, 1986, p. 425). Prospective subjects, to
the degree that they are capable, should have the
opportunity to choose whether or not to participate in
research. With careful accommodations, the subjects
may include persons with cognitive impairment
(Simpson, 2010), psychiatric diagnosis (Beebe &
Smith, 2010), or dementia (Beattie, 2009).
This definition of informed consent provided a
basis for the discussion of consent in all subsequent
codes and regulations and has general acceptance in
the research community. As the definition indicates,
informed consent consists of four elements: (1) disclo-
sure of essential information; (2) comprehension;
(3) competence; and (4) voluntarism. This section
describes the elements of informed consent and the
methods of documenting consent.
Information Essential for Consent
Informed consent requires the researcher to disclose
specific information to each prospective subject. The
following information has been identified as essential
content for informed consent in research by federal
regulations (U.S. FDA, 2010a; U.S. DHHS (2009).
Introduction of Research Activities
The introduction of the research must indicate that a
study is being conducted and provide key information
about the study. Each prospective subject is provided
a statement that he or she is being asked to participate
in research and a description of the purpose and the
expected duration of participation in the study. In
clinical nursing research, the patient, serving as a
subject, must know which nursing activities are
research activities and which are routine nursing inter-
ventions. If at any point the prospective subject dis-
agrees with the researchers’ goals or the intent of the
study, he or she can decline participation or withdraw
from the study.
Prospective subjects also need to receive a com-
plete description of the procedures to be followed and
identification of any procedures in the study that are
experimental (U.S. FDA, 2010a; U.S. DHHS, 2009).
Thus, researchers need to describe the research vari-
ables and the procedures or mechanisms that will be
used to observe, examine, manipulate, or measure
these variables. In addition, they must inform pro-
spective subjects about when the study procedures
CHAPTER 9  Ethics in Research 177
will be implemented, how many times, and in what
setting.
Research participants also need to know the funding
source(s) of a study and whether the study is spon-
sored by specific individuals, organizations, or com-
panies. For example, researchers studying the effects
of a specific drug must identify any sponsorship by a
pharmaceutical company. If the study is being con-
ducted as part of an academic requirement, researchers
should also share that information (Fry et al., 2011).
Description of Risks and Discomforts
Prospective subjects need to be informed about any
foreseeable risks or discomforts (physical, emotional,
social, or economic) that might result from the study
(U.S. FDA, 2010b; U.S. DHHS, 2009). They also
need to know how the risks of the study were mini-
mized and the benefits were maximized. If the study
involves greater than minimal risk, it is a good idea to
encourage the prospective subjects to consult another
person regarding their participation. A trusted advisor,
such as a friend, family member, or another nurse,
could serve as a consultant.
Description of Benefits
You should also describe any benefits to the subject or
to others that may be expected from the research. The
study might benefit the current subjects or might gen-
erate knowledge that will provide evidence-based care
to patients and families in the future (U.S. FDA,
2010a; U.S. DHHS, 2009).
Disclosure of Alternatives
Study participants must receive a disclosure of alter-
natives related to their participating in a study. They
must be informed about appropriate, alternative pro-
cedures or courses of treatment, if any, that might be
advantageous to them (U.S. DHHS, 2009). For
example, nurse researchers examining the effect of a
distraction intervention on the chronic pain of patients
with osteoarthritis would need to make potential
subjects aware of the other alternatives for pain
management.
Assurance of Anonymity and Confidentiality
Prospective subjects must be assured that the confi-
dentiality of their records and PHI will be maintained
during and following their study participation (U.S.
FDA, 2010a; U.S. DHHS, 2003, 2009). Thus, subjects
need to know that their responses and the information
obtained from their records during a study will be kept
confidential and their identities will remain anony-
mous in presentations, reports, and publications of the
178 UNIT TWO  The Research Process
study. Depending on the study design, some partici-
pants’ identities will be anonymous to the researchers
to decrease the potential for bias.
Compensation for Participation in Research
For research involving more than minimal risks, pro-
spective subjects must be given an explanation as to
whether any compensation or medical treatments or
both are available if injury occurs. If medical treat-
ments are available, you need to describe the type and
extent of the treatments. Female prospective subjects
need to know whether the study treatment or proce-
dure may involve potential risks to them or their
fetuses if they are or may become pregnant during the
study (U.S. FDA, 2010a; U.S. DHHS, 2009). Potential
subjects also need to know whether they will receive
a small financial payment ($10 to $30) to compensate
them for their time and effort related to participating
in the study.
Offer to Answer Questions
You need to offer to answer any questions that the
prospective subjects may have during the consent
process. Study participants also need an explanation
of whom to contact for answers to questions about the
research during the conduct of the study and of whom
to contact in the event of a research-related problem
or injury as well as how to do so (U.S. FDA, 2010a;
U.S. DHHS, 2009).
Noncoercive Disclaimer
A noncoercive disclaimer is a statement that partici-
pation is voluntary and refusal to participate will
involve no penalty or loss of benefits to which the
subject is entitled (U.S. FDA, 2010a; U.S. DHHS,
2009). This statement can facilitate a relationship
between you and your prospective subjects, especially
if the relationship has a potential for coercion.
Option to Withdraw
Subjects may discontinue participation in or may
withdraw from a study at any time without penalty or
loss of benefits. However, researchers do have the
right to ask subjects whether they think that they will
be able to complete the study, to decrease the number
of subjects withdrawing early. There may be circum-
stances under which the subject’s participation may
be terminated by the researcher without regard to the
subject’s consent (U.S. DHHS, 2009). For example,
if a particular treatment becomes potentially danger-
ous to a subject, you have an obligation to discon-
tinue the subject’s participation in the study. Thus,
describe for prospective subjects the circumstances
under which they might be withdrawn from the study,
and make a general statement about the circum-
stances that could lead to the termination of the entire
project.
Consent to Incomplete Disclosure
In some studies, subjects are not completely informed
of the study purpose because that knowledge would
alter the subjects’ actions. However, prospective sub-
jects must know that certain information is being with-
held deliberately. You must ensure that there are no
undisclosed risks to the subjects that are more than
minimal and that their questions are truthfully
answered regarding the study. Subjects who are
exposed to nondisclosure of information must know
when and how they will be debriefed about the study.
Subjects are debriefed by informing them of the
actual purpose of the study and the results that were
obtained. At this point, subjects have the option to
have their data withdrawn from the study. If the sub-
jects experience adverse effects related to the study,
you need to make every attempt to reconcile the
effects (APA, 2010; U.S. DHHS, 2009).
Comprehension of Consent Information
Informed consent implies not only the imparting of
information by the researcher but also the comprehen-
sion of that information by the subject. Studies exam-
ining subjects’ levels of comprehension of consent
information have found their comprehension to be
limited (Erlen, 2010). The potential subjects’ compre-
hension of the consent depended on the complexity of
the study, the amount of information communicated,
and the process for communicating the information.
The amount of information to be taught depends on
the subjects’ knowledge of research and the specific
research project proposed. Federal regulations require
that the information given to subjects or their repre-
sentatives must be in a language they can understand
(U.S. FDA, 2010a; U.S. DHHS, 2009). Thus, the
consent information must be written and verbalized in
lay terminology, not professional jargon, and must be
presented without the use of biased terms that might
coerce a subject into participating in a study. Meade
(1999) identified the following tips for promoting the
comprehension of a consent document by potential
research subjects:
• Introduce the purpose of the study early in the
consent form.
• Outline the study treatment with specificity and
conciseness.
• Convey the elements of informed consent in an
organized fashion.
 CHAPTER 9  Ethics in Research 179

• Define technical terms, and be consistent in the use
of terminology.
• Use clear terminology, and avoid professional
jargon.
• Develop the document using headings, uppercase
and lowercase letters, and spacing to make it easy
to read.
• Use headings for major elements in the consent
form, such as “Purpose,” “Benefits,” and “Risks.”
• Use a readable font, that is, a minimum of 12- to
14-point font for text and a 16- to 18-point font for
the headers.
• Address the subject directly, using phrases such
as, “You are being asked to take part in this
study.…”
• Estimate the reading level of the document with the
use of a computerized readability formula, and
revise it to achieve no higher than an eighth grade
reading level.
Meade (1999) used these tips to simplify a para-
graph from an example consent form, as shown in Box
9-4. Once you have developed the consent document,
pilot-test it with patients who are comparable to the
proposed subjects for the study. These patients can
give feedback on the ease of reading, clarity, and
understandability of the consent document. You can
then revise the document as needed on the basis of the
feedback. These guidelines will help you develop a
clear, concise consent document that your study sub-
jects can comprehend.

Box 9-4 Simplification of Consent Document

Origin Consent Document prolonged patient survival and result in a
Example A*: decrease in the number of patients experiencing
Side effects of the marrow infusion are a recurrence.
uncommon and consist primarily of an unusual Revised Simplified Consent Document
taste from the preservative, occasional nausea
Example A*:
and vomiting, and, rarely, fever and chills. In
Side effects of getting stem cells:
addition, your chest may feel tight for a while,
• An unusual or funny taste in your mouth
but that will pass.
• Mild nausea and vomiting
Example B†:
• Fever and chills (rarely)
The standard approach to treating breast cancer
• Tightness in chest (rarely)
is to give several “cycles” (repeated doses at
Example B†:
regularly specified intervals) of a combination of
Why is this study being done?
two or more chemotherapy drugs (drugs that
The purpose of this research study is to find
kill cancer cells). Recent information suggests
out whether adding the drug Taxol (paclitaxel)
that it may be more beneficial to give several
to a commonly used chemotherapy is better
cycles of one drug followed by several cycles of
than the commonly used chemotherapy by itself
another drug. Some researchers think that the
at preventing your cancer from coming back.
second approach may kill more cells that are
The study also will see what side effects there
resistant to chemotherapy. The approach being
are from adding Taxol to the commonly used
tested in this study is to administer four cycles
chemotherapy. Taxol has been found to be
of standard chemotherapy (doxorubicin/
effective in treating patients with advanced
cyclophosphamide) followed by four cycles of
breast cancer. In this study, we want to see
the drug paclitaxel. Researchers hope to show
whether Taxol will help to treat patients with
that cancer cells resistant to the doxorubicin/
early stage breast cancer and whether the side
cyclophosphamide chemotherapy may be
effects seem to be worth the possible benefit.
sensitive to paclitaxel. This may then result in
NCI Model Document Sub-Group: Comprehensive Working Group on Informed Consent, 1998.
*Standard doses versus myeloablative therapy for previously untreated symptomatic multiple myeloma. Phase III.
SWOG 9321.
†
A randomized trial evaluating the worth of paclitaxel (Taxol) following doxorubicin (Adriamycin)/
cyclophosphamide (Cytoxan) in breast cancer.
From Broome, M. E. (1999). Consent (assent) for research with pediatric patients. Seminars in Oncology Nursing,
15(2), 130.
180 UNIT TWO  The Research Process
Researchers can also take steps to determine the
prospective subjects’ level of comprehension by
having them complete a survey or questionnaire
examining their understanding of consent information
(Cahana & Hurst, 2008). In complex, high-risk studies,
it is more difficult for subjects to comprehend consent
information, so some researchers might require pro-
spective subjects to pass a test on consent information
before becoming research subjects.
In qualitative research, the participants might com-
prehend their participation in a study at the beginning,
but unexpected events or consequences might occur
during the study to obscure that understanding. These
events might precipitate a change in the focus of the
research and the type of participation by the partici-
pants. For example, the topics of an interview might
change with an increased need for information from
the participants to address these topics. Thus, informed
consent is an ongoing, evolving process in qualitative
research. The researcher must renegotiate the partici-
pants’ consent and determine their comprehension of
that consent as changes occur in the study. By continu-
ally informing and determining the comprehension of
participants, you will establish trust with them and
promote the conduct of an ethical study (Munhall,
2012a).
Competence to Give Consent
Autonomous individuals, who are capable of under-
standing and weighing the benefits and risks of a pro-
posed study, are competent to give consent. The
competence of the subject is often determined by the
researcher using an assessment of decisional capacity
(Beattie, 2009). Persons with diminished autonomy
resulting from legal or mental incompetence, terminal
illness, or confinement to an institution might not be
legally competent to consent to participate in research
(see the earlier discussion of the right to self-
determination). However, the researcher makes every
effort to present the consent information at a level
prospective subjects can understand, so that they can
assent to the research. In addition, researchers need to
clearly present essential information for consent to the
legally authorized representative, such as the parents
or guardian, of the prospective subject (U.S. DHHS,
2009).
Voluntary Consent
Voluntary consent means that the prospective subject
has decided to take part in a study of his or her own
volition without coercion or any undue influence. Vol-
untary consent is obtained after the prospective subject
has been given essential information about the study
and has shown comprehension of this information
(U.S. FDA, 2010a; U.S. DHHS, 2009). Some research-
ers, because of their authority, expertise, or power,
have the potential to coerce subjects into participating
in research. Researchers need to ensure that their per-
suasion of prospective subjects is not coercive. Thus,
the rewards offered in a study ought to be congruent
with the risks taken by the subjects.
Documentation of Informed Consent
The documentation of informed consent depends on
(1) the level of risk involved in the study and (2) the
discretion of the researcher and those reviewing the
study for institutional approval. Most studies require
a written consent form, although in some studies, the
consent form may be replaced by oral consent or the
consent form may be used but the subject’s signature
is waived.
Written Consent Waived
The requirements for written consent may be waived
in research that “presents no more than minimal risk
of harm to subjects and involves no procedures for
which written consent is normally required outside of
the research context” (U.S. DHHS, 2009, 45 CFR
Section 46.117c). For example, if you were using
questionnaires to collect relatively harmless data, you
would not need to obtain a signed consent form from
the subjects. The subject’s completion of the question-
naire may serve as consent. The top of the question-
naire might contain a statement such as “Your
completion of this questionnaire indicates your consent
to participate in this study.”
Written consent is also waived when the only
record linking the subject and the research would be
the consent document and the principal risk is the
harm that could result from a breach of confidentiality.
The subject needs to be given the option of signing a
consent form or not and the subject’s wishes will
govern (U.S. DHHS, 2009). However, the four ele-
ments of consent—disclosure, comprehension, com-
petence, and voluntarism—are essential in all studies
whether written consent is waived or required.
Written Consent Documents
Short-Form Written Consent Document
The short-form consent document includes the fol-
lowing statement: “The elements of informed consent
required by Section 46.116 [see the section on infor-
mation essential for consent] have been presented
orally to the subject or the subject’s legally authorized
representative” (U.S. DHHS, 2009, 45 CFR Section
46.117b). The researcher must develop a written
 CHAPTER 9  Ethics in Research 181

summary of what is to be said to the subject in the oral
presentation, and the summary must be approved by
an IRB. When the oral presentation is made to the
subject or to the subject’s representative, a witness is
required. The subject or representative must sign the
short-form consent document. The witness must sign
both the short-form and a copy of the summary, and
the person actually obtaining consent must sign a copy
of the summary. Copies of the summary and short
form are given to the subject and the witness; the
researcher retains the original documents and must
keep these documents for 3 years after the end of the
study. The short-form written consent documents
might be used in studies that present minimal or mod-
erate risk to the subjects.
Formal Written Consent Document
The written consent document or consent form
includes the elements of informed consent required by
the U.S. DHHS (2009) and U.S. FDA (2010a) regula-
tions (see the previous section on information essen-
tial for consent). In addition, a consent form might
include other information required by the institution
where the study is to be conducted or by the agency
funding the study. Most universities provide consent
form guidelines for researchers to use. A sample
consent form is presented in Figure 9-2 with the
common essential consent information. The subject
can read the consent form, or the researcher can read
it to the subject; however, it is wise also to explain the
study to the subject. The subject signs the form, and

Study title: The Needs of Family Members of Critically Ill Adults

Investigator: Linda L. Norris, R.N.

Ms. Norris is a registered nurse studying the emotional and social needs of family members of patients in the
Intensive Care Units (research purpose). Although the study will not benefit you directly, it will provide information that
might enable nurses to identify family members’ needs and to assist family members with those needs (potential
benefits).
The study and its procedures have been approved by the appropriate people and review boards at The University of
Texas at Arlington and X hospital (IRB approval). The study procedures might cause fatigue for you or your family
(potential risks). The procedures include: (1) responding to a questionnaire about the needs of family members of
critically ill patients and (2) completing a demographic data sheet (explanation of procedures). Participation in this
study will take approximately 20 minutes (time commitment). You are free to ask any questions about the study or
about being a subject and you may call Ms. Norris at (999) 999-9999 (work) or (999) 999-9999 (home) if you have
further questions (offer to answer questions).
Your participation in this study is voluntary; you are under no obligation to participate (alternative option and
voluntary consent). You have the right to withdraw at any time and the care of your family member and your
relationship with the healthcare team will not be affected (option to withdraw).
The study data will be coded so they will not be linked to your name. Your identity will not be revealed while the
study is being conducted or when the study is reported or published. All study data will be collected by Ms. Norris,
stored in a secure place, and not shared with any other person without your permission (assurance of anonymity and
confidentiality).

I have read this consent form and voluntarily consent to participate in this study.
(If Appropriate)

Subject’s Signature Date Legal Representative Date

I have explained this study to the above subject and have sought his/her understanding
for informed consent

Investigator’s Signature Date

Figure 9-2  Sample consent form. Words in parentheses and boldface identify common essential consent information and would not appear in an
actual form.
182 UNIT TWO  The Research Process
the investigator or research assistant collecting the
data witnesses it. This type of consent can be used for
minimal- to moderate-risk studies. All persons signing
the consent form must receive a copy of it. The
researcher keeps the original consent form for 3 years
in a secure location, such as a locked file cabinet in a
locked room.
Studies that involve subjects with diminished
autonomy require a written consent form. If these pro-
spective subjects have some comprehension of the
study and agree to participate as subjects, they must
sign the consent form. However, each subject’s legally
authorized representative also must sign the form. The
representative indicates his or her relationship with the
subject under the signature (see Figure 9-2).
The written consent form used in a high-risk study
often contains the signatures of two witnesses, the
researcher, and an additional person. The additional
person signing as a witness must observe the informed
consent process and must not be otherwise connected
with the study. The best witnesses are research or
patient advocates who are employed by the institution.
Sometimes nurses are asked to sign a consent form as
a witness for a biomedical study. They must know the
study purpose and procedures and the subject’s com-
prehension of the study before signing the form (Fry
et al., 2011). The role of the witness is more important
in the consent process if the prospective subject is in
awe of the investigator and does not feel free to ques-
tion the procedures of the study.
Kravits, McAllister-Black, Grant, and Kirk (2010)
conducted a study to examine the effect of a psycho-
educational intervention on the stress reduction and
prevention of burnout in registered nurses (RNs).
These researchers provided a detailed description of
their consent process in the following quotation. The
researchers found that the intervention was useful in
reducing the stress and exhaustion of the RNs but
recommended additional research on the impact of the
intervention.
“Informed Consent
The study was reviewed and approved by the Insti­
tutional Review Board. Consent was obtained prior
to initiating the intervention. The primary investigator
and/or the co-investigator explained the study includ­
ing risks and benefits to all attendees including com­
munity healthcare providers. The potential participants
were assured that participation was strictly voluntary
and confidential. Consent was obtained by the project
director, an RN, and the research assistant after the
primary and co-investigators left the room to protect
the anonymity of the participants. All instruments and
materials were blind coded with an alphanumeric
code to further protect participant anonymity. When
explaining the study, permission to answer only the
questions they were comfortable with was empha­
sized. All participants, regardless of consent status,
then received the psycho-educational intervention.”
(Kravits et al., 2010, p. 133)
Recording of the Consent Process
A researcher might elect to audio-record or obtain a
DVD of the consent process. These methods docu-
ment what was said to the prospective subject and
record the subject’s questions and the investigator’s
answers. Audio-recording and DVD creation are time-
consuming and costly, however, and are not appropri-
ate for studies of minimal or moderate risk. If your
study is considered high risk, it might be wise to com-
pletely document the consent process, because doing
so might protect you and your subjects. Both of you
would retain a copy of the recording.
Authorization for Research Uses
and Disclosure
The HIPAA Privacy Rule provides individuals the
right, as research subjects, to authorize covered enti-
ties (healthcare provider, health plan, and healthcare
clearinghouse) to use or disclose their private health
information (PHI) for research purposes. This autho-
rization is regulated by the HIPAA and is separate
from the informed consent that is regulated by the U.S.
DHHS (2009) and the U.S. FDA (2010a). The autho-
rization focuses on the privacy risks and states how,
why, and with whom the PHI will be shared. The
authorization form must include the following
information:
Authorization Core Elements (see Privacy Rule,
45 CFR Section 164.508[c][1])
• “Description of PHI to be used or disclosed (iden-
tifying the information in a specific and meaningful
manner).
• The name(s) of person(s) authorized to make the
requested use or disclosure.
• The name(s) of person(s) who may use the PHI or
to whom the covered entity may make the requested
disclosure.
• Description of each purpose of the requested use
or disclosure. Researchers should note that this

element must be study specific, not for future
unspecified research.
• Authorization expiration date. The terms ‘end of
the research study’ or ‘none’ may be used for
research, including for the creation and mainte-
nance of a research database or repository.
• Signature of the individual and date. If the autho-
rization is signed by an individual’s personal rep-
resentative, a description of the representative’s
authority to act for the individual must be included.”
(U.S. DHHS, 2004)
The authorization information can be included as
part of the consent form, but it is probably best to have
two separate forms (Olsen, 2003). U.S. DHHS (2004)
developed a sample authorization form, which is pre-
sented in Figure 9-3.
Institutional Review
In institutional review, a committee of the research-
er’s peers, examines his or her study for ethical con-
cerns. The first federal policy statement on protection
of human subjects by institutional review was issued
by the U.S. PHS in 1966. The statement required that
research involving human subjects must be reviewed
by a committee of peers or associates to confirm that
(1) the rights and welfare of subjects will be protected,
(2) appropriate methods will be used to secure
informed consent, and (3) the potential benefits of the
investigation are greater than the risks (Levine, 1986).
In 1974, DHEW passed the National Research Act,
which required that all research involving human sub-
jects undergo institutional review. Currently, the U.S.
DHHS (2009, 45 CFR Sections 46.107-46.115) and
the U.S. FDA (2010b, 21 CFR Sections 56.101-
56.124) have similar regulations for institutional
review of research. These regulations describe the
membership, functions, and operations of an institu-
tional review board. An institutional review board
(IRB) is a committee that reviews research to ensure
that the investigator is conducting the research ethi-
cally. Universities, hospital corporations, and many
managed care centers have IRBs to promote the
conduct of ethical research and protect the rights of
prospective subjects at these institutions.
Each IRB has at least five members of various
backgrounds (cultural, economic, educational, gender,
racial) to promote a complete, scholarly, and fair
review of research that is commonly conducted in an
institution. If an institution regularly reviews studies
with vulnerable subjects, such as children, neonates,
pregnant women, prisoners, and mentally disabled
persons, the IRB should include one or more members
with knowledge about and experience in working with
CHAPTER 9  Ethics in Research 183
these individuals. The members must have sufficient
experience and expertise to review a variety of studies,
including quantitative, outcomes, intervention, and
qualitative research (Munhall, 2012b). The IRB
members must not have a conflicting interest related
to a study conducted by an institution. Any member
having a conflict of interest with a research project
being reviewed must excuse himself or herself from
the review process, except to provide information
requested by the IRB. The IRB also must include other
members whose primary concern is nonscientific,
such as an ethicist, a lawyer, or a minister. At least one
of the IRB members must be someone who is not
affiliated with the institution (U.S. FDA, 2010b; U.S.
DHHS, 2009). The IRBs in hospitals are often com-
posed of physicians, nurses, lawyers, scientists, clergy,
and community laypersons.
In 2009, the U.S. FDA and U.S. DHHS regulations
were revised to require all IRBs to register through a
system maintained by the DHHS. The registration
information includes contact information for the insti-
tution with the IRB and the official overseeing the
activities performed by the IRB (such as names,
addresses, emails, and telephone numbers), the
number of active protocols involving federally regu-
lated products reviewed during the preceding 12
months, and a description of the types of products
involved in the protocols reviewed (U.S. FDA, 2010b;
U.S. DHHS, 2009). The IRB registration requirement
was implemented to make it easier for the DHHS to
inspect IRBs and communicate information to them.
This rule was made effective in July of 2009 and
requires each IRB to renew its registration every 3
years.
Levels of Reviews Conducted by Institutional
Review Boards
Universities and healthcare agencies have IRBs that
function in a similar way to review research following
federal regulations (U.S. FDA. 2010b; U.S. DHHS,
2009). Faculty and students must receive IRB approval
from their university prior to seeking IRB approval at
the agency where the study is to be conducted. The
functions and operations of an IRB involve the review
of research at three different levels: (1) exempt from
review; (2) expedited review; and (3) complete review.
We want to stress with students and practicing nurses
that the level of the review required for each study is
decided by the IRB chairperson and/or committee, not
by the researcher.
Studies are usually exempt from review if they
pose no apparent risks for the research subjects. The
studies that are usually considered exempt from
IRB review by the federal regulations are identified in
184 UNIT TWO  The Research Process

AUTHORIZATION TO USE OR DISCLOSE (RELEASE) HEALTH INFORMATION

THAT IDENTIFIES YOU FOR A RESEARCH STUDY
REQUIRED ELEMENTS:
If you sign this document, you give permission to [name or other identification of specific healthcare provider(s) or
description of classes of persons, e.g., all doctors, all healthcare providers] at [name of covered entity or entities] to use or
disclose (release) your health information that identifies you for the research study described below:
[Provide a description of the research study, such as the title and purpose of the research.]

The health information that we may use or disclose (release) for this research includes
[complete as appropriate]:
[Provide a description of information to be used or disclosed for the research project. This description may include, for
example, all information in a medical record, results of physical examinations, medical history, lab tests, or certain health
information indicating or relating to a particular condition.]
The health information listed above may be used by and/or disclosed (released) to:
[Name or class of persons involved in the research; i.e., researchers and their staff**]

[Name of covered entity] is required by law to protect your health information. By signing
this document, you authorize [name of covered entity] to use and/or disclose (release)
your health information for this research. Those persons who receive your health
information may not be required by Federal privacy laws (such as the Privacy Rule) to
protect it and may share your information with others without your permission, if permitted
by laws governing them.

Please note that [include the appropriate statement]:

• You do not have to sign this Authorization, but if you do not, you may not receive research-related treatment.
(When the research involves treatment and is conducted by the covered entity or when the covered
entity provides health care solely for the purpose of creating protected health information to disclose to
a researcher)

• [Name of covered entity] may not condition (withhold or refuse) treating you on whether you sign this
Authorization.
(When the research does not involve research-related treatment by the covered
entity or when the covered entity is not providing health care solely for the
purpose of creating protected health information to disclose to a researcher)

Please note that [include the appropriate statement]:

• You may change your mind and revoke (take back) this Authorization at any time,
except to the extent that [name of covered entity(ies)] has already acted based on
this Authorization. To revoke this Authorization, you must write to: [name of the
covered entity(ies) and contact information].
(When the research study is conducted by an entity other than the covered
entity)

• You may change your mind and revoke (take back) this Authorization at any time. Even if you revoke this
Authorization, [name or class of persons at the covered entity involved in the research] may still use or disclose
health information they already have obtained about you as necessary to maintain the integrity or reliability of the
current research. To revoke this Authorization, you must write to: [name of the covered entity(ies) and contact
information].
(When the research study is conducted by the covered entity)

Signature of participant or participant’s Date

personal representative

Printed name of participant or participant’s If applicable, a description of the personal

personal representative representative’s authority to sign for the participant

** When a covered entity conducts the research study, the Authorization must list ALL names or other identification, or ALL
classes, of persons who will have access through the covered entity to the protected health information (PHI) for the
research study (e.g., research collaborators, sponsors, and others who will have access to data that includes
PHI). Examples may include, but are not limited to, the following:

• Data coordinating centers that will receive and process PHI;

• Sponsors who want access to PHI or who will actually own the research data; and/or
• Institutional Review Boards or Data Safety and Monitoring Boards.

If the research study is conducted by an entity other than the covered entity, the authorization need only list the name or other
identification of the outside researcher (or class of researchers) and any other entity to whom the covered entity is expected to
make the disclosure.

Figure 9-3  Sample of authorization language for research uses and disclosures of individually identifiable health information by a covered healthcare
provider. Words in parentheses and boldface are explanations for the reader and would not appear in an actual form.

Box 9-5 Research Qualifying for Exemption from Review
Unless otherwise required by department or
agency heads, research activities in which the
only involvement of human subjects will be in
one or more of the following categories are
exempt from review:
(1) Research conducted in established or
commonly accepted educational settings,
involving normal educational practices, such
as (i) research on regular and special
education instructional strategies, or (ii)
research on the effectiveness of or the
comparison among instructional techniques,
curricula, or classroom management
methods.
(2) Research involving the use of educational
tests (cognitive, diagnostic, aptitude,
achievement), survey procedures, interview
procedures or observation of public
behavior, unless: (i) information obtained is
recorded in such a manner that human
subjects can be identified, directly or
through identifiers linked to the subjects;
and (ii) any disclosure of the human
subjects’ responses outside the research
could reasonably place the subjects at risk of
criminal or civil liability or be damaging to
the subjects’ financial standing,
employability, or reputation.
(3) Research involving the use of educational
tests (cognitive, diagnostic, aptitude,
achievement), survey procedures, interview
procedures, or observation of public
behavior that is not exempt under
paragraph (b)(2) of this section, if: (i) the
human subjects are elected or appointed
public officials or candidates for public
office; or (ii) Federal statute(s) require(s)
From U.S. Department of Health and Human Services (U.S. DHHS, 2009). Protection of human subjects. Code of
Federal Regulations, Title 45, Part 46. Retrieved from http://www.hhs.gov/ohrp/policy/ohrpregulations.pdf/.
Box 9-5. For example, studies by nurses and other
health professionals that have no foreseeable risks or
are a mere inconvenience for subjects might be identi-
fied as exempt from review by the chairperson of the
IRB committee. Studies incorporating previously col-
lected data from which PHI has been de-identified are
usually exempt as well (U.S. DHHS, 2004).
Studies that have some risks, which are viewed as
minimal, are expedited in the review process. Minimal
CHAPTER 9  Ethics in Research 185
without exception that the confidentiality of
the personally identifiable information will
be maintained throughout the research and
thereafter.
(4) Research involving the collection or study of
existing data, documents, records,
pathological specimens, or diagnostic
specimens if these sources are publicly
available or if the information is recorded by
the investigator in such a manner that
subjects cannot be identified, directly or
through identifiers linked to the subjects.
(5) Research and demonstration projects which
are conducted by or subject to the approval
of Department or Agency heads, and which
are designed to study, evaluate, or otherwise
examine: (i) Public benefit or service
programs; (ii) procedures for obtaining
benefits or services under those programs;
(iii) possible changes in or alternatives to
those programs or procedures; or (iv)
possible changes in methods or levels of
payment for benefits or services under those
programs.
(6) Taste and food quality evaluation and
consumer acceptance studies (i) if
wholesome foods without additives are
consumed or (ii) if a food is consumed that
contains a food ingredient at or below the
level and for a use found to be safe, or
agricultural chemical or environmental
contaminant at or below the level found to
be safe, by the Food and Drug
Administration or approved by the
Environmental Protection Agency or the
Food Safety and Inspection Service of the
U.S. Department of Agriculture.
risk means “that the risks of harm anticipated in the
proposed research are not greater, considering proba-
bility and magnitude, than those ordinarily encoun-
tered in daily life or during the performance of routine
physical or psychological examinations or tests” (U.S.
DHHS, 2009, 45 CFR Section 46.102). Expedited
review procedures can also be used to review minor
changes in previously approved research. Under expe-
dited IRB review procedures, the review may be
186 UNIT TWO  The Research Process
Box 9-6 Research Qualifying for Expedited Institutional Review Board Review
Expedited review (by committee chairpersons or
designated members) is authorized for the
following research involving no more than
minimal risk:
1. Collection of hair and nail clippings, in a
nondisfiguring manner; deciduous teeth and
permanent teeth if patient care indicates a
need for extraction.
2. Collection of excreta and external secretions
including sweat, uncannulated saliva,
placenta removed at delivery, and amniotic
fluid at the time of rupture of the
membrane before or during labor.
3. Recording of data from subjects 18 years of
age or older using noninvasive procedures
routinely employed in clinical practice. This
includes the use of physical sensors that are
applied either to the surface of the body or
at a distance and do not involve input of
matter or significant amounts of energy into
the subject or an invasion of the subject’s
privacy. It also includes such procedures as
weighing, testing sensory acuity,
electrocardiography, electroencephalography,
thermography, detection of naturally
occurring radioactivity, diagnostic
echography, and electroretinography. It does
not include exposure to electromagnetic
radiation outside the visible range (for
example, x-rays, microwaves).
From U.S. Department of Health and Human Services (U.S. DHHS, 2009). Protection of human subjects. Code of
Federal Regulations, Title 45, Part 46. Retrieved from http://www.hhs.gov/ohrp/policy/ohrpregulations.pdf/.
carried out by the IRB chairperson or by one or more
experienced reviewers designated by the chairperson
from among members of the IRB. In reviewing the
research, the reviewers may exercise all of the authori-
ties of the IRB except disapproval of the research. A
research proposal may be disapproved only after a
complete review by the IRB (U.S. FDA, 2010b; U.S.
DHHS, 2009). Box 9-6 identifies research that usually
qualifies for expedited review.
A study involving greater than minimal risks to
research subjects requires a complete IRB review. To
obtain IRB approval, researchers must ensure that “(1)
risks to subjects are minimized, (2) risks to subjects
are reasonable in relation to anticipated benefits, (3)
selection of subjects is equitable, (4) informed consent
will be sought from each prospective subject or the
4. Collection of blood samples by venipuncture,
in amounts not exceeding 450 mL in an
8-week period and no more than two times
per week, from subjects 18 years of age or
older and who are in good health and not
pregnant.
5. Collection of both supragingival and
subgingival dental plaque and calculus,
provided the procedure is not more invasive
than routine prophylactic scaling of the
teeth and the process is accomplished in
accordance with accepted prophylactic
techniques.
6. Voice recordings made for research purposes
such as investigations of speech defects.
7. Moderate exercise by healthy volunteers.
8. The study of existing data, documents,
records, pathological specimens, or
diagnostic specimens.
9. Research on individual or group behavior or
characteristics of individuals, such as studies
of perception, cognition, game theory, or
test development, where the investigator
does not manipulate subjects’ behavior and
research will not involve stress to subjects.
10. Research on drugs or devices for which an
investigational new drug exemption or an
investigational device exemption is not
required.
subject’s legally authorized representative, (5)
informed consent will be appropriately documented,
(6) the research plan makes adequate provision for
monitoring data collection for subjects’ safety, and (7)
adequate provisions are made to protect the privacy of
subjects and to maintain the confidentiality of data.”
(U.S. FDA, 2010b, 21 CFR 56.111; U.S. DHHS, 2009,
45 CFR 46.111).
Every research report must indicate that the study
had IRB approval and whether the approval was from
a university and/or clinical agency. All the reports
used as examples in this chapter indicated the studies
had appropriate IRB approval. For example, Riegel et
al. (2011) provided the following description of their
IRB approval. This study involved a secondary data
analysis using a national database of HF patients to

determine their levels of self-care performance. These
researchers ensured the studies in the database had
IRB approval and that they obtained IRB approval
from their university.
“All studies had been approved by local institutional
review boards. In each, eligibility was confirmed by a
trained nurse research assistant who then explained
study requirements and obtained written informed
consent. This secondary analysis was approved by the
institutional review board of the University of Penn­
sylvania.” (Riegel et al., 2011, p. 134)
Influence of HIPAA Privacy Rule on
Institutional Review Boards
Under the 2003 HIPAA Privacy Rule, an IRB or an
institutionally established privacy board can act on
requests for a waiver or an alteration of the authoriza-
tion requirement for a research project. If an IRB and
a privacy board both exist in an agency, the approval
of only one board is required, and it will probably be
the IRB for research projects. Researchers can choose
to obtain a signed authorization form from potential
subjects or can ask for a waiver or an alteration of the
authorization requirement. An altered authorization
requirement occurs when an IRB approves a request
that some but not all of the required 18 elements be
removed from health information that is to be used in
research. The researcher can also request a partial or
complete waiver of the authorization requirement
from the IRB. For a partial waiver, discussed earlier,
the researcher obtains PHI to contact and recruit
potential subjects for a study. An IRB can give a
researcher a complete waiver of authorization in
studies in which the informed consent requirements
might also be waived. Thus, a waiver or alteration of
the authorization requirement might occur when the
following criteria have been met:
• The PHI use or disclosure involves no more than
minimal risk to the privacy for research subjects
based on (1) an adequate plan presented to the IRB
to protect the PHI identifiers from improper use of
disclosure; (2) an adequate plan exists to destroy
the identifiers at the earliest opportunity; and (3)
written assurance the PHI will not be reused or
disclosed to any other person.
• The research could not reasonably be conducted
without the waiver or alteration of the authorization
requirement.
• The research cannot be done without access to and
use of the PHI. (U.S. DHHS, 2003)
CHAPTER 9  Ethics in Research 187
The healthcare provider, health plan, or healthcare
clearinghouse cannot release the PHI to the researcher
until the following documentation has been received:
(1) the identity of the approving IRB; (2) the date the
waiver or alteration was approved; (3) IRB documen-
tation that the criteria for waiver or alteration have
been met; (4) a brief description of the PHI to which
the researcher has been granted access or use; (5) a
statement as to whether the waiver was approved
under normal or expedited review procedures; and
(6) the signature of the IRB chair or the chair’s
designee.
The HIPAA Privacy Rule does not change the IRB
membership and functions that are designated under
the U.S. DHHS and U.S. FDA regulations. For clari-
fication, the responsibilities of the IRB/privacy board
for HIPAA (U.S. DHHS, 2007b) and the responsibili-
ties of the IRB under the U.S. DHHS (2009) and U.S.
FDA (2010a) are outlined in Table 9-3.
Research Misconduct
The goal of research is to generate sound scientific
knowledge, which is possible only through the honest
conduct, reporting, and publication of studies.
However, since the 1980s, a number of fraudulent
studies have been conducted and published in presti-
gious scientific journals. An example of research mis-
conduct was evident in the publications of Dr. Robert
Slutsky, a heart specialist at the University of Califor-
nia, San Diego, School of Medicine, whose study
results raised questions of data fabrication (Friedman,
1990). In 6 years, Slutsky published 161 articles, and
at one time, he was completing an article every 10
days. Eighteen of the articles were found to be fraudu-
lent and have retraction notations, and 60 articles were
judged to be questionable.
Another example of research misconduct is the
work of Stephen Breuning, a psychologist at the Uni-
versity of Pittsburgh, who engaged in deceptive and
misleading practices in reporting his research on
retarded children. He used his fraudulent research to
obtain more than $300,000 in federal grants. In 1988,
he was criminally charged with research fraud, pleaded
guilty, was fined $20,000, and was sentenced to up to
10 years in prison (Garfield & Welljams-Dorof, 1990).
Research misconduct is also evident in nursing as
identified by a survey of nurse research directors,
coordinators, and deans of nursing programs by
Rankin and Esteves (1997). Of the 88 nurses surveyed,
27.2% reported cheating on data collection, 33.2%
identified misinterpretations of findings, 65.8%
reported protocol violations related to the study site
188 UNIT TWO  The Research Process

TABLE 9-3  Comparison of IRB/Privacy Board Responsibilities for HIPAA, U.S. DHHS, and FDA
Area of
Distinction HIPAA Privacy Rule
Permissions for Authorization
research
IRB/privacy board Requires the covered entity to
responsibilities obtain authorization for
research use or disclosure
of PHI unless a regulatory
permission applies. Because
of this, the IRB or privacy
board would see only
requests to waive or alter
the authorization
requirement. In exercising
privacy rule authority, the
IRB or privacy board does
not review the authorization
form.
U.S. DHHS Protection of
Human Subjects
Regulations Title 45 CFR
Part 46
Informed consent and
authorization
Requires the covered entity to
obtain authorization for
research use or disclosure
of PHI unless a regulatory
permission applies. Because
of this, the IRB or privacy
board would see only
requests to waive or alter
the authorization
requirement. In exercising
privacy rule authority, the
IRB or privacy board does
not review the authorization
form.
U.S. FDA Protection of Human
Subjects Regulations Title 21 CFR
Parts 50 and 56
Informed consent and authorization
The IRB must ensure that informed
consent will be sought from, and
documented for, each prospective
subject or the subject’s legally
authorized representative in accordance
with, and to the extent required by,
FDA regulations. If specified criteria
are met, the requirements for either
obtaining informed consent or
documenting informed consent may be
waived. The IRB must review and
approve the authorization form if it is
combined with the informed consent
document. Privacy boards have no
authority under the FDA Protection of
Human Subjects Regulations.

CFR, Code of Federal Regulations; DHHS, U.S. Department of Health and Human Services; U.S. FDA, U.S. Food and Drug Administration; HIPAA, Health
Insurance Portability and Accountability Act; IRB, institutional review board; PHI, protected health information.
From U.S. Department of Health and Human Services. (2007b). How can covered entities use and disclose protected health information for research and
comply with the Privacy Rule? Retrieved from http://privacyruleandresearch.nih.gov/pr_08.asp/.

and subjects, and 49.7% reported an occasional inci-
dence of plagiarism. Habermann, Broome, Pryor, and
Ziner (2010) studied 266 research coordinators, pre-
dominately RNs, who indicated they had firsthand
knowledge of scientific misconduct in the past year.
The types and frequencies of research misconduct
included: 50% protocol violations, 26.6% consent vio-
lations, 13.9% fabrication, 5.2% financial conflict of
interest, and 5% falsification.
In response to the increasing incidences of scien-
tific misconduct, the federal government developed
the Office of Research Integrity (ORI) in 1989 within
the U.S. DHHS. The ORI was to supervise the imple-
mentation of the rules and regulations related to
research misconduct and to manage any investigations
of misconduct. The most current regulations imple-
mented by the ORI (2005) are CFR 42, Parts 50 and
93, Policies of General Applicability, which are dis-
cussed in the following section.
Role of the ORI in Promoting the Conduct of
Ethical Research
The ORI was responsible for defining important terms
used in the identification and management of research
misconduct. Research misconduct was defined as
“the fabrication, falsification, or plagiarism in pro-
cessing, performing, or reviewing research, or in
reporting research results. It does not include honest
error or differences in opinion” (ORI, 2005, 42 CFR
Section 93.103). Fabrication in research is the
making up of results and the recording or reporting of
them. Falsification of research is manipulating
research materials, equipment, or processes or chang-
ing or omitting data or results such that the research
is not accurately represented in the research record.
Fabrication and falsification of research data are two
of the most common acts of research misconduct
managed by the ORI (2012) over the past 5 years.
Plagiarism is the appropriation of another person’s
ideas, processes, results, or words without giving
appropriate credit, including those obtained through
confidential review of others’ research proposals and
manuscripts.
Currently, the ORI promotes the integrity of bio-
medical and behavioral research in approximately
4000 institutions worldwide (ORI, 2011). The office
applies federal policies and regulations to protect the
integrity of the U.S. PHS’s extramural and intramural

research programs. The extramural program provides
funding to research institutions, and the intramural
program provides funding for research conducted
within the federal government. The ORI carries out its
responsibilities by:
• Developing policies, procedures, and regulations
related to the detection, investigation, and preven-
tion of research misconduct and the responsible
conduct of research.
• Reviewing and monitoring research misconduct
investigations.
• Recommending research misconduct findings and
administrative actions to the assistant secretary for
health for decision, subject to appeal (ORI, 2012).
• Assisting the Office of the General Counsel (OGC)
to present cases before the U.S. DHHS departmen-
tal appeals board.
• Providing technical assistance to institutions that
respond to allegations of research misconduct.
• Implementing activities and programs to teach
responsible conduct of research, promote research
integrity, prevent research misconduct, and improve
the handling of allegations of research misconduct.
• Conducting policy analyses, evaluations, and
research to build the knowledge base in research
misconduct, research integrity, and prevention and
to improve the DHHS research integrity policies
and procedures.
• Administering programs for maintaining institu-
tional assurances, responding to allegations of
retaliation against whistleblowers, approving intra-
mural and extramural policies and procedures, and
responding to Freedom of Information Act and
Privacy Act requests (ORI, 2005).
The ORI classifies research misconduct as (1) an
act that involves a significant departure from the
acceptable practice of the scientific community for
maintaining the integrity of the research record; (2) an
act that was committed intentionally; and (3) an alle-
gation that can be proved by a preponderance of evi-
dence. The office has a section on its website titled
“Handling Misconduct,” which includes a summary of
the allegations and investigations managed by its
office from 1994 to 2012 (ORI, 2012). The most
common sites for the investigations were medical
schools (68%), hospitals (11%), and research insti-
tutes (10%). The individuals charged with misconduct
were primarily males holding a PhD or medical degree
(MD) and were mostly associate professors, profes-
sors, and postdoctoral fellows. When research miscon-
duct was documented, the actions taken against the
researchers or agencies might have included debar-
ment from receiving federal funding for periods
CHAPTER 9  Ethics in Research 189
ranging from 18 months to 8 years; prohibition from
U.S. PHS advisory service; and other actions requiring
supervised research, certification of data, certification
of sources, and correction or retraction of articles
(ORI, 2012).
Role of Journal Editors and Researchers in
Preventing Scientific Misconduct
Editors of journals also have a major role in monitor-
ing and preventing research misconduct in the pub-
lished literature. Friedman (1990, p. 1416) identified
criteria for classifying a publication as fraudulent,
questionable, or valid and indicated that research arti-
cles were “fraudulent if there was documentation or
testimony from coauthors that the publication did not
reflect what had actually been done.” Articles were
considered questionable if no coauthor could produce
the original data or if no coauthor had personally
observed or performed each phase of the research or
participated in the research publication. A research
article was considered valid if one or more coauthors
had personally participated in each aspect of the
research and publication.
Preventing the publication of fraudulent research
requires the efforts of authors, coauthors, research
coordinators, reviewers of research reports for publi-
cation, and editors of professional journals (Hansen &
Hansen, 1995; Hawley & Jeffers, 1992; Wager, 2007).
Authors who are primary investigators for research
projects must be responsible in their conduct, report-
ing, and publication of research. Coauthors and
coworkers should question and, if necessary, chal-
lenge the integrity of a researcher’s claims. Some-
times, well-known scientists’ names have been added
to a research publication as coauthors to give it cred-
ibility. Individuals should not be listed as coauthors
unless they were actively involved in the conduct and
publication of the research.
Research coordinators in large, funded studies also
have a role to promote integrity in research and to
identify research misconduct activities. These indi-
viduals are often the ones closest to the actual conduct
of the study, during which misconduct often occurs.
In the Habermann et al. (2010) study introduced
earlier, research coordinators were identified as having
firsthand experiences with both scientific misconduct
and research integrity. Research coordinators often
learned of the misconduct firsthand, and the principal
investigator was usually identified as the responsible
party. The five major categories of misconduct identi-
fied were protocol violations, consent violations, fab-
rication, falsification, and financial conflict of interest.
Thus, Habermann et al. recommended that the
190 UNIT TWO  The Research Process
definition of research misconduct might need to be
expanded beyond fabrication, falsification, and
plagiarism.
Peer reviewers have a key role in determining the
quality and publishability of a manuscript. They are
considered experts in the field, and their role is to
examine research for inconsistencies and inaccuracies.
Editors must monitor the peer review process and
must be cautious about publishing manuscripts that
are at all questionable. Editors also need procedures
for responding to allegations of research misconduct.
They must decide what actions to take if their journal
contains an article that has proven to be fraudulent.
Usually, fraudulent publications require retraction
notations and are not to be cited by authors in future
publications (ORI, 2005).
The publication of fraudulent research is a major
concern in medicine and nursing (Habermann et al.,
2010; Njie & Thomas, 2001; Rankin & Esteves, 1997).
The smaller pool of funds available for research and
the greater emphasis on research publications could
lead to a higher incidence of fraudulent publications.
However, the ORI (2011; 2012) has made major
advances in addressing research misconduct and the
management of fraudulent publications by: (1) identi-
fying the administrative actions for acts of research
misconduct; (2) developing a process for notifying
funding agencies and journals of acts of research mis-
conduct; and (3) providing for public disclosure of the
incidents of research misconduct.
Each researcher is responsible for monitoring the
integrity of his or her research protocols, results, and
publications. In addition, nursing professionals and
journal editors must foster a spirit of intellectual
inquiry, mentor prospective scientists regarding the
norms for good science, and stress quality, not quan-
tity, in publications (Wager, 2007; Wocial, 1995).
Animals as Research Subjects
The use of animals as research subjects is a controver-
sial issue of growing interest to nurse researchers. A
small but increasing number of nurse scientists are
conducting physiological studies that require the use
of animals. Many scientists, especially physicians,
believe that the current animal rights’ movement could
threaten the future of health research. The animal
rights groups are active in anti-research campaigns
and are often backed by financial resources estimated
in the millions of dollars (Pardes, West, & Pincus,
1991). The goal of these groups is to raise the con-
sciousness of researchers and society to ensure that
animals are used wisely in the conduct of research and
treated humanely. However, some of these animal
rights groups have tried to frighten the public with
sometimes distorted stories about inhumane treatment
of animals in research. Some of the activist leaders
have made broad comparisons between human life
and animal life. For example, a major animal rights
group called People for the Ethical Treatment
of Animals (PETA) has a website that posts videos
and blogs about the unethical treatment of animals
in research (see http://www.peta.org/tv/videos/
investigations-animal-experimentation/). Some of
these activists have now progressed to violence, using
physical attacks, including real bombs, arson, and van-
dalism. Even more damage is being done to research
through lawsuits that have blocked the conduct of
research and the development of new research centers.
Medical schools now spend millions of dollars annu-
ally for security, public education, and other efforts to
defend research.
The use of animals in research is a complicated
issue that requires careful consideration by investiga-
tors, considering the knowledge that is needed to
manage healthcare problems. Two important ques-
tions must be addressed: Should animals be used as
subjects in research, and, if animals are used in
research, what mechanisms ensure that they are treated
humanely? In regard to the first question, the type of
research project developed influences the selection of
subjects. Animals are just one of a variety of types of
subjects used in research; others are human beings,
plants, and computer data sets. Most researchers use
nonanimal subjects in their studies because they are
generally less expensive for the type of study being
conducted. In studies that are low risk, which most
nursing studies are, human beings are commonly used
as subjects.
Some studies, however, require the use of animals
to answer the research question. Animals are more
commonly used in laboratory studies that involve
investigation of high-risk physiological variables.
Approximately 17 to 22 million animals are used in
research each year, and 90% of them are rodents, with
the combined percentage of dogs and cats being only
1% to 2% (Goodwin & Morrison, 2000). Studies using
animals account for about one eighth of the published
studies (Osborne, Payne, & Newman, 2009). Because
animals are deemed valuable subjects for selected
research projects, the second question, concerning
their humane treatment, must also be answered. At
least five separate types of regulations exist to protect
research animals from mistreatment. The federal gov-
ernment, state governments, independent accreditation
organizations, professional societies, and individual

institutions work to ensure that research animals are
used only when necessary and only under humane
conditions. At the federal level, animal research is
conducted according to the guidelines of U.S. PHS
Policy on Humane Care and Use of Laboratory
Animals, which was adopted in 1986, reprinted essen-
tially unchanged in 1996, and is available on the Office
of Laboratory Animal Welfare (OLAW, 2008) website
(http://grants.nih.gov/grants/olaw/olaw.htm/).
The Humane Care and Use of Laboratory Animals
Regulations define animal as any live, vertebrate
animal used or intended for use in research, research
training, experimentation, or biological testing or for
related purposes. Any institution proposing research
involving animals must have a written Animal Welfare
Assurance statement acceptable to the U.S. PHS that
documents compliance with the U.S. PHS policy.
Every assurance statement is evaluated by the National
Institutes of Health’s Office for Protection from
Research Risks (OPRR) to determine the adequacy of
the institution’s proposed program for the care and use
of animals in activities conducted or supported by the
U.S. PHS (OLAW, 2011). The studies including
animals require an animal-use protocol that describes
the following elements: (1) research project; (2) ratio-
nale for animal use and consideration of alternatives;
(3) justification for the choice of species and numbers
of animals; (4) research procedures involving animals;
(5) procedures to minimize pain and distress; (6)
animal living conditions and veterinary care; (7)
names and qualifications of personnel who will
perform work with animals; (8) method of euthanasia;
and (9) endpoint criteria. OLAW (2011) website
includes guidelines for the care and use of animals in
research (http://grants.nih.gov/grants/olaw/Investiga
torsNeed2Know.pdf/).
Institutions’ assurance statements about compli-
ance with the U.S. PHS policy have promoted the
humane care and treatment of animals in research. In
addition, more than 700 institutions conducting health-
related research have sought accreditation by the
American Association for Accreditation of Laboratory
Animal Care (AAALAC), which was developed to
ensure the humane treatment of animals in research
(OLAW, 2011). In conducting research, each investi-
gator must carefully select the type of subject needed;
if animals are used as subjects, they require humane
treatment. Osborne et al. (2009) conducted a study to
determine the journal editorial policies regarding use
of animals in research. These researchers found that
journals need clear polices on the essential informa-
tion to be included in a research report to reflect the
fair treatment of animals in studies.
CHAPTER 9  Ethics in Research 191
KEY POINTS
• The ethical conduct of research starts with the
identification of the study topic and continues
through the publication of the study if quality
research evidence is going to be developed for
practice.
• The debate about ethics and research must continue
probably because of (1) the complexity of human
rights issues; (2) the focus of research in new, chal-
lenging arenas of technology and genetics; (3) the
complex ethical codes and regulations governing
research; and (4) the variety of interpretations of
these codes and regulations.
• Two historical documents that have had a strong
impact on the conduct of research are the Nurem-
berg Code and the Declaration of Helsinki. More
recently, the U.S. Department of Health and Human
Services (U.S. DHHS, 2009) and the U.S. Food and
Drug Administration (U.S. FDA, 2010a, 2010b)
have promulgated regulations that direct the ethical
conduct of research. These regulations include (1)
general requirements for informed consent; (2)
documentation of informed consent; (3) institu-
tional review board (IRB) review of research; (4)
exempt and expedited review procedures for certain
kinds of research; and (5) criteria for IRB approval
of research.
• The Council for International Organizations of
Medical Sciences revises and updates ethical
guidelines for biomedical research conducted
internationally.
• Public Law 104-191, the Health Insurance Portabil-
ity and Accountability Act (HIPAA), was enacted
in 1996 and implemented in 2003 to protect an
individual’s health information.
• Conducting research ethically requires protection
of the human rights of subjects. Human rights are
claims and demands that have been justified in the
eyes of an individual or by the consensus of a group
of individuals. The human rights that require pro-
tection in research are (1) self-determination; (2)
privacy; (3) anonymity and confidentiality; (4) fair
treatment; and (5) protection from discomfort and
harm.
• The rights of research subjects can be protected by
balancing benefits and risks of a study, securing
informed consent, and submitting the research for
institutional review.
• To balance the benefits and risks of a study, the
type, level, and number of risks are examined,
and the potential benefits are identified. If possible,
the risks must be minimized and the benefits
192 UNIT TWO  The Research Process
maximized to achieve the best possible benefit-risk
ratio.
• Informed consent involves the transmission of
essential information, comprehension of that infor-
mation, competence to give consent, and voluntary
consent of the prospective subject.
• In institutional review, a committee of peers called
an institutional review board (IRB) examines each
study for ethical concerns. The IRB conducts three
levels of review: exempt, expedited, and complete.
The process for accessing protected health infor-
mation according to the HIPAA Privacy Rule is
also detailed.
• Research misconduct includes fabrication, falsifi-
cation, and plagiarism during the conduct, report-
ing, or publication of research. The Office of
Research Integrity (ORI) was developed to inves-
tigate and manage incidents of research misconduct
to protect the integrity of research in all
disciplines.
• Another current ethical concern in research is the
use of animals as subjects. Two important questions
are addressed: Should animals be used as research
subjects? and If animals are used in research, what
mechanisms ensure that they are treated humanely?
The U.S. Public Health Service Policy on Humane
Care and Use of Laboratory Animals provides
direction for the conduct of research with animals
as subjects.
REFERENCES
American Nurses Association (ANA, 2001). Code of ethics for
nurses with interpretive statements. Washington, DC: American
Nurses Association.
American Psychological Association (APA, 2010). Ethical princi-
ples of psychologists and code of conduct. Washington, DC:
American Psychological Association. Retrieved June 20, 2011,
from http://www.apa.org/ethics/code/index.aspx/.
Beattie, E. (2009). Research participation of individuals with
dementia: Decisional capacity, informed consent, and consider-
ations for nurse investigators. Research in Gerontological
Nursing, 2(2), 94–102.
Beebe, L. H. & Smith, K. (2010). Informed consent to research in
persons with schizophrenia spectrum disorders. Nursing Ethics,
17(4), 425–434.
Beecher, H. K. (1966). Ethics and clinical research. New England
Journal of Medicine, 274(24), 1354–1360.
Brandt, A. M. (1978). Racism and research: The case of the Tuske-
gee Syphilis Study. Hastings Center Report, 8(6), 21–29.
Broome, M. E. (1999). Consent (assent) for research with pediatric
patients. Seminars in Oncology Nursing, 15(2), 96–103.
Broome, M. E. & Stieglitz, K. A. (1992). The consent process
and children. Research in Nursing & Health, 15(2), 147–
152.
Bruce, S. L. & Grove, S. K. (1994). The effect of a coronary artery
risk evaluation program on serum lipid values and cardiovascular
risk levels. Applied Nursing Research, 7(2), 67–74.
Cahana, A. & Hurst, S. A. (2008). Voluntary informed consent
in research and clinical care: An update. Pain Practice, 8(6),
446–451.
Council for International Organizations of Medical Sciences
(CIOMS, 2010). CIOMS: About us. Retrieved from http://www.
cioms.ch/about/frame_about.htm/.
Damrosch, S. P. (1986). Ensuring anonymity by use of subject-
generated identification codes. Research in Nursing & Health,
9(1), 61–63.
Erlen, J. A. (2010). Informed consent: Revisiting the issues. Ortho-
paedic Nursing, 29(4), 276–280.
Fawcett, J. & Garity, J. (2009). Evaluation research for evidence-
based practice. Philadelphia, PA: F. A. Davis.
Ford, J. S. & Reutter, L. I. (1990). Ethical dilemmas associated
with small samples. Journal of Advanced Nursing, 15(2),
187–191.
Friedman, P. J. (1990). Correcting the literature following fraudu-
lent publication. JAMA, 263(10), 1416–1419.
Fry, S. T., Veatch, R. M. & Taylor, C. (2011). Case studies in
nursing ethics (4th ed.). Sudbury, MA: Jones & Bartlett
Learning.
Garfield, E. & Welljams-Dorof, A. (1990). The impact of fraudulent
research on the scientific literature: The Stephen E. Breuning case.
JAMA, 263(10), 1424–1426.
Goodwin, F. K. & Morrison, A. R. (2000). Science and self-doubt.
Reason, 32(5), 22–28.
Habermann, B., Broome, M., Pryor, E. R., & Ziner, K. W. (2010).
Research coordinators’ experiences with scientific misconduct
and research integrity. Nursing Research, 59(1), 51–57.
Hansen, B. C. & Hansen, K. D. (1995). Academic and scientific
misconduct: Issues for nursing educators. Journal of Professional
Nursing, 11(1), 31–39.
Hawley, D. J. & Jeffers, J. M. (1992). Scientific misconduct as a
dilemma for nursing. Image—Journal of Nursing Scholarship,
24(1), 51–55.
Hershey, N. & Miller, R. D. (1976). Human experimentation and
the law. Germantown, MD: Aspen.
Hinds, P. S., Burghen, E. A., & Pritchard, M. (2007). Conducting
end-of-life studies in pediatric oncology. Western Journal of
Nursing Research, 29(4), 448–465.
Holaday, B., Gonzales, O., & Mills, D. (2007). Assent of school-age
bilingual children. Western Journal of Nursing Research, 29(4),
466–485.
Jones, D. J., Munro, C. L., Grap, M. J., Kitten, T., & Edmond, M.
(2010). Oral care and bacteremia risk in mechanically ventilated
adults. Heart & Lung, 39(6S) S57–S65.
Kelman, H. C. (1967). Human use of human subjects: The problem
of deception in social psychological experiments. Psychological
Bulletin, 67(1), 1–11.
Kravits, K., McAllister-Black, R., Grant, M., & Kirk, C. (2010).
Self-care strategies for nurses: A psycho-educational intervention
for stress reduction and the prevention of burnout. Applied
Nursing Research, 23(3), 130–138.
Levine, R. J. (1986). Ethics and regulation of clinical research (2nd
ed.). Baltimore-Munich: Urban & Schwarzenberg.

McClain, N., Laughon, K., Steeves, R., & Parker, B. (2007). Bal-
ancing the needs of scientist and the subject in trauma research.
Western Journal of Nursing Research, 29(1), 121–128.
Meade, C. D. (1999). Improving understanding of the informed
consent process and document. Seminars in Oncology Nursing,
15(2), 124–137.
Milgram, S. (1963). Behavioral study of obedience. Journal of
Abnormal and Social Psychology, 67(4), 371–378.
Munhall, P. L. (2012a). Ethical considerations in qualitative
research. In P. L. Munhall (Ed.), Nursing research: A qualitative
perspective (5th ed.) (pp. 491–502). Sudbury, MA: Jones &
Bartlett Learning.
Munhall, P. L. (2012b). Institutional review of qualitative research
proposals: A task of no small consequence. In P. L. Munhall (Ed.),
Nursing research: A qualitative perspective (5th ed.) (pp. 503–
515). Sudbury, MA: Jones & Bartlett Learning.
National Commission for the Protection of Human Subjects of Bio-
medical and Behavioral Research. (1978). Belmont report: Ethical
principles and guidelines for research involving human subjects
(DHEW Publication No. [05] 78-0012). Washington, DC: U.S.
Government Printing Office.
Njie, V. P. S. & Thomas, A. C. (2001. Quality issues in clinical
research and the implications on health policy (QICHRHP).
Journal of Professional Nursing, 17(5), 233–242.
Nuremberg Code. (1949). Trials of War Criminals before the
Nuremberg Military Tribunals under Control Council Law No. 10,
Vol. 2 (pp. 181–182) Washington, D.C.: U.S. Government Printing
Office, 1949. Retrieved from http://ohsr.od.nih.gov/guidelines/
nuremberg.html/.
Office of Laboratory Animal Welfare (OLAW, 2008). Policy and
guidance: Policy and guidance for the care and use of animals.
Retrieved from http://grants.nih.gov/grants/olaw/olaw.htm/.
Office of Laboratory Animal Welfare (OLAW, 2011). For research-
ers and institutions: Good animal care and good science go hand-
in-hand. Retrieved from http://grants.nih.gov/grants/policy/air/
researchers_institutions.htm/.
Office of Research Integrity (ORI, 2005). Public Health Service
Policies on Research Misconduct. Code of Federal Regulations,
Title 42, Parts 50 and 93, Policies of General Applicability.
Retrieved from http://ori.dhhs.gov/documents/FR_Doc_05–9643
.shtml.
Office of Research Integrity (ORI, 2011). About ORI—History.
Retrieved from http://ori.dhhs.gov/about/history.shtml/.
Office of Research Integrity (ORI, 2012). Handling misconduct—
Case summaries. Retrieved from http://ori.dhhs.gov/misconduct/
cases/.
Olsen, D. P. (2003). Methods: HIPAA privacy regulations and
nursing research. Nursing Research, 52(5), 344–348.
Orb, A., Eisenhauer, L., & Wynaden, D. (2001). Ethics in qualitative
research. Journal of Nursing Scholarship, 33(1), 93–96.
Osborne, N. J., Payne, D., & Newman, M. L. (2009). Journal edito-
rial policies, animal welfare, and the 3 Rs. The American Journal
of Bioethics, 9(12), 55–59.
Pardes, H., West, A., & Pincus, H. A. (1991). Physicians and the
animal-rights movement. New England Journal of Medicine,
324(23), 1640–1643.
Ramos, M. C. (1989). Some ethical implications of qualitative
research. Research in Nursing & Health, 12(1), 57–63.
CHAPTER 9  Ethics in Research 193
Rankin, M. & Esteves, M. D. (1997). Perceptions of scientific mis-
conduct in nursing. Nursing Research, 46(5), 270–275.
Rew, L., Horner, S. D., & Fouladi, R. T. (2010). Factors associated
with health behaviors in middle childhood. Journal of Pediatric
Nursing, 25(3), 157–166.
Reynolds, P. D. (1979). Ethical dilemmas and social science
research. San Francisco, CA: Jossey-Bass.
Riegel, B., Lee, C. S., Albert, N., Lennie, T., Chang, M., Song,
E. K., Bentley, B., et al. (2011). From novice to expert: Confi-
dence and activity status determine heart failure self-care perfor-
mance. Nursing Research, 60(2), 132–138.
Rothman, D. J. (1982). Were Tuskegee and Willowbrook “studies
in nature”? Hastings Center Report, 12(2), 5–7.
Simpson, C. (2010). Decision-making capacity and informed
consent to participate in research by cognitively impaired indi-
viduals. Applied Nursing Research, 23(4), 221–226.
Steinfels, P. & Levine, C. (1976). Biomedical ethics and the shadow
of Naziism. Hastings Center Report, 6(4), 1–20.
Thompson, P. J. (1987). Protection of the rights of children as
subjects for research. Journal of Pediatric Nursing, 2(6), 392–399.
U.S. Department of Health and Human Services (U.S. DHHS, 1981,
January 26). Final regulations amending basic HHS policy for the
protection of human research subjects. Code of Federal Regula-
tions, Title 45, Part 46.
U.S. Department of Health and Human Services (U.S. DHHS,
2003). Health information privacy: Summary of the HIPAA
Privacy Rule. Retrieved from http://www.hhs.gov/ocr/privacy/
hipaa/understanding/summary/index.html/.
U.S. Department of Health and Human Services (U.S. DHHS,
2004). Institutional review boards and the HIPAA Privacy
Rule. Retrieved from http://privacyruleandresearch.nih.gov/
irbandprivacyrule.asp/.
U.S. Department of Health and Human Services (U.S. DHHS,
2007a). How do other privacy protections interact with the
privacy rule? Retrieved from http://privacyruleandresearch
.nih.gov/pr_05.asp/.
U.S. Department of Health and Human Services (U.S. DHHS,
2007b). How can covered entities use and disclose protected
health information for research and comply with the Privacy
Rule? Retrieved from http://privacyruleandresearch.nih.gov/pr_
08.asp/.
U.S. Department of Health and Human Services (U.S. DHHS,
2009). Protection of human subjects. Code of Federal Regula-
tions, Title 45, Part 46. Retrieved from http://www.hhs.gov/ohrp/
policy/ohrpregulations.pdf.
U.S. Department of Health and Human Services (U.S. DHHS,
2010). HIPAA Privacy Rule Information for researchers: Over-
view. Retrieved from http://privacyruleandresearch.nih.gov/.
U.S. Department of Health and Human Services (U.S. DHHS,
2012). Office for Human Research Protections (OHRP). Retrieved
from http://www.hhs.gov/ohrp/.
U.S. Food and Drug Administration (U.S. FDA, 2007). Regulatory
information: Food and Drug Administration Amendments Act
(FDAAA) of 2007. Public Law 110-185. Retrieved from http://
www.fda.gov/RegulatoryInformation/Legislation/FederalFood
DrugandCosmeticActFDCAct/SignificantAmendments
totheFDCAct/FoodandDrugAdministrationAmendmentsActof
2007/FullTextofFDAAALaw/default.htm/.
194 UNIT TWO  The Research Process
U.S. Food and Drug Administration (FDA, 2010a). Protection of
human subjects (informed consent). Code of Federal Regulations,
Title 21, Part 50. Retrieved from http://www.accessdata.fda.gov/
scripts/cdrh/cfdocs/cfcfr/CFRsearch.cfm?CFRPart=50/.
U.S. Food and Drug Administration (FDA, 2010b). Institutional
review boards. Code of Federal Regulations, Title 21, Part 56. Re­
trieved from http://www.accessdata.fda.gov/scripts/cdrh/cfdocs/
cfcfr/CFRsearch.cfm?CFRPart=56/.
Wager, E. (2007). What do journal editors do when they suspect
research misconduct? Medicine & Law, 26(3), 535–544.
Wocial, L. D. (1995). The role of mentors in promoting integrity
and preventing scientific misconduct in nursing research. Journal
of Professional Nursing, 11(5), 276–280.
World Medical Association (WMA) General Assembly. (1964).
Declaration of Helsinki (1964). Helsinki, Finland: Author.
Retrieved June 20, 2011, from http://www.cirp.org/library/ethics/
helsinki/.
World Medical Association (WMA) General Assembly. (2008).
World Medical Association Declaration of Helsinki: Ethical prin-
ciples for medical research involving human subjects. Seoul,
Korea: Author. Retrieved from http://www.wma.net/en/30publi
cations/10policies/b3/.
   http://evolve.elsevier.com/Grove/practice/
Understanding Quantitative
Research Design
A
research design is the blueprint for conduct-
ing a study. It maximizes control over factors
that could interfere with the validity of the
study findings. Being able to identify the study design
and to evaluate design flaws that might threaten the
validity of findings is an important part of critically
appraising studies. When you are conducting a study,
the research design guides you in planning and imple-
menting the study in a way to achieve accurate results.
The control achieved through the quantitative study
design increases the probability that your study find-
ings are an accurate reflection of reality.
The term research design is used in two ways in
the nursing literature. Some consider research design
to be the entire strategy for the study, from identifica-
tion of the problem to final plans for data collection.
Others limit design to clearly defined structures within
which the study is implemented. In this text, the first
definition refers to the research methodology and the
second is a definition of the research design. The
research design of a study is the end result of a series
of decisions you will make concerning how best to
implement your study. The design is closely associ-
ated with the framework of the study. As a blueprint,
the design is not specific to a particular study but
rather is a broad pattern or guide that can be applied
to many studies (see Chapter 11 for different types of
quantitative research designs). Just as the blueprint for
a house must be individualized to the house being
built, so the design must be made specific to a study.
Using the problem statement, framework, research
questions, and clearly defined variables, you can map
out the design to achieve a detailed research plan for
collecting and analyzing data.
This chapter gives you a background for under-
standing the elements of a design and critically
appraising the designs in published quantitative
studies. You are introduced to (1) the concepts impor-
tant to design; (2) design validity; and (3) the elements
10 CHAPTER
of a good design. You are also provided questions to
assist you in selecting and implementing a design in
a study. The chapter concludes with a discussion of
mixed methods, which are relatively recent approaches
used in nursing that combine quantitative and qualita-
tive research designs.
Concepts Important to Design
Many terms used in discussing research design have
special meanings within this context. An understand-
ing of these concepts is essential for recognizing the
purpose of a specific design. Some of the major con-
cepts used in relation to research design are causality,
bias, manipulation, control, and validity.
Causality
The first assumption you must make in examining
causality is that causes lead to effects. Some of the
ideas related to causation emerged from the logical
positivist philosophical tradition. Hume, a positivist,
proposed that the following three conditions must be
met to establish causality (1) there must be a strong
relationship between the proposed cause and the
effect; (2) the proposed cause must precede the effect
in time; and (3) the cause has to be present whenever
the effect occurs. Cause, according to Hume, is not
directly observable but must be inferred (Kerlinger &
Lee, 2000; Shadish, Cook, & Campbell, 2002).
A philosophical group known as essentialists pro-
posed that two concepts must be considered in deter-
mining causality: necessary and sufficient. The
proposed cause must be necessary for the effect to
occur. (The effect cannot occur unless the cause first
occurs.) The proposed cause must also be sufficient
(requiring no other factors) for the effect to occur. This
leaves no room for a variable that may sometimes, but
not always, serves as the cause of an effect. John
Stuart Mill, another philosopher, added a third idea
195
196 UNIT TWO  The Research Process
related to causation. He suggested that, in addition to
the preceding criteria for causation, there must be no
alternative explanations for why a change in one vari-
able seems to lead to a change in a second variable
(Campbell & Stanley, 1963).
Causes are frequently expressed within the propo-
sitions of a theory. Testing the accuracy of these theo-
retical statements indicates the usefulness of the theory
(Fawcett & Garity, 2009). A theoretical understanding
of causation is considered important because it
improves our ability to predict and, in some cases, to
control events in the real world. The purpose of an
experimental design is to examine cause and effect.
The independent variable in a study is expected to be
the cause, and the dependent variable is expected to
reflect the effect of the independent variable.
Multicausality
Multicausality, the recognition that a number of inter-
relating variables can be involved in causing a particu-
lar effect, is a more recent idea related to causality.
Because of the complexity of causal relationships, a
theory is unlikely to identify every variable involved
in causing a particular phenomenon. A study is unlikely
to include every component influencing a particular
change or effect.
Cook and Campbell (1979) have suggested three
levels of causal assertions that one must consider in
establishing causality. Molar causal laws relate to
large and complex objects. Intermediate mediation
considers causal factors operating between molar and
micro levels. Micromediation examines causal con-
nections at the level of small particles, such as atoms.
Cook and Campbell (1979) used the example of
turning on a light switch, which causes the light to
come on (molar). An electrician would tend to explain
the cause of the light coming on in terms of wires and
electrical current (intermediate mediation). However,
the physicist would explain the cause of the light
coming on in terms of ions, atoms, and subparticles
(micromediation).
The essentialists’ ideas of necessary and sufficient
do not hold up well when one views a phenomenon
from the perspective of multiple causation. The light
switch may not be necessary to turn on the light if the
insulation has worn off the electrical wires. Addition-
ally, even though the switch is turned on, the light will
not come on if the light bulb is burned out. Although
this is a concrete example, it is easy to relate it to
common situations in nursing.
Few phenomena in nursing can be clearly reduced
to a single cause and a single effect. However, the
greater the proportion of causal factors that can be
identified and explored, the clearer the understanding
of the phenomenon. This greater understanding
improves our ability to predict and control. For
example, currently nurses have a limited understand-
ing of patients’ preoperative attitudes, knowledge,
and behaviors and their effects on postoperative atti-
tudes and behaviors. Nurses assume that high preop-
erative anxiety leads to less healthy postoperative
responses and that providing information before
surgery improves healthy responses in the postopera-
tive period. Many nursing studies have examined this
particular phenomenon. However, the causal factors
involved are complex and have not been clearly delin-
eated. The research evidence needed to reduce patients’
anxiety and improve their postoperative recovery is
still evolving.
Probability
The original criteria for causation required that a
variable should have an identified effect each time
the cause occurred. Although this criterion may apply
in the basic sciences, such as chemistry or physics,
it is unlikely to apply in the health sciences or
social sciences. Because of the complexity of the
nursing field, nurses deal in probabilities. Probabil-
ity addresses relative, rather than absolute, causality.
From the perspective of probability, a cause will not
produce a specific effect each time that particular
cause occurs.
Reasoning changes when one thinks in terms of
probabilities. The researcher investigates the probabil-
ity that an effect will occur under specific circum-
stances. Rather than seeking to prove that A causes B,
a researcher would state that if A occurs, there is a
50% probability that B will occur. The reasoning
behind probability is more in keeping with the com-
plexity of multicausality. In the example about preop-
erative attitudes and postoperative outcomes, nurses
could seek to predict the probability of unhealthy post-
operative patient outcomes when preoperative anxiety
levels are high.
Causality and Nursing Philosophy
Traditional theories of prediction and control are built
on theories of causality. The first research designs
were also based on causality theory. Nursing science
must be built within a philosophical framework of
multicausality and probability. The strict senses of
single causality and of “necessary and sufficient” are
not in keeping with the progressively complex, holis-
tic philosophy of nursing. To understand multicausal-
ity and increase the probability of being able to predict
and control the occurrence of an effect, researchers
 CHAPTER 10  Understanding Quantitative Research Design 197
need to comprehend both wholes and parts (Fawcett
& Garity, 2009; Shadish et al., 2002).
Practicing nurses must be aware of the molar,
intermediate mediational, and micromediational
aspects of a particular phenomenon. A variety of
differing approaches, reflecting both qualitative and
quantitative research, are necessary to develop a
knowledge base for nursing. Some see explanation
and causality as different and perhaps opposing forms
of knowledge. Nevertheless, the nurse must join these
forms of knowledge, sometimes within the design of
a single study, to acquire the knowledge needed
for nursing practice (Creswell, 2009; Marshall &
Rossman, 2011).
Bias
The term bias means to slant away from the true or
expected. A biased opinion has failed to include both
sides of the question. A biased witness is one who is
strongly for or against one side of the situation. A
biased scale is one that does not provide a valid mea-
surement of a concept.
Bias is of great concern in research because of the
potential effect on the meaning of the study findings.
Any component of the study that deviates or causes a
deviation from true measure leads to error and dis-
torted findings. Many factors related to research can
be biased: the researcher, the measurement methods,
the individual subjects, the sample, the data, and the
statistics (Grove, 2007; Thompson, 2002; Waltz,
Strickland, & Lenz, 2010). In critically appraising a
study, you need to look for possible biases in these
areas. An important concern in designing a study is to
identify possible sources of bias and eliminate or
avoid them. If they cannot be avoided, you need to
design your study to control them. Designs, in fact,
are developed to reduce the possibilities of bias
(Shadish et al., 2002).
Manipulation
Manipulation tends to have a negative connotation and
is associated with one person underhandedly maneu-
vering a second person so that he or she behaves or
thinks in the way the first person desires. Denotatively,
to manipulate means to move around or to control the
movement of something, such as manipulating a
syringe to give an injection. The major role of nurses
is to implement interventions that involve manipula-
tion of events related to patients and their environment
to improve their health. Manipulation has a specific
meaning when used in experimental or quasi-
experimental research because it is the manipulation
or implementation of the study treatment or
intervention. The experimental group receives the
treatment or intervention during a study and the
control group does not. For example, in a study on
preoperative care, preoperative relaxation therapy
might be manipulated so that the experimental group
receives the treatment and the control group does not.
In a study on oral care, the frequency of care might be
manipulated to determine its effect on patient out-
comes (Doran, 2011).
In nursing research, when experimental designs are
used to explore causal relationships, the nurse must be
free to manipulate the variables under study. For
example, in a study of pain management, if the
freedom to manipulate pain control measures is under
the control of someone else, a bias is introduced into
the study. In qualitative, descriptive, and correlational
studies, the researcher does not attempt to manipulate
variables. Instead, the purpose is to describe a situa-
tion as it exists (Marshall & Rossman, 2011; Munhall,
2012).
Control
Control means having the power to direct or manipu-
late factors to achieve a desired outcome. In a study
of pain management, one must be able to control inter-
ventions to relieve pain. The idea of control is impor-
tant in research, particularly in experimental and
quasi-experimental studies. The more control the
researcher has over the features of the study, the more
credible the study findings. The purpose of research
designs is to maximize control factors in the study
(Shadish et al., 2002).
Study Validity
Study validity, a measure of the truth or accuracy of
a claim, is an important concern throughout the
research process. Study validity is central to building
sound evidence for practice. Questions of validity
refer back to the propositions from which the study
was developed and address their approximate truth or
falsity. Is the theoretical proposition from the study
framework an accurate reflection of reality? Was the
study designed to provide a valid test of the proposi-
tion? Validity is a complex idea that is important to
the researcher and to those who read the study report
and consider using the findings in their practice. Criti-
cal appraisal of research requires that we think through
threats to validity and make judgments about how
seriously these threats affect the integrity of the find-
ings. Validity provides a major basis for making deci-
sions about which findings are sufficiently valid to add
to the evidence base for practice.
198 UNIT TWO  The Research Process
Shadish et al. (2002) have described four types
of validity: statistical conclusion validity, internal
validity, construct validity, and external validity. These
types of design validity need to be critically appraised
for strengths and possible threats in published studies.
When conducting a study, you will be confronted with
major decisions about the four types of design validity.
To make these decisions, you must address a variety
of questions, such as the following:
1. Is there a relationship between the two variables?
(statistical conclusion validity)
2. Given that there is a relationship, is it possibly
causal from the independent variable to the depen-
dent variable, or would the same relationship have
been obtained in the absence of any treatment or
intervention? (internal validity)
3. Given that the relationship is probably causal and
is reasonably known to be from one variable to
another, what are the particular cause-and-effect
constructs involved in the relationship? (construct
validity)
4. Given that there is probably a causal relationship
from construct A to construct B, can this relation-
ship be generalized across persons, settings, and
times? (external validity) (Cook & Campbell,
1979; Shadish et al., 2002)
Statistical Conclusion Validity
The first step in inferring cause is to determine whether
the independent and dependent variables are related.
You can determine this relationship (covariation)
through statistical analysis. Statistical conclusion
validity is concerned with whether the conclusions
about relationships or differences drawn from statisti-
cal analysis are an accurate reflection of the real world.
The second step is to identify differences between
groups. There are reasons why false conclusions can
be drawn about the presence or absence of a relation-
ship or difference. The reasons for the false conclusions
are called threats to statistical conclusion validity.
These threats are described in the following section.
Low Statistical Power
Low statistical power increases the probability of
concluding that there is no significant difference
between samples when actually there is a difference
(Type II error, failing to reject a false null) (see Chapter
8 for discussion of the null hypothesis). A Type II error
is most likely to occur when the sample size is small
or when the power of the statistical test to determine
differences is low (Aberson, 2010). The concept of
statistical power and strategies to improve it are dis-
cussed in Chapters 15 and 21.
Violated Assumptions of Statistical Tests
Most statistical tests have assumptions about the data
collected, such as the following: (1) the data are at
least at the interval level; (2) the sample was randomly
obtained; and (3) the distribution of scores was normal.
If these assumptions are violated, the statistical analy-
sis may provide inaccurate results (Corty, 2007;
Grove, 2007). The assumptions of statistical tests
commonly conducted in nursing studies are provided
in Chapters 23, 24, and 25.
Fishing and the Error Rate Problem
A serious concern in research is incorrectly conclud-
ing that a relationship or difference exists when it does
not (Type I error, rejecting a true null). The risk of
Type I error increases when the researcher conducts
multiple statistical analyses of relationships or differ-
ences; this procedure is referred to as fishing. When
fishing is used, a given portion of the analyses shows
significant relationships or differences simply by
chance. For example, the t-test is commonly used to
make multiple statistical comparisons of mean differ-
ences in a single sample (Kerlinger & Lee, 2000). This
procedure increases the risk of a Type I error because
some of the differences found in the sample occurred
by chance and are not actually present in the popula-
tion. Multivariate statistical techniques have been
developed to deal with this error rate problem
(Goodwin, 1984). Fishing and error rate problems are
discussed in Chapter 21.
Reliability of Measures
The technique of measuring variables must be reliable
to reveal true differences. A measure is a reliable
measure if it gives the same result each time the same
situation or factor is measured. If a scale is used to
measure anxiety, it should give the same score (be
reliable) if repeatedly given to the same person in a
short time (unless, of course, repeatedly taking the
same test causes anxiety to increase or decrease) (Waltz
et al., 2010). Physiological measurement methods that
consistently measure physiological variables are con-
sidered precise (Ryan-Wenger, 2010). For example, a
thermometer would be precise if it showed the same
temperature reading when tested repeatedly on the
same patient within a limited time (see Chapter 16).
Reliability of Intervention Implementation
Intervention reliability ensures that the research
treatment is standardized and applied consistently
each time it is administered in a study. In some studies,
the consistent implementation of the treatment is
referred to as intervention fidelity. Intervention
 CHAPTER 10  Understanding Quantitative Research Design 199
fidelity often includes a protocol to standardize the
elements of the treatment and a plan for training to
ensure consistent implementation of the treatment pro-
tocol (Forbes, 2009; Santacroce, Maccarelli, & Grey,
2004). If the method of administering a research inter-
vention varies from one person to another, the chance
of detecting a true difference decreases. During the
planning and implementation phases, researchers must
ensure that the study intervention is provided in
exactly the same way each time it is administered to
prevent a threat to statistical conclusion design valid-
ity. Chapter 14 provides a detailed discussion of types
of interventions, intervention development, and inter-
vention fidelity.
Random Irrelevancies in the Experimental Setting
Environmental extraneous variables in complex field
settings (e.g., a clinical unit) can influence scores on
the dependent variable. These variables increase the
difficulty of detecting differences. Consider the activi-
ties occurring on a nursing unit. The numbers and
variety of staff, patients, crises, and work patterns
merge into a complex arena for the implementation of
a study. Any of the dynamics of the unit can influence
manipulation of the independent variable or measure-
ment of the dependent variable.
Random Heterogeneity of Respondents
Subjects in a treatment or intervention group can differ
in ways that correlate with the dependent variable, a
situation referred to as random heterogeneity. This
difference can influence the outcome of the interven-
tion and prevent detection of a true relationship
between the independent variable and the dependent
variable. For example, subjects may have a variety of
responses to preoperative interventions to lower
anxiety because of unique characteristics of patients
associated with their differing levels of anxiety.
Internal Validity
Internal validity is the extent to which the effects
detected in the study are a true reflection of reality
rather than the result of extraneous variables. Although
internal validity should be a concern in all studies, it
is addressed more commonly in relation to studies
examining causality than in other studies. When
examining causality, the researcher must determine
whether the independent and dependent variables may
have been affected by a third, often unmeasured, vari-
able (an extraneous variable). Chapter 8 describes
the different types of extraneous variables. The pos-
sibility of an alternative explanation of cause is some-
times referred to as a rival hypothesis (Shadish et al.,
2002). Any study can contain threats to internal design
validity, and these validity threats can lead to false-
positive or false-negative conclusions. The researcher
must ask, “Is there another reasonable (valid) explana-
tion (rival hypothesis) for the finding other than the
one I have proposed?” Threats to internal validity are
described here.
Controlling the Environment
History effect results when an event that is not related
to the planned study occurs during the time of the
study and creates an effect on the study outcome.
History could influence a subject’s response to the
treatment and alter the measurements obtained on the
dependent variables. For example, if you are studying
the effect of an emotional support intervention on sub-
jects’ completion of their cardiac rehabilitation
program and several of the nurses quit their jobs in the
rehabilitation center during your study, this event
could influence the subjects’ rehabilitation program
completion rate in your study.
Maturation
In research, maturation is defined as growing older,
wiser, stronger, hungrier, more tired, or more experi-
enced during the study. Such unplanned and unrecog-
nized changes are a threat to the study internal design
validity and can influence the findings of the study.
Testing Effect
Sometimes, the effect being measured (referred to as
the testing effect) can be due to the number of times
the subject’s responses have been tested. The subject
may remember earlier, inaccurate responses and then
modify them, thus altering the outcome of the study.
The test itself may influence the subject to change
attitudes or may increase the subject’s knowledge.
Instrumentation
Effects can be due to changes in measurement instru-
ments (instrumentation) between the pretest and the
posttest rather than a result of the treatment. For
example, a weight scale that was accurate when the
study began (pretest) could now show subjects to
weigh 2 lbs less than they actually weigh (posttest).
Instrumentation is also involved when people serving
as observers or data collectors become more experi-
enced between the pretest and the posttest, thus alter-
ing in some way the data they collect.
Selection
Selection addresses the process by which subjects are
chosen to take part in a study and how subjects are
200 UNIT TWO  The Research Process
grouped within a study. A selection threat is more
likely to occur in studies in which randomization is
not possible (Kerlinger & Lee, 2000; Thompson
2002). In some studies, people selected for the study
may differ in some important way from people not
selected for the study. In other studies, the threat is
due to differences in subjects selected for study
groups. For example, people assigned to the control
group could be different in some important way from
people assigned to the experimental group. This dif-
ference in selection could cause the two groups to
react differently to the treatment; in this case, the
treatment would not have caused the differences in
group responses.
Subject Attrition
The subject attrition threat is due to subjects
who drop out of a study before completion. Partici-
pants’ attrition becomes a threat when (1) those who
drop out of a study are different types of people
from those who remain in the study or (2) there is
a difference between the kinds of people who drop
out of the experimental group and the people who
drop out of the control or comparison group (see
Chapter 15).
Interactions with Selection
The aforementioned threats can interact with selection
to further complicate the validity of the study. The
threats most likely to interact with selection are history
effect, maturation, and instrumentation. For example,
if the control group you selected for your study has a
different history from that of the experimental group,
responses to the treatment may be due to this differ-
ence rather than to the treatment.
Diffusion or Imitation of Treatments
The control group may gain access to the treatment
intended for the experimental group (diffusion) or a
similar treatment available from another source (imi-
tation). For example, suppose your study examined
the effect of teaching specific information to hyperten-
sive patients as a treatment and then measured the
effect of the teaching on blood pressure readings and
adherence to treatment protocols. Suppose that the
experimental group patients shared the teaching infor-
mation with the control patients (treatment diffu-
sion). This sharing changed the behavior of the control
group. The control group patients’ responses to the
outcome measures may show no differences from
those of the experimental group even though the
teaching actually did make a difference (Type II error;
fail to reject a false null).
Compensatory Equalization of Treatments
When the experimental group receives a treatment
seen as desirable, such as a new treatment for AIDS,
administrative people and other health professionals
may not tolerate the difference and may insist that the
control group also receive the treatment. The researcher
therefore no longer has a control group and cannot
document the effectiveness of the treatment through
the study. In health care, both giving treatment and
withholding treatment have ethical implications.
Resentful Demoralization of Respondents Receiving
Less Desirable Treatments
If control group subjects believe that they are receiv-
ing less desirable treatment, they may withdraw, give
up, or become angry. Changes in behavior resulting
from this reaction rather than from the treatment can
lead to differences that cannot be attributed to the
treatment.
Construct Validity
Construct validity examines the fit between the con-
ceptual definitions and operational definitions of vari-
ables. Theoretical constructs or concepts are defined
within the study framework (conceptual definitions).
These conceptual definitions provide the basis for the
operational definitions of the variables. Operational
definitions (methods of measurement) must validly
reflect the theoretical constructs. (Theoretical con-
structs are discussed in Chapter 7; conceptual and
operational definitions of concepts and variables are
discussed in Chapter 8.)
Is use of the measure a valid inference about the
construct? By examining construct validity, we can
determine whether the instrument actually measures
the theoretical construct it purports to measure. The
process of developing construct validity for an instru-
ment often requires years of scientific work. When
selecting methods of measurement, the researcher
must determine the previous development of instru-
ment construct validity (DeVon et al., 2007; Waltz
et al., 2010). The threats to construct validity are
related both to previous instrument development and
to the development of measurement techniques as part
of the methodology of a particular study. Threats to
construct validity are described here.
Inadequate Preoperational Clarification
of Constructs
Measurement of a construct stems logically from a
concept analysis of the construct, either by the theorist
who developed the construct or by the researcher. The
conceptual definition should emerge from the concept
 CHAPTER 10  Understanding Quantitative Research Design 201
analysis, and the method of measurement (operational
definition) should clearly reflect both. A deficiency in
the conceptual or operational definition leads to low
construct validity.
Mono-Operation Bias
Mono-operation bias occurs when only one method
of measurement is used to assess a construct. When
only one method of measurement is used, fewer
dimensions of the construct are measured. Construct
validity greatly improves if the researcher uses more
than one instrument (Waltz et al., 2010). For example,
if anxiety were a dependent variable, more than one
measure of anxiety could be used. It is often possible
to apply more than one measurement of the dependent
variable with little increase in time, effort, or cost.
Monomethod Bias
In monomethod bias, the researcher uses more than
one measure of a variable, but all the measures use the
same method of recording. Attitude measures, for
example, may all be paper and pencil scales. Attitudes
that are personal and private, however, may not be
detected through the use of paper and pencil tools.
Paper and pencil tools may be influenced by feelings
of nonaccountability for responses, acquiescence, or
social desirability. For example, construct validity
would improve if anxiety were measured by a paper
and pencil test, verbal messages of anxiety, the gal-
vanic skin response, and the observer’s recording of
incidence and frequency of behaviors that have been
validly linked with anxiety.
Hypothesis Guessing within
Experimental Conditions
Hypothesis guessing occurs when subjects within a
study guess the hypotheses of the researcher. The
validity concern relates to behavioral changes that
may occur in the subjects as a consequence of knowing
the hypothesis. The extent to which this issue modifies
study findings is not currently known.
Evaluation Apprehension
Subjects want researchers to see them in a favorable
light. They want to be seen as competent and psycho-
logically healthy. Evaluation apprehension occurs
when the subject’s responses in the experiment are due
to this desire rather than the effects of the independent
variable.
Experimenter Expectancies (Rosenthal Effect)
The expectancies of the researcher can bias the
data. For example, experimenter expectancy occurs
if a researcher expects a particular intervention to
relieve pain. The data he or she collects may be
biased to reflect this expectation. If another researcher
who does not believe the intervention would be
effective had collected the data, results could have
been different. The extent to which this effect actu-
ally influences studies is not known. Because of
their concern about experimenter expectancy, some
researchers are not involved in the data collection
process. In other studies, data collectors do not know
which subjects are assigned to treatment and control
groups.
Another way to control this threat is to design the
study so that the various data collectors have different
expectations. If the sample size is large enough, the
researcher could compare data gathered by the differ-
ent data collectors. Failing to determine differences in
the data collected by the collectors would give evi-
dence that the construct is valid.
Confounding Constructs and Levels of Constructs
When developing the methodology of a study, you
must decide about the intensity of the variable that
will be measured or provided as a treatment. This
intensity influences the level of the construct that
will be reflected in the study. These decisions can
affect validity, because the method of measuring the
variable influences the outcome of the study and
the understanding of the constructs in the study
framework.
For example, in reviewing your research, you
might find that variable A does not affect variable B
when, in fact, it does, but either not at the level of A
that was manipulated or not at the level of B that was
measured. This issue is a particular problem when A
is not linearly related to B or when the effect being
studied is weak. To control this threat, you will need
to include several levels of A in the design and will
have to measure many levels of B. For example, in a
study in which A is preoperative teaching and B is
anxiety, (1) the instrument being used to measure
anxiety measures only high levels of anxiety or (2) the
preoperative teaching is provided for 15 minutes but
30 minutes or an hour of teaching is required to cause
significant changes in anxiety.
In some cases, confounding of variables occurs,
leading to mistaken conclusions. Few measures of a
construct are pure measures. Rather, a selected method
of measuring a construct can measure a portion of the
construct as well as other related constructs. Thus,
the measure can lead to confusing results, because
the variable measured does not accurately reflect the
construct.
202 UNIT TWO  The Research Process
Interaction of Different Treatments
The interaction of different treatments is a threat to
construct validity if subjects receive more than one
treatment in a study. For example, your study might
examine the effectiveness of pain relief measures, and
subjects might receive medication, massage, distrac-
tion, and relaxation strategies. In this case, each one
of the treatments interacts with the others, and the
effect of any single treatment on pain relief would be
impossible to extract. Your study findings could not
be generalized to any situation in which patients did
not receive all four pain treatments. Chapter 14 pro-
vides direction in preventing and managing interac-
tions of treatments.
Interaction of Testing and Treatment
In some studies, pretesting the subject is thought to
modify the effect of the treatment. In this case, the
findings can be generalized only to subjects who have
been pretested. Although there is some evidence that
pretest sensitivity does not have the impact that was
once feared, it must be considered when the validity
of the study is examined. The Solomon four-group
design (discussed in Chapter 11) tests this threat to
validity. Repeated posttests can also lead to an interac-
tion of testing and treatment.
External Validity
External validity is concerned with the extent to
which study findings can be generalized beyond the
sample used in the study. With the most serious threat,
the findings would be meaningful only for the group
being studied. To some extent, the significance of the
study depends on the number of types of people and
situations to which the findings can be applied. Some-
times, the factors influencing external validity are
subtle and may not be reported in research reports;
however, the researcher must be responsible for these
factors. Generalization is usually narrower for a single
study than for multiple replications of a study using
different samples, perhaps from different populations
in different settings. The threats to the ability to gen-
eralize the findings (external validity) in terms of
study design are described here.
Interaction of Selection and Treatment
Seeking subjects who are willing to participate in a
study can be difficult, particularly if the study requires
extensive amounts of time or other investments by
subjects. If a large number of the persons approached
to participate in a study decline to participate, the
sample actually selected will be limited in ways that
might not be evident at first glance. For example, the
researcher knows the subjects well and probably why
they consented to participate in the study. Subjects
might be volunteers, “do-gooders,” or people with
nothing better to do. In this case, generalizing the
findings to all members of a population, such as all
nurses, all hospitalized patients, or all persons experi-
encing diabetes, is not easy to justify.
The study must be planned to limit the investment
demands on subjects and thereby improve participa-
tion. The researcher must report the number of persons
who were approached and refused to participate in the
study (refusal rate) so that those who are examining
the study can judge any threats to external validity. As
the percentage of those who decline to participate
increases, external design validity decreases. Suffi-
cient data need to be collected on the subjects to allow
the researcher to be familiar with the characteristics
of subjects and, to the extent possible, the character-
istics of those who decline to participate.
Interaction of Setting and Treatment
Bias exists in types of settings and organizations that
agree to participate in studies. This bias has been par-
ticularly evident in nursing studies. For example,
some hospitals welcome nursing studies and encour-
age employed nurses to conduct studies. Others are
resistant to the conduct of nursing research. These two
types of hospitals may be different in important ways;
thus, there might be an interaction of setting and treat-
ment that limits the generalizability of the findings. As
a researcher, you must consider this factor when
making statements about the population to which your
findings can be generalized.
Interaction of History and Treatment
The circumstances in which a study was conducted
(history) influence the treatment and thus the general-
izability of the findings. Logically, one can never gen-
eralize to the future; however, replicating the study
during various periods strengthens the usefulness of
findings over time. In critically appraising studies, you
must always consider the period of history during
which the study was conducted and the effect of
nursing practice and societal events during that period
on the reported findings (see Chapter 14 for more
details on research with interventions).
Elements of a Good Design
The purpose of design is to set up a situation that
maximizes the possibilities of obtaining accurate
responses to objectives, questions, or hypotheses.
Select a design that is (1) appropriate to the purpose
of the study; (2) feasible given realistic constraints;
and (3) effective in reducing threats to design validity.
 CHAPTER 10  Understanding Quantitative Research Design 203
In most studies, comparisons are the basis of obtaining
valid answers. A good design provides the subjects,
the setting, and the protocol within which those com-
parisons can be clearly examined. The comparisons
may focus on differences or relationships or both. The
study may require that comparisons be made between
or among individuals, groups, or variables. A compari-
son may also be made of measures taken before a
treatment (pretest) and measures taken after a treat-
ment (posttest). After these comparisons have been
made, you can compare the sample values with statis-
tical tables reflecting population values. In some cases,
the study may involve comparing group values with
population values.
Designs were developed to reduce threats that
might invalidate the comparisons. However, some
designs are more effective in reducing threats than
others. It may be necessary to modify the design to
reduce a particular threat. Before selecting a design,
you must identify the design validity threats that are
most likely to invalidate your study.
Strategies for reducing threats to design validity are
sometimes addressed in terms of control. Selecting a
design involves decisions related to control of the
environment, sample, treatment, and measurement.
Increasing control (to reduce threats to validity) will
require you to carefully think through every facet of
your design. An excellent description of one research
team’s efforts to develop a good design and control
threats to validity is offered by McGuire et al. (2000)
in their study, “Maintaining Study Validity in a Chang-
ing Clinical Environment.”
Controlling the Environment
The study environment has a major effect on research
outcomes. An uncontrolled environment introduces
many extraneous variables into the study situation.
Therefore, the study design may include strategies for
controlling that environment. In many studies, it is
important that the environment be consistent for all
subjects. Elements in the environment that may influ-
ence the application of a treatment or the measurement
of variables must be identified and, when possible,
controlled.
Controlling Equivalence of Subjects
and Groups
When comparisons are made, it is assumed that the
individual units of the comparison are relatively
equivalent except for the variables being measured.
The researcher does not want to be comparing “apples
and oranges.” To establish equivalence, the researcher
defines sampling criteria. Deviation from this equiva-
lence is a threat to internal design validity. Deviation
occurs when sampling criteria have not been ade-
quately defined or when unidentified extraneous vari-
ables increase variation in the group.
The most effective strategy for achieving equiva-
lence consists of random sampling followed by
random assignment to groups. However, this strategy
does not guarantee equivalence. Even when random-
ization has been used, the researcher must examine the
extent of equivalence by measuring and comparing
characteristics for which the groups must be equiva-
lent. This comparison is usually reported in the
description of the sample (Shadish et al., 2002).
Contrary to the aforementioned need for equiva-
lence, groups must be as different as possible in rela-
tion to the research variables. Small differences or
relationships are more difficult to distinguish than
large differences. These differences are often addressed
in terms of effect size. Although sample size plays an
important role, effect size is maximized by a good
design. Effect size is greatest when variance within
groups is small.
Control and Comparison Groups
If the study involves an experimental treatment, the
design usually calls for a comparison. Outcome mea-
sures for individuals who receive the experimental
treatment are compared with outcome measures for
those who do not receive the experimental treatment.
This comparison requires a control group, subjects
who do not receive the experimental treatment.
However, in nursing studies, all patients require care
and those who do not receive the study intervention
receive standard or usual care. Nurse researchers often
refer to the group receiving standard care, but no treat-
ment, as the comparison group rather than the control
group.
One threat to validity is the lack of equivalence
between the experimental and control groups. This
threat is best controlled by random assignment to
groups. Another strategy is for the subjects to serve as
their own controls. With this design strategy, pretest
and posttest measures are taken of the subjects in the
absence of a treatment, as well as before and after
the treatment. In this case, the timing of measures
must be comparable between control and treatment
conditions.
Controlling the Treatment
In a well-designed experimental study, the researcher
has complete control of any treatment provided. The
first step in achieving control is to develop a detailed
description of the treatment, such as an intervention
protocol, to ensure standardization of the treatment.
The next step is to use strategies to ensure consistency
204 UNIT TWO  The Research Process
in implementing the treatment. Consistency may
involve elements of the treatment such as equipment,
time, intensity, sequencing, and staff skill. This process
is referred to as intervention fidelity and is discussed
in more detail in Chapter 14.
Variations in the treatment reduce the effect size. It
is likely that subjects who receive fewer optimal
applications of the treatment will have a smaller
response, resulting in more variance in posttest mea-
sures for the experimental group. To avoid this
problem, the treatment is administered to each subject
in exactly the same way. This consideration requires
the researcher to think carefully through every
element of the treatment to reduce variation wherever
possible (Morrison et al., 2009; Santacroce et al.,
2004). For example, if information is being provided
as part of the treatment, some researchers record the
information, present it to each subject in the same
environment, and attempt to decrease variation in the
subject’s experience before and during the viewing of
the DVD content. Variations include elements such as
time of day, mood, anxiety, experience of pain, inter-
actions with others, and amount of time spent waiting.
Yamada, Stevens, Sidani, Watt-Watson, and Silva
(2010) provide a detailed discussion of the process
they used to measure intervention implementation
fidelity in their study.
In many nursing studies, the researcher does not
have complete control of the treatment. It may be
costly to control the treatment carefully; it may be
difficult to persuade staff to be consistent in the treat-
ment, or the time required to implement a carefully
controlled treatment may seem prohibitive. In some
cases, the researcher may be studying causal outcomes
of an event occurring naturally in the environment.
Regardless of the reason for the researcher’s decision,
internal design validity is reduced when the treatment
is inconsistently applied. The risk of a Type II error is
higher owing to greater variance and a smaller effect
size. Thus, studies with uncontrolled treatments need
larger samples to reduce the risk of a Type II error.
External validity may improve if the treatment is
studied as it typically occurs clinically. If the study
does not reveal a statistically significant difference,
then perhaps the typical clinical application of the
treatment does not have an important effect on patient
outcomes. The question then becomes whether a dif-
ference might have been found if the treatment had
been consistently applied.
Counterbalancing
In some studies, each subject receives several different
treatments sequentially (e.g., relaxation, distraction,
and visual imagery) or various levels of the same treat-
ment (e.g., different doses of a drug or varying lengths
of relaxation time). Sometimes the application of one
treatment can influence the response to later treat-
ments, a phenomenon referred to as a carryover
effect. If a carryover effect is known to occur, it is not
advisable for a researcher to use this design strategy
for the study. However, even when no carryover effect
is known, the researcher may take precautions against
the possibility that this effect will influence outcomes.
In one such precaution, known as counterbalancing,
the various treatments are administered in random
order rather than being provided consistently in the
same sequence.
Controlling Measurement
Measurement methods play a key role in the validity
of a study. Instruments such as scales must have docu-
mented validity and reliability, and physiological mea-
sures require accuracy and precision (Ryan-Wenger,
2010; Waltz et al., 2010). When measurement is crude
or inconsistent, variance within groups is high, and it
is more difficult to detect differences or relationships
among groups. Thus, the study does not provide a
valid test of the hypotheses. However, the consistent
implementation of measurements enhances validity.
For example, each subject must receive the same
instructions about completing a pain scale. Data col-
lectors must be trained and observed for consistency.
Designs define the timing of measures (e.g., pretest,
posttest). Sometimes, the design calls for multiple
measures over time. The researcher must specify the
points in time during which measures will be taken.
The research report must include a rationale for the
timing of measures.
Controlling Extraneous Variables
When designing a study, you must identify variables
not included in the design (extraneous variables) that
could explain some of the variance that occurs when
the study variables are measured. In a good design,
the effect of these variables on variance is controlled.
The extraneous variables commonly encountered in
nursing studies are age, education, gender, social
class, severity of illness, level of health, functional
status, and attitudes. For a specific study, you must
think carefully through the variables that could have
an impact on that study.
Design strategies used to control extraneous vari-
ables include random sampling, random assignment
to groups, selecting subjects who are homogeneous
in terms of a particular extraneous variable, selecting
a heterogeneous sample, blocking, stratification,
 CHAPTER 10  Understanding Quantitative Research Design 205

TABLE 10-1  Studies Using Control Strategies for Good Design

Design Strategy Example Studies
Control group McCorkle, R., Jeon, S., Ercolano, E., & Schwartz, P. (2011). Healthcare utilization in women after
abdominal surgery for ovarian cancer. Nursing Research, 60(1), 47–57.
Chen, K., Fan, J., Wang, H., Wu, S., Li, C., & Lin, H. (2010). Silver yoga exercises improved physical
fitness of transitional frail elders. Nursing Research, 59(5), 364–370.
Counterbalancing Cacciola, J. S., Alterman, A. I., McLellan, A. T., Lin, Y., & Lynch, K. G. (2007). Initial evidence for the
reliability and validity of a “Lite” version of the Addiction Severity Index. Drug and Alcohol
Dependence, 87(2–3), 297–302.
Ivarsson, B., Larsson, S., Lührs, C., & Sjöberg, T. (2007). Patients’ perceptions of information about risks
at cardiac surgery. Patient Education and Counseling, 67(1–2), 32–38.
Random sampling Laschinger, H. K. S., Finegan, J., & Wilk, P. (2011). Situational and dispositional influences on nurses’
workplace well-being: The role of empowering unit leadership. Nursing Research, 60(2), 124–131.
Simons, S. R., Stark, R. B., & DeMarco, R. F. (2011). A new, four-item instrument to measure workplace
bullying. Research in Nursing & Health, 34(2), 132–140.
Random assignment Jones, D., Duffy, M. E., & Flanagan, J. (2011). Randomized clinical trial testing efficacy of a nurse-
coached intervention in arthroscopy patients. Nursing Research, 60(2), 92–99.
Lee, K. A., & Gay, C. L. (2011). Can modifications to the bedroom environment improve sleep of new
parents? Two randomized controlled trials. Research in Nursing & Health, 34(1), 7–19.
Homogeneity Hodgin, R. F., Chandra, A., & Weaver, C. (2010). Correlates to long-term-care nurse turnover: Survey
results from the State of West Virginia. Hospital Topics, 88(4), 91–97.
Estok, P. J., Sedlak, C. A., Doheny, M. O., & Hall, R. (2007). Structural model for osteoporosis preventing
behavior in postmenopausal women. Nursing Research, 56(3), 148–158.
Heterogeneity Kwok, C. S., Loke, Y. K., Hale, R., Potter, J. F., & Myint, P. K. (2011). Atrial fibrillation and incidence of
dementia: A systematic review and meta-analysis. Neurology, 76(10), 914–922.
Neufeld, A., & Harrison, M. J. (2003). Unfulfilled expectations and negative interactions: Nonsupport in
the relationships of women caregivers. Journal of Advanced Nursing, 41(4), 323–331.
Blocking Rousaud, A., Blanch, J., Hautzinger, M., De Lazzari, E., Peri, J. M., Puig, O., et al. (2007). Improvement of
psychosocial adjustment to HIV-1 infection through a cognitive-behavioral oriented group psychotherapy
program: A pilot study. AIDS Patient Care and STDs, 21(3), 212–222.
Tsay, S. L., Wang, J. C., Lin, K. C., & Chung, U. L. (2005). Effects of acupressure therapy for patients
having prolonged mechanical ventilation support. Journal of Advanced Nursing, 52(2), 142–150.
Stratification Botticello, A. L., Chen, Y., Cao, Y., & Tulsky, D. S. (2011). Do communities matter after rehabilitation?
The effect of socioeconomic and urban stratification on well-being after spinal cord injury. Archives of
Physical Medicine & Rehabilitation, 92(3), 464–471.
Carey, T. A. (2006). Estimating treatment duration for psychotherapy in primary care. Journal of Public
Mental Health, 5(3), 23–28.
Matching Mehta, S., Chen, H., Johnson, M. L., & Aparasu, R. R. (2010). Risk of falls and fractures in older adults
using antipsychotic agents: A propensity-matched retrospective cohort study. Drugs & Aging, 27(10),
815–829.
Trevisanuto, D., Micaglio, M., Pitton, M., Magarotto, M., Piva, D., & Zanardo, V. (2006). Laryngeal mask
airway: Is the management of neonates requiring positive pressure ventilation at birth changing? Journal
of Neonatal Nursing, 12(5), 185–192.
Statistical control Griffin-Blake, C. S., & DeJoy, D. M. (2006). Evaluation of social-cognitive versus stage-matched, self-help
(partialing out) physical activity interventions at the workplace. American Journal of Health Promotion, 20(3), 200–209.
Roberts, J. E., Burchinal, M. R., & Zeisel, S. A. (2002). Otitis media in early childhood in relation to
children’s school-age language and academic skills. Pediatrics, 110(4), 696–706.

matching subjects between groups in relation to Random Sampling

a particular variable, and statistical control. Table Random sampling increases the probability that
10-1 identifies some nursing studies and the various subjects with various levels of an extraneous variable
strategies they have used to control extraneous are included and are randomly dispersed throughout
variables. the groups within the study (Thompson, 2002).
206 UNIT TWO  The Research Process

This strategy is particularly important for controlling
uni­dentified extraneous variables. Whenever possible,
however, extraneous variables must be identified,
measured, and reported in the description of the
sample.
Random Assignment
Random assignment enhances the probability that
subjects with various levels of extraneous variables
are equally dispersed in treatment and control or com-
parison groups. When subjects are randomly assigned
to groups, these groups are considered independent.
Independent groups exist when the selection and
assignment of subjects to one group, such as the
experimental group, are unrelated to the subjects
selected and assigned to the control group. Whenever
possible, however, this dispersion must be evaluated
rather than assumed.
Homogeneity
Homogeneity is a more extreme form of equiva-
lence in which the researcher limits the subjects to
only one level of an extraneous variable to reduce its
impact on the study findings. To use this strategy,
you must have previously identified the extraneous
variables. You might choose to include subjects with
only one level of an extraneous variable in the study.
For example, only subjects between the ages of 20
and 30 years may be included in a study, or only
subjects with a particular level of education. The
study may include only breast cancer patients who
have been diagnosed within 1 month, are at a par-
ticular stage of disease, and are receiving a specific
treatment for cancer. The difficulty with this strategy
is that it limits generalization to the types of subjects
included in the study. Findings could not justifiably
be generalized to types of people excluded from the
study.
Heterogeneity
In studies using nonrandom sampling methods, the
researcher may attempt to obtain subjects with a wide
variety of characteristics (or who are heterogeneous)
to reduce the risk of biases. When using the strategy
of heterogeneity, you may seek subjects from multiple
diverse sources. The strategy is designed to increase
generalizability of the study findings. Characteristics
of the sample must be described in the research report
to indicate the heterogeneity of the sample.
Blocking
In blocking, the researcher includes subjects with
various levels of an extraneous variable in the sample
but controls the numbers of subjects at each level of
the variable and their random assignment to groups
within the study. Designs using blocking are referred
to as randomized block designs (see Chapter 11).
The extraneous variable is then used as an independent
variable in the data analysis. Therefore, the extraneous
variable must be included in the framework and the
study hypotheses.
Using this strategy, you might randomly assign
equal numbers of subjects in three age categories
(younger than 18 years, 18 to 60 years of age, and
older than 60 years) to each group in the study. You
could use blocking for several extraneous variables.
For example, you could block the study in relation to
both age and ethnic background (African American,
Hispanic, Caucasian, and Asian). Table 10-2 summa-
rizes an example of this approach.
During data analysis for the randomized block
design, each cell in the analysis is treated as a group.

TABLE 10-2  Example of Blocking Using Age and Ethnic Background

Age Ethnic Group Experimental Control
Younger than 18 years (n = 160) African American n = 40 n = 20 n = 20
Hispanic, nonwhite n = 40 n = 20 n = 20
White, non-Hispanic n = 40 n = 20 n = 20
Asian n = 40 n = 20 n = 20
19 to 60 years of age (n = 160) African American n = 40 n = 20 n = 20
Hispanic, nonwhite n = 40 n = 20 n = 20
White, non-Hispanic n = 40 n = 20 n = 20
Asian n = 40 n = 20 n = 20
Older than 60 years (n = 160) African American n = 40 n = 20 n = 20
Hispanic, nonwhite n = 40 n = 20 n = 20
White, non-Hispanic n = 40 n = 20 n = 20
Asian n = 40 n = 20 n = 20
 CHAPTER 10  Understanding Quantitative Research Design 207
Therefore, you must evaluate the cell sample size and
the effect size to ensure adequate power to detect dif-
ferences in the study. A minimum of 20 subjects per
group is recommended but the final group size is best
determined by power analysis (Aberson, 2010). Thus,
the example described for Table 10-2 would require a
minimal sample of 480 subjects. Using randomized
block designs usually require very large sample sizes
to implement.
Stratification
Stratification involves the distribution of subjects
throughout the sample, using sampling techniques
similar to those used in blocking, but the purpose of
the procedure is even distribution throughout the
sample. The extraneous variable is not included in the
data analysis. Distribution of the extraneous variable
is included in the description of the sample.
Matching
To ensure that subjects in the control or comparison
group are equivalent to subjects in the experimental
group, some studies are designed to match subjects in
the two groups. Matching is used when a subject in
the experimental group is randomly selected and then
a subject similar in relation to important extraneous
variables is randomly selected for the control group.
This matching of subjects on selected characteristics
to be included in both the experimental and control
groups results in dependent groups. For example,
subjects in the experimental and control groups might
be matched for age, gender, severity of illness, or
number of chronic illnesses. Clearly, the pool of avail-
able subjects would have to be large to accomplish this
goal. In quasi-experimental studies, matching may be
performed without randomization (Cook & Campbell,
1979).
Statistical Control
In some studies, it is not considered feasible to control
extraneous variables through the design. However, the
researcher recognizes the possible impact of extrane-
ous variables on variance and effect size. Therefore,
measures are obtained for the identified extraneous
variables. Data analysis strategies that have the capac-
ity to remove (partial out) the variance explained by
the extraneous variable are performed before the anal-
ysis of differences or relationships between or among
the variables of interest in the study. One statistical
procedure commonly used for this purpose is analysis
of covariance (Corty, 2007). Although statistical
control seems to be a quick and easy solution to
the problem of extraneous variables, its results are not
as satisfactory as those of the various methods of
design control.
Questions to Direct Design
Development and
Implementation in a Study
Developing and implementing a study design requires
the researcher to consider multiple details such as
those discussed in the sections on design validity and
elements of a good design. The more carefully thought
out these details are, the stronger the design. Strong
research designs are essential to generate valid research
evidence for nursing (Brown, 2009; Melnyk &
Fineout-Overholt, 2011). The elements central to the
study design include the presence or absence of a
treatment, the number of groups in the sample, the
number and timing of measurements, the sampling
method, the time frame for data collection, planned
comparisons, and the control of extraneous variables.
Finding answers to the following questions will help
you to develop a study design:
1. Is the primary purpose of the study to describe
variables and groups within the study situation, to
examine relationships, or to examine causality
within the study situation? (Kerlinger & Lee,
2000; Shadish et al., 2002)
2. Will a treatment or intervention be implemented
in the study? (Forbes, 2009)
3. If an intervention is implemented, will the re-
searcher control the intervention? (Forbes, 2009;
Morrison et al., 2009; Santacroce et al., 2004)
4. Will the sample be pretested before the
intervention?
5. Will the sample be randomly or nonrandomly
selected?
6. What sampling method is used to obtain study
participants? (Thompson, 2002)
7. Will the sample be studied as a single group or
divided into groups? (Fawcett & Garity, 2009)
8. How many groups will there be?
9. What will be the size of each group? (Aberson,
2010)
10. Will there be a control group, or will a comparison
group or standard care group be compared with
the experimental group?
11. Will the study participants be randomly assigned
to the groups? If the participants are randomly
assigned, how is this assignment accomplished?
12. What instruments will be used to measure the
variables? (Bialocerkowski, Klupp, & Bragge,
2010; Waltz et al., 2010)
208 UNIT TWO  The Research Process
13. Are the measurement methods valid and reliable
or precise and accurate? (DeVon et al., 2007)
14. Will the variables be measured more than once?
15. Will the data be collected cross-sectionally or
over time?
16. Have extraneous variables been identified?
17. Are data being collected on extraneous variables?
18. What strategies are being used to control for
extraneous variables?
19. What strategies are being used to compare vari-
ables or groups?
20. Will data be collected at a single site or at multiple
sites?
21. What strategies are used to ensure consistent col-
lection of data? (Creswell, 2009; Fawcett &
Garity, 2009; Shadish et al., 2002)
Mixed Methods
There is controversy among researchers about the
relative validity of various approaches to research.
Designing quantitative experimental studies with rig-
orous controls may provide strong external validity
but questionable or limited internal validity. Qualita-
tive studies may have strong internal validity but ques-
tionable external validity. A single approach to
measuring a concept may be inadequate to justify a
claim that it is a valid measure of a theoretical concept.
Testing a single theory may leave the results open to
the challenge of rival hypotheses from other theories
(Creswell, 2009).
As research methodologies continue to evolve,
mixed-methods approaches offer investigators the
ability to utilize the strengths of both qualitative and
quantitative research designs. Mixed-methods research
is characterized as research that contains elements of
both qualitative and quantitative approaches (Coward,
1990; Creswell, 2009; Duffy, 1987; Marshall &
Rossman, 2011; Mitchell, 1986; Morse, 1991; Myers
& Haase, 1989; Patton, 2002; Porter, 1989). The phil-
osophical underpinnings of mixed-methods research
and what paradigms best fit these research methods
are still evolving. It is recognized that all researchers
bring assumptions to their studies, consciously or
unconsciously, and investigators decide whether they
are going to view their study from a post-positivist
(quantitative) or constructivist (qualitative) perspec-
tive, or through an “advocacy lens,” such as feminism
(Fawcett & Garity, 2009; Munhall, 2012).
Over the last few years, many researchers have
departed from the idea that one paradigm or one
research strategy is right and have taken the perspec-
tive that the search for the truth requires the use of all
available strategies. To capitalize on the representa-
tiveness and generalizability of quantitative research
and the in-depth, contextual nature of qualitative
research, several methods are combined in a single
research study (Creswell, 2009).
The idea of using mixed-methods approaches to
conduct studies has a long history. More than 50 years
ago, quantitative researchers Campbell and Fiske
(1959) recommended mixed methods to more accu-
rately measure a psychological trait. The multi-trait–
multi-method matrix was designed to rule out method
effects to allow researchers to attribute individual
variation to the personality trait itself (Rocco, Bliss,
Gallagher, & Perez-Prado, 2003). This mixed method-
ology was later expanded into what Denzin (1989)
identified as “triangulation.” Denzin (1989) believed
that combining multiple theories, methods, observers,
and data sources can help researchers overcome the
intrinsic bias that comes from single-theory, single-
methods, and single-observer studies. Triangulation
evolved to include using multiple data collection and
analysis methods, multiple data sources, multiple ana-
lysts, and multiple theories or perspectives (Patton,
2002). The concept of triangulation has now been
replaced by the idea of mixed-methods approaches
(Creswell, 2009).
Because phenomena are complex, combining qual-
itative and quantitative methods enables researchers to
be more likely to capture the essence of the phenom-
enon. There are selected techniques or strategies
associated with conducting mixed-methods studies.
Strategies frequently associated with the mixed-
methods approaches include the following:
Sequential procedures are those in which the
researcher investigates a phenomenon so that find-
ings of one method may elaborate on or elucidate
the findings of another method. This may involve
beginning with a qualitative method to explore and
then following up with a quantitative method using
a large sample so that results can be generalized to
a population. Alternatively, the study may begin
with a quantitative method so that theories or prop-
ositions can be tested, followed by a qualitative
method using in-depth interviews with study par-
ticipants to expand a theory (Creswell, 2009).
Concurrent procedures is an approach in which the
researcher merges quantitative and qualitative data
in order to present an all-inclusive analysis of the
research problem. Collection of quantitative and
qualitative data is done concurrently during the
study. Then the investigator assimilates the results
obtained from both methods in order to interpret
the overall findings. The researcher may nest one
 CHAPTER 10  Understanding Quantitative Research Design 209
form of data within another larger data collection
to analyze different questions (Creswell, 2009).
Transformative procedure is one in which the
researcher uses a “theoretical lens” in order to
encompass a broad perspective within a design that
contains both quantitative and qualitative data. This
broad perspective is based on a theoretical frame-
work and provides methods for data collection and
anticipated outcomes from the study. This method
usually serves a larger purpose to advocate for
minority or marginalized groups. Data could be
sequential or concurrent (Creswell, 2009).
In each of these three strategies, both quantitative
and qualitative data are collected and the researcher
may integrate the data at different stages of the
research process. The priority may be given to one
type of data over the other (i.e., quantitative over
qualitative or vice versa), or they can be equal (con-
current). If the sequential approach is implemented
and qualitative data are collected first, then qualitative
data are the priority, and integration of the two data
sets occurs during data interpretation and is then expli-
cated in the discussion. If quantitative data are col-
lected first, then these data are the priority, and
qualitative data are used to augment the quantitative
data. Data integration is completed during analysis
and interpretation and is expressed in the discussion.
With the concurrent approach, data are considered
equal and therefore are collected at the same time, or
concurrently. Integration of the data occurs during
data collection.
Creswell (2009) identifies six types of mixed-
methods approaches, expanding the three strategies
previously introduced. The four approaches usually
implemented in nursing research are: (1) sequential
explanatory strategy; (2) sequential exploratory
Quantitative Qualitative
Data Data Data
Collection Analysis Collection
Interpretation of entire analysis
Purpose is to use qualitative results to assist in explaining and
interpreting the findings of a primarily quantitative study.
Easy to implement, describe and report.
Weakness is the length of time involved in data collection, with the
two separate phases.
strategy; (3) sequential transformative strategy; and
(4) concurrent triangulation strategy. Models of these
mixed-methods approaches and examples are pro-
vided to expand your understanding of these designs.
Sequential Explanatory Strategy
With the sequential explanatory strategy, the
researcher collects and analyzes quantitative data and
then collects and analyzes qualitative data. Integration
of the data occurs during the interpretation phase. The
purpose of this approach is to assist in explaining and
interpreting quantitative data (see Figure 10-1). It is
useful when unexpected quantitative results are
revealed. Qualitative examination of the phenomenon
facilitates a fuller understanding and is well suited to
explaining and interpreting relationships. There may
or may not be a theoretical perspective to the study.
This approach is easy to implement because the steps
fall in sequential stages; however the two-stage
approach extends the time involved in data collection
and is seen as a weakness of the design.
This type of methodology was used in a study by
Carr (2009), who examined the experience of postop-
erative pain. Women undergoing surgery completed
questionnaires to measure pain, anxiety, and depres-
sion (quantitative data). Follow-up telephone inter-
views explored their pain experiences (qualitative
data). Using a second series of patients, Carr once
again looked at the frequency and patterns of anxiety
in the immediate preoperative and postoperative
periods, followed by semi-structured telephone inter-
views. During the interviews, patients identified
events/situations occurring during their hospitaliza-
tion that contributed to their anxiety.
As indicated by the use of sequential explanatory
strategy, priority was given to the quantitative data.
Data
Analysis Figure 10-1  Sequential explanatory strategy. (Adapted
from Creswell, J. W. [2009]. Research design: Qualitative,
quantitative, and mixed methods approaches [3rd ed.].
Los Angeles, CA: Sage; and Hanson, W. E., Creswell, J. W.,
Plano Clark, V. L., Petska, K. S., & Creswell, J. D. [2005].
Mixed methods research designs in counseling
psychology. Journal of Counseling Psychology, 52[2],
224–235.)
210 UNIT TWO  The Research Process
However, during interpretation, the qualitative data
seemed to indicate that lack of knowledge about the
operative procedure, which is traditionally thought to
be a contributor to anxiety, was not as contributory to
patient anxiety as expected. By interpreting both quan-
titative and qualitative data in the same study, the
investigator was able to broaden the understanding
of the pain and anxiety of postoperative patients.
Carr (2009) also noted that some of the events that
led to patients’ anxiety and pain were amenable to
nursing interventions and not solely due to lack of
knowledge.
Sequential Exploratory Strategy
Sequential exploratory strategy is very similar to the
sequential explanatory strategy except the collection
and analysis of qualitative data precedes the collection
of quantitative data. Integration of the data occurs
during the interpretation phase, and the quantitative
data are used to understand the qualitative data (see
Figure 10-2). There may or may not be a theoretical
perspective to the study. The purpose of this approach
is to: (1) explore a phenomenon’s distribution within
a population; (2) test elements of an emerging theory;
or (3) develop and test new measurement instruments.
The sequential exploratory strategy is also easy to
implement because of the staged approach, but the
length of time required for data collection is consid-
ered a weakness.
An example of sequential exploratory strategy is
the research by Choi and Harwood (2004), who con-
ducted a qualitative study to explore the themes of
Korean women’s response to domestic violence. Five
themes emerged centered on intolerance to abusive
husbands and were the basis for the development of a
Figure 10-2  Sequential exploratory strategy. (Adapted
from Creswell, J. W. [2009]. Research design: Qualitative,
quantitative, and mixed methods approaches [3rd ed.].
Los Angeles, CA: Sage; and Hanson, W. E., Creswell, J. W.,
Plano Clark, V. L., Petska, K. S., & Creswell, J. D. [2005].
Mixed methods research designs in counseling
psychology. Journal of Counseling Psychology, 52[2],
224–235.)
quantitative scale to explore each item. A pilot study
was conducted to examine the women’s intolerance to
physical abuse by their husbands, and a third phase of
the study was conducted to establish construct validity
of the scale (Choi, Phillips, Figueredo, Insel, & Min,
2008).
Sequential Transformative Strategy
When the sequential transformative strategy is
employed, qualitative or quantitative data collection
and analysis can come first. The results are integrated
during the interpretation phase. But unlike the prior
two approaches, sequential transformative strategy is
guided by a theory or through the advocacy lens. The
purpose of this approach is to employ methods that
will best serve the theoretical perspective. The theo-
retical perspective drives the entire research process
from the introduction of the problem to the directional
research question, which generates a sensitive
approach to data collection and ends with an appeal
to act (see Figure 10-3). Using this strategy, research-
ers may have the opportunity to speak out about dif-
ferent perspectives, serve as an advocate for study
participants, or better understand a phenomenon that
is changing as a result of being studied. The staging
inherent in the design is a strength, but once again,
this approach extends the length of time required to
collect data. There is little written on this approach, so
there is little guidance on how to use the “transforma-
tive vision” to guide the methods.
The sequential transformative strategy was used by
Park, Knapp, and Shin (2009) in the conduct of their
mixed-methods study of social engagement in assisted
living communities. These researchers used a frame-
work of social relationships and support among older
Qualitative Quantitative
Data Data Data Data
Collection Analysis Collection Analysis
Interpretation of entire analysis
Purpose is to use quantitative data and results to assist in the
interpretation of qualitative findings.
Primary focus of this model is to explore a phenomenon.
Weakness is the length of time involved in data collection, with the two
separate phases.
 CHAPTER 10  Understanding Quantitative Research Design 211
Quantitative Qualitative
Vision Advocate Ideas
Purpose is to employ methods that will best serve the theoretical
perspective of the researcher.
Caters to diverse perspectives and helps understand a phenomenon
or process changing as a result of being studied.
adults to guide the research design. Although there
was substantial information on female older adults in
assisted living communities, a review of the literature
revealed that little was known about how social
engagement is experienced by men in later life. Using
a quantitative approach followed by a qualitative
approach, the researchers were able to examine gender
differences in social engagement and psychological
well-being among residents in assisted living com-
munities. In addition, they explored experiences and
challenges concerning social relationships for older
men. Their findings suggested that older men are less
likely to be satisfied with their lives in an assisted
living community because they feel that their social
worlds are limited. The most striking finding was that
assisted living communities are not well designed for
men’s emotional or social needs. These researchers
gave a “voice” to older men and a call for assisted
living administrators, social workers, and staff to be
aware of the specialized social and emotional needs
of male residents (Park et al., 2009).
Concurrent Triangulation Strategy
Concurrent triangulation strategy is a more familiar
approach to researchers. This model is selected when
a researcher wishes to use quantitative and qualitative
methods in an attempt to confirm, cross-validate, or
corroborate findings within a single study. This model
generally uses separate quantitative and qualitative
methods as a mechanism to compensate for the weak-
nesses intrinsic in one method with the strengths of
the other method. Therefore, the quantitative and qual-
itative data collection processes are conducted concur-
rently. This strategy usually integrates the results of
the two methods during the interpretation phase; and
convergence strengthens the knowledge claims or
explains lack of convergence (see Figure 10-4). Great
researcher effort and expertise are needed to study
a phenomenon with two methods. Because of the dif-
ferent methods, researchers are challenged with the
Figure 10-3  Sequential transformative strategy. (Adapted
from Creswell, J. W. [2009]. Research design: Qualitative,
Theory
quantitative, and mixed methods approaches [3rd ed.].
Los Angeles, CA: Sage; and Hanson, W. E., Creswell, J. W.,
Plano Clark, V. L., Petska, K. S., & Creswell, J. D. [2005].
Mixed methods research designs in counseling
psychology. Journal of Counseling Psychology, 52[2],
224–235.)
Quantitative + Qualitative
Data Data
Collection Collection
Data Data
Analysis Analysis
Data Results Compared
Used as a means to offset the weakness inherent with
one method with the strengths of the other method.
Interpretation can either note the convergence of the
findings as a way to strengthen the knowledge claims
of the study or explain any lack of convergence that
may result.
Figure 10-4  Concurrent triangulation strategy. (Adapted from Creswell,
J. W. [2009]. Research design: Qualitative, quantitative, and mixed
methods approaches [3rd ed.]. Los Angeles, CA: Sage; and Hanson, W.
E., Creswell, J. W., Plano Clark, V. L., Petska, K. S., & Creswell, J. D.
[2005]. Mixed methods research designs in counseling psychology.
Journal of Counseling Psychology, 52[2], 224–235.)
difficulty of comparing the study results and determin-
ing the study findings. It is still unclear how to best
resolve discrepancies in findings between methods
(Creswell, 2009).
An example of the concurrent mixed-methods
or triangulation approach is the study by Manuel et al.
(2007). These researchers concurrently conducted
quantitative and qualitative approaches to examine the
coping strategies used by young women diagnosed
with breast cancer. The modified Ways of Coping–
Cancer Version (WOC-CA) scale was used to
collect quantitative data. After the WOC-CA scale,
 212 UNIT TWO  The Research Process
participants were asked two open-ended questions
(qualitative) because studies have consistently found
that younger women show greater psychological dis-
tress in response to breast cancer than older women.
The researchers thought the open-ended responses
would be particularly important because they provided
the opportunity to explore whether existing scales
adequately cover the strategies used by younger
women to cope with breast cancer. In the discussion
section of their report, the researchers indicated that
the data did not converge to confirm current knowl-
edge. The qualitative data provided useful insights
about coping in the population of younger women. For
example, although quantitative data revealed that
“wishful thinking” was commonly used in dealing
with the cancer, responses of the younger women to
the open-ended items indicated this strategy was not
used. Qualitative analysis revealed that being physi-
cally active, seeking information, resting, and using
medications and complementary and alternative thera-
pies were the most effective strategies used by younger
women. However, these items are not included in
either the WOC-CA or other commonly used cancer
coping scales. These findings suggest that researchers
and healthcare professionals should be aware that fre-
quently used coping scales may not include significant
strategies used by younger women dealing with the
diagnosis of breast cancer.
KEY POINTS
• Research design is a blueprint for the conduct of a
study that maximizes the researcher’s control over
factors that could interfere with the desired
outcomes.
• Before selecting a design, the researcher must
understand certain concepts: causality, multicausal-
oasis-ebl|Rsalles|1475856587
ity, probability, bias, manipulation, and control.
• The purpose of design is to set up a situation
that maximizes the possibilities of obtaining
valid answers to research questions or testing
hypotheses.
• Study validity is a measure of the truth or accuracy
of the research findings and is an important concern
throughout the research process.
• The four types of study design validity requiring
examination in a study are (1) statistical conclusion
validity; (2) internal validity; (3) construct validity;
and (4) external validity.
• Statistical conclusion validity is concerned with
whether the conclusions about relationships or dif-
ferences drawn from statistical analyses are an
accurate reflection of the real world. Several poten-
tial threats to statistical conclusion design validity
are addressed.
• Internal validity is the extent to which the effects
detected in the study are a true reflection of reality
rather than the result of extraneous variables, and
potential threats to internal validity are discussed.
• Construct validity examines the fit between the
conceptual and operational definitions of a vari-
able, and threats to this type of design validity are
discussed.
• External validity is concerned with the extent to
which study findings can be generalized beyond the
sample used in the study. The potential threats to
external design validity are identified.
• A good design provides the subjects, the setting,
and the protocol within which these comparisons
can be clearly examined.
• Designs are developed to reduce threats to the
validity of the comparisons. However, some designs
are more effective in reducing threats than others.
• The elements of a good design include control of
(1) the environment; (2) equivalence of subjects
and groups; (3) treatment; (4) measurement; and
(5) extraneous variables.
• In designing the study, the researcher must identify
extraneous variables that could explain some of the
variance in measurement of the study variables.
• Questions are provided to direct the reader in
developing and implementing a study design.
• Mixed-methods approaches include the combined
use of quantitative and qualitative research methods.
Data are collected either sequentially or concur-
rently, and theory may or may not be used in
conducting the study. Merging of data occurs at
different points of the study depending on how
the qualitative and quantitative methods are
implemented.
• The four mixed-methods approaches usually imple-
mented in nursing research are (1) sequential
explanatory strategy; (2) sequential exploratory
strategy; (3) sequential transformative strategy; and
(4) concurrent triangulation strategy.
REFERENCES
Aberson, C. L. (2010). Applied power analysis for the behavioral
sciences. New York, NY: Routledge Taylor & Francis Group.
Bialocerkowski, A., Klupp, N., & Bragge, P. (2010). Research
methodology series: How to read and critically appraise a reli-
ability article. International Journal of Therapy and Rehabilita-
tion, 17(3), 114–120.
Brown, S.J. (2009). Evidence-based nursing: The research-practice
connection. Boston, MA: Jones & Bartlett.
 CHAPTER 10  Understanding Quantitative Research Design 213
Campbell, D. T., & Fiske, D. W. (1959). Convergent and discrimi-
nate validation by the multitrait-multimethod matrix. Psychologi-
cal Bulletin, 56(2), 81–105.
Campbell, D. T., & Stanley, J. C. (1963). Experimental and quasi-
experimental designs for research. Chicago, IL: Rand McNally.
Carr, E. C. (2009). Understanding inadequate pain management in
the clinical setting: The value of the sequential explanatory mixed
method study. Journal of Clinical Nursing, 18(1), 124–131.
Choi, M., & Harwood, J. (2004). A hypothesized Model of Korean
Women’s Responses to Abuse. Journal of Transcultural Nursing,
15(3), 207–216.
Choi, M., Phillips, L. R., Figueredo, A. J., Insel, K., & Min, S. K.
(2008). Construct validity of the Korean Women’s Abuse Intoler-
ance Scale. Nursing Research, 57(1), 40–50.
Cook, T. D., & Campbell, D. T. (1979). Quasi-experimentation:
Design and analysis issues for field settings. Chicago, IL: Rand
McNally.
Corty, E. W. (2007). Using and interpreting statistics: A practical
text for the health, behavioral, and social sciences. St. Louis, MO:
Mosby Elsevier.
Coward, D. D. (1990). Critical multiplism: A research strategy for
nursing science. Image—Journal of Nursing Scholarship, 22(3),
163–167.
Creswell, J. W. (2009). Research design: Qualitative, quantitative,
and mixed methods approaches (3rd ed.). Los Angeles, CA: Sage.
Denzin, N. K. (1989). The research act: A theoretical introduction
to sociological methods (3rd ed.). New York: McGraw-Hill.
DeVon, H. A., Block, M. E., Moyle-Wright, P., Ernst, D. M.,
Hayden, S. J., Lazzara, D. J., et al. (2007). A psychometric toolbox
for testing validity and reliability. Journal of Nursing Scholarship,
39(2), 155–164.
Doran, D. M. (2011). Nursing outcomes: The state of the science
(2nd ed.). Sudbury, MA: Jones & Bartlett Learning.
Duffy, M. E. (1987). Methodological triangulation: A vehicle for
merging quantitative and qualitative research methods. Image—
Journal of Nursing Scholarship, 19(3), 130–133.
Fawcett, J., & Garity, J. (2009). Evaluating research for evidence-
based nursing practice. Philadelphia, PA: F. A. Davis.
Forbes, A. (2009). Clinical intervention research in nursing. Inter-
national Journal of Nursing Studies, 46(4), 557–568.
Goodwin, L. D. (1984). Increasing efficiency and precision of data
analysis: Multivariate vs. univariate statistical techniques. Nursing
Research, 33(4), 247–249.
Grove, S. K. (2007). Statistics for health care research: A practical
workbook. Philadelphia, PA: Saunders Elsevier.
Hanson, W. E., Creswell, J. W., Plano Clark, V. L., Petska, K. S., &
Creswell, J. D. (2005). Mixed methods research designs in coun-
seling psychology. Journal of Counseling Psychology, 52(2),
224–235.
Kerlinger, F. N., & Lee, H. B. (2000). Foundations of behavioral
research (4th ed.). Fort Worth, TX: Harcourt College
Publishers.
Manuel, J. C., Burwell, S. R., Crawford, S. L., Lawrence, R. H.,
Farmer, D. F., Hege, A., et al. (2007). Younger women’s
perceptions of coping with breast cancer. Cancer Nursing, 30(2),
85–94.
McGuire, D. B., DeLoney, V. G., Yeager, K. A., Owen, D. C.,
Peterson, D. E., Lin, L. S., et al. (2000). Maintaining study valid-
ity in a changing clinical environment. Nursing Research, 49(4),
231–235.
Marshall, C., & Rossman, G. B. (2011). Designing qualitative
research (5th ed.). Thousand Oaks, CA: Sage.
Melnyk, B. M., & Fineout-Overholt, E. (2011). Evidence-based
practice in nursing & healthcare: A guide to best practice (2nd
ed.). Philadelphia, PA: Lippincott Williams, & Wilkins.
Mitchell, E. S. (1986). Multiple triangulation: A methodology for
nursing science. Advances in Nursing Science, 8(3), 18–26.
Morrison, D. M., Hoppe, M. J., Gillmore, M. R., Kluver, C., Higa,
D., & Wells, E. A. (2009). Replicating an intervention: The
tension between fidelity and adaptation. AIDS Education and Pre-
vention, 21(2), 128–140.
Morse, J. M. (1991). Approaches to qualitative-quantitative meth-
odological triangulation. Nursing Research, 40(1), 120–123.
Munhall, P. L. (2012). Nursing research: A qualitative perspective
(5th ed.). Miami: Jones & Bartlett.
Myers, S. T., & Haase, J. E. (1989). Guidelines for integration of
quantitative and qualitative approaches. Nursing Research, 38(5),
299–301.
Park N. S., Knapp, M. A., & Shin, H. J. (2009). Mixed methods
study of social engagement in assisted living communities: Chal-
lenges. Journal of Gerontological Social Work, 52(8), 767–783.
Patton, M. Q. (2002). Qualitative research and evaluation methods
(3rd ed.). Thousand Oaks, CA: Sage.
Porter, E. J. (1989). The qualitative-quantitative dualism. Image—
Journal of Nursing Scholarship, 21(2), 98–102.
Rocco, T. S., Bliss, L. A., Gallagher, S., & Perez-Prado, A. (2003).
Taking the next step: Mixed methods research in organizational
systems. Information Technology, Learning, and Performance
Journal, 21(1), 19–29.
Ryan-Wenger, N. A. (2010). Evaluation of measurement precision,
accuracy, and error in biophysical data for clinical research and
practice. In C. F. Waltz, O. L. Strickland, & E. R. Lenz (Eds.),
Measurement in nursing and health research (4th ed.) (pp. 371–
383). New York, NY: Springer Publishing Company.
Santacroce, S. J., Maccarelli, L. M., & Grey, M. (2004). Methods:
Intervention fidelity. Nursing Research, 53(1), 63–66.
Shadish, W. R., Cook, T. D., & Campbell, D. T. (2002). Experimen-
tal and quasi-experimental designs for generalized causal infer-
ence. Chicago, IL: Rand McNally.
Thompson, S. K. (2002). Sampling (2nd ed.). New York, NY: John
Wiley & Sons.
Waltz, C. F., Strickland, O. L., & Lenz, E. R. (2010). Measurement
in nursing and health research (4th ed.). New York, NY: Springer
Publishing Company.
Yamada, J., Stevens, B., Sidani, S., Watt-Watson, J., & De Silva, N.
(2010). Content validity of a process evaluation checklist to
measure intervention implementation fidelity of the EPIC inter-
vention. Worldviews of Evidence-Based Nursing, 7(3), 158–164.

11
CHAPTER
Selecting a Quantitative
Research Design
A
design is the blueprint for conducting a study
that maximizes control over factors that
could interfere with the validity of the find-
ings. A research design gives you greater control and
thus improves the validity of your study. To select
an appropriate research design, you will need to
integrate many elements. Chapter 10 began with
questions that will help you select a design or iden-
tify by name the design of a study you are critically
appraising. Identifying the design of a published
study is not always easy, because many published
studies do not identify the design used. Determining
the design may require you to put together bits
of information from various parts of the research
report.
This chapter describes the designs most commonly
used in nursing research, using the study categories
described in Chapter 3: descriptive, correlational,
quasi-experimental, and experimental. Descriptive
and correlational designs examine variables in natural
environments, such as home, and do not include
researcher-designed treatments or interventions. Quasi-
experimental and experimental designs examine the
effects of an intervention by comparing differences
between groups that have received the intervention
and those that have not received the intervention. As
you review each design, note the threats to validity
that are controlled by the design, keeping in mind that
uncontrolled threats in the design you choose may
weaken the validity of your study. Box 11-1 lists the
designs discussed in this chapter. Each of the designs
is briefly described, and a model is provided so you
can see the different elements of the designs. After the
descriptions of the designs, we provide a series of
decision trees or algorithms that will help you to select
the appropriate design for your study or to identify the
design used in a published study.
Investigators have always developed designs to
meet emerging research needs. In the 1930s, Sir
214
  http://evolve.elsevier.com/Grove/practice/
Ronald A. Fisher (1935) developed the first experi-
mental designs that were published in a book titled
The Design of Experiments. However, most work
on design has been conducted since the 1970s, and
the designs of the last 20 years have become
much more sophisticated and varied. There is no
universal standard for categorizing designs. Names
of designs change as various writers discuss them.
Researchers sometimes merge elements of several
designs to meet the research needs of a particular
study. From these developments, new designs some-
times emerge.
Originally, only experimental designs were con-
sidered of value. In addition, many believed that
the only setting in which an experiment can be
conducted is a laboratory, where stricter controls
can be maintained than in a field or natural setting.
This approach is appropriate for the natural sciences
but not for the social sciences. From the social sci-
ences have emerged additional quantitative designs
(descriptive, correlational, and quasi-experimental),
methodological designs, and qualitative designs
(Cook & Campbell, 1979; Creswell, 2009; Fawcett
& Garity, 2009). The epidemiology, public health,
and community health fields have presented time-
series designs, health promotion designs, and preven-
tion designs.
Currently, most nurse researchers are using
designs developed in other disciplines, such as psy-
chology, medicine, sociology, epidemiology, and
education, that meet the needs of those disciplines.
Will these designs be effective in adding to the
knowledge base required for nursing? These designs
are a useful starting point, but nurse scientists must
go beyond them to develop designs that will more
appropriately meet the needs of the nursing commu-
nity. To go beyond current designs, nurse scientists
must have a working knowledge of available designs
and of the logic on which they are based. Designs
 CHAPTER 11  Selecting a Quantitative Research Design 215
Box 11-1 Research Designs
Descriptive Study Designs
Typical descriptive study designs
Comparative descriptive designs
Time-dimensional designs:
Longitudinal designs
Cross-sectional designs
Trend designs
Event-partitioning designs
Case study designs
Correlational Study Designs
Descriptive correlational designs
Predictive designs
Model-testing designs
Quasi-experimental Study Designs
Nonequivalent comparison group designs:
One-group posttest-only design
Posttest-only design with comparison group
One-group pretest-posttest design
Pretest and posttest design with a
comparison group
Pretest and posttest design with two
comparison treatments
Pretest and posttest design with two
comparison treatments and a standard or
routine care group
Pretest and posttest design with a removed
treatment
Pretest and posttest design with a reversed
treatment
Interrupted time-series designs:
Simple interrupted time-series designs
Interrupted time-series design with a
no-treatment comparison group
Interrupted time-series design with multiple
treatment replications
Experimental Study Designs
Classic experimental design
Experimental posttest-only comparison group
design
Randomized block design
Factorial design
Nested design
Crossover or counterbalanced design
Clinical trials
Randomized controlled trials
created to meet nursing needs should be congruent
with nursing philosophy. They must provide a means
for nurses to examine dimensions of nursing within a
holistic framework and to review those dimensions
over time. Designs must be developed that can seek
answers to important nursing questions rather than
answering only questions that can be examined by
existing designs.
Innovative design strategies are beginning to
appear within nursing research. One example is the
intervention research design described in Chapter
14. Developing designs to study the outcomes of
nursing actions is also important. The emerging
field of outcomes research in nursing is described in
Chapter 13. Nurse researchers must see themselves
as credible scientists before they will dare to develop
new design strategies to explore little-understood
aspects of nursing. To develop a new design, the
researcher must carefully consider possible threats
to validity and ways to diminish them. Nurses must
also be willing to risk the temporary failures that
are always inherent in the development of some-
thing new.
Descriptive Study Designs
Descriptive study designs (see Box 11-1) are crafted
to gain more information about characteristics within
a particular field of study. Their purpose is to provide
a picture of situations as they naturally happen. In
many aspects of nursing, a phenomenon must be
clearly delineated before prediction or causality can
be examined. A descriptive design may be used to
develop theory, identify problems with current prac-
tice, justify current practice, make judgments, or
determine what others in similar situations are doing.
Variables are not manipulated, and there is no treat-
ment or intervention. Dependent and independent
variables are not appropriate for use within a descrip-
tive design, because the design involves no attempt to
establish causality.
Descriptive designs vary in levels of complexity.
Some contain only two variables, whereas others may
have multiple variables. The relationships among vari-
ables present an overall picture of the phenomenon
being examined, but examination of types and degrees
of relationships is not the primary purpose of a descrip-
tive study. Protection against bias (or threat to the
validity) in a descriptive design is achieved through
(1) links between conceptual and operational defini-
tions of variables (Fawcett & Garity, 2009); (2) sample
selection and size (Aberson, 2010; Thompson, 2002);
(3) the use of valid and reliable instruments (Waltz,
216 UNIT TWO  The Research Process
CLARIFICATION MEASUREMENT
Variable 1
Variable 2
Phenomenon
of interest
Variable 3
Variable 4
Figure 11-1  Typical descriptive study design.
Strickland, & Lenz, 2010) or accurate and precise
biophysical measures (Ryan-Wenger, 2010); and (4)
data collection procedures that achieve some environ-
mental control (Bialocerkowski, Klupp, & Bragge,
2010; Creswell, 2009; DeVon et al., 2007; Kerlinger
& Lee, 2000).
Typical Descriptive Study Designs
Figure 11-1 presents the commonly used descriptive
study design that examines characteristics of a single
sample. The design identifies a phenomenon of inter-
est and the variables within the phenomenon, develops
conceptual and operational definitions of the vari-
ables, and describes the variables. The description of
the variables leads to an interpretation of the theoreti-
cal meaning of the findings and provides knowledge
of the variables and the study population that can be
used for future research in the area.
Most studies contain descriptive components;
however, the methodology of some studies is confined
to the typical descriptive design. This is a critically
important design for acquiring knowledge in an area
in which little research has been conducted. Peterson,
Schwab, van Oostrom, Gravenstein, and Caruso
(2010) implemented a descriptive study design to
examine the effects of various patient positions on the
development of pressure in common areas such as the
sacrum and hips. Minimizing skin–support surface
interface pressure is important in preventing and con-
trolling pressure ulcers, but the effects of standard
patient repositioning on skin interface pressure has not
been clearly described. The study design and key
results are presented in the abstract, which is reprinted
in full:
DESCRIPTION INTERPRETATION
Description
of variable 1
Description
of variable 2
Interpretation
of meaning
Description
of variable 3
Description Development
of variable 4 of hypotheses
“Aim: This paper is a report of a study of the effects
of lateral turning on skin-bed interface pressures in
the sacral, trochanteric and buttock regions, and its
effectiveness in unloading at-risk tissue.
Design: This was a descriptive, observational study.
Data were collected from 15 healthy adults from a
university-affiliated hospital. Mapped 24-inch ×
24-inch (2304 half-inch sensors) interface pressure
profiles were obtained in the supine position, fol-
lowed by lateral turning with pillow or wedge support
and subsequent head-of-bed elevation to 30°.”
(Peterson et al., 2010, pp. 1557-1558)
“Results: Raising the head-of-bed to 30° in the
lateral position statistically significantly increased peak
interface pressures and total area ≥ 32 mm Hg. Com-
paring areas ≥ 32 mm Hg from all positions, 93% of
participants had skin areas with interface pressures ≥
32 mm Hg throughout all positions (60 ± 54 cm2),
termed ‘triple jeopardy areas’. The triple jeopardy
area increased statistically significantly with wedges as
compared to pillows (153 ± 99 cm2 vs. 48 ± 47 cm2,
p < 0.05).
Conclusions: Standard turning by experienced
intensive care unit nurses does not reliably relieve
elevated skin-bed interface pressures as intended.
These areas of the body remain at risk for skin break-
down, and help to explain why pressure ulcers occur
despite the implementation of standard preventive
measures. Support materials for maintaining lateral
turned positions can also influence tissue unloading
and triple jeopardy areas.” (Peterson et al., 2010,
p. 1556)
 CHAPTER 11  Selecting a Quantitative Research Design 217
A descriptive research design is appropriate when
the current practice (that is, routine turning of patients)
does not seem sufficient to alleviate the problem (risk
for pressure ulcers). To discover why traditional
nursing interventions have failed to eliminate the pres-
sure ulcers on three at-risk areas of the body, Peterson
et al. (2010) decided to explore the currently used
techniques (pillows versus wedges) and objectively
measure and describe the pressures that participants
experienced using digital pressure sensors. This study
involved no treatment and had one study group of
participants who were measured for pressures in the
at-risk areas.
Some descriptive studies use questionnaires
(surveys) to describe an identified area of concern. For
example, Strand and Lindgren (2010) used a descrip-
tive design to study nurses’ knowledge, attitude, and
barriers regarding preventing pressure ulcers in
patients in intensive care units (ICUs). The following
excerpt from the abstract describes the design of their
study:
Background: “Pressure ulcer incidence varies between
1 and 56% in intensive care and prevention is an
important quest for nursing staff. Critically ill patients
that develop pressure ulcers suffer from increased
morbidity and mortality and also require prolonged
intensive care.
Purpose: The aim of this study was to investigate
registered nurses’ and enrolled nurses’ (1) attitudes,
(2) knowledge, and (3) perceived barriers to oppor-
tunities towards pressure ulcer prevention, in an ICU
setting. These are important aspects in the Theory of
Planned Behaviour, a conceptual framework when
trying to predict, understand, and change specific
behaviours.
Method: The study is descriptive. Questionnaires
were distributed to registered nurses and enrolled
nurses in four ICUs in a Swedish hospital.
Results: The mean score regarding attitude was 34
± 4. Correct categorization of pressure ulcers was
made by 46.8% of the nursing staff…. Pressure relief
(97.3%) and nutritional support (36.1%) were the
most frequently reported preventive measures.
Reported barriers were lack of time (57.8%) and
severely ill patients (28.9%); opportunities were
knowledge (38%) and access to pressure relieving
equipment (35.5%).” (Strand & Lindgren, 2010,
p. 335)
The results of this descriptive study indicated that
nurses are lacking in comprehensive knowledge of
pressure ulcers; additionally, time, acuity of the
patient’s condition, and the availability of supportive
equipment were barriers to the prevention of pressure
ulcers. The study focused on surveying nurses’ knowl-
edge of, attitudes toward, and perceived barriers to
opportunities for pressure ulcer prevention using a
questionnaire. A descriptive design was appropriate
for identifying variables that might influence nursing
care of patients at risk for pressure ulcers in ICU
settings.
Some descriptive studies obtain data from retro-
spective chart review. For example, Kline and Edwards
(2007) conducted a chart review to describe the effec-
tiveness of intrapartum intravenous (IV) insulin on
antepartum and intrapartum control of the mother’s
diabetes and on the occurrence and severity of hypo-
glycemia in the neonate. Researchers measured and
described the variables of intrapartum IV insulin, ante-
partum diabetic control, intrapartum diabetic control,
and hypoglycemia in the neonate.
It is not uncommon for researchers using a descrip-
tive design to combine quantitative descriptive
methods and qualitative methods (mixed methods)
(Creswell, 2009). Chapter 10 includes a discussion of
different types of mixed-methods approaches. To use
this strategy, consult with a researcher experienced in
using qualitative methods or include this person as a
research partner to appropriately collect, analyze, and
interpret qualitative data.
Meghani and Keane (2007) used quantitative and
qualitative methods in their study of preference
for analgesic treatment in African-American cancer
patients. The investigators used demographic data, the
Brief Pain Inventory, and in-depth semi-structured
interviews. Their sample of 35 patients was taken
from three outpatient oncology clinics. Their study
identified the major sources of anxiety described by
this sample. The goal of this study was to improve our
understanding of patients’ needs and assist in the
development of specific interventions that might alle-
viate their problems.
Comparative Descriptive Designs
The comparative descriptive design (Figure 11-2)
examines and describes differences in variables in two
or more groups that occur naturally in the setting.
Descriptive statistics and inferential statistical analy-
ses may be used to examine differences between or
among groups. Commonly, the results obtained from
these analyses are not generalized to a population
because the description is for a very specific sample
and would not necessarily apply to a larger population.
An example of this design is the study by Cramer,
218 UNIT TWO  The Research Process
Group I
(variable[s] Describe
measured)
Comparison
of groups on
select variables
Group II
(variable[s] Describe
measured)
Figure 11-2  Comparative descriptive design.
Chen, Roberts, and Clute (2007) of the social and
economic impact of community-based prenatal care.
The abstract for this study, which is reprinted in full,
describes the focus, design, and major findings:
“Objective: This article describes the evaluation and
findings of a community-based prenatal care program,
Omaha Healthy Start (OHS), designed to reduce local
racial disparities in birth outcome.
Design: This evaluative study used a comparative
descriptive design, and Targeting Outcomes of Pro-
grams was the conceptual framework for evaluation.
Sample: The evaluation followed 3 groups for 2 years:
OHS birth mothers (N = 79; N = 157); non-OHS
participant birth mothers (N = 746; N = 774); and
Douglas County birth mothers (N = 7,962; N =
7,987).
Measurement: OHS provided case management,
home visits, screening, referral, transportation, and
health education to participants. Program outcome
measures included low birth weight, infant mortality,
adequacy of care, trimester of care, and costs of care.
Results: OHS birth outcomes improved during year
2, and there was a 31% cost saving in the average
hospital expenditure compared with the nonpartici-
pant groups. Preliminary evaluative analysis indicates
that prenatal case management and community out-
reach can improve birth outcomes for minority
women, while producing cost savings.
Conclusions: Further prospective study is needed to
document trends over a longer period of time regard-
ing the relationship between community-based case
management programs for minority populations,
birth outcomes, and costs of care.” (Cramer et al.,
2007, p. 329)
Interpretation
of meaning
Development
of hypotheses
Using a comparative descriptive design, Cramer
and colleagues (2007) described the variables inci-
dence of case management, home visits, screening,
referral, transportation, and health education, as well
as the outcomes low birth weight, infant mortality,
adequacy of care, trimester of care, and costs of care
in three groups (OHS birth mothers, non-OHS birth
mothers, and Douglas County birth mothers) yearly
for 3 years. The results of the study were comparisons
across the 3 years and among the three groups.
Time-Dimensional Designs
Time-dimensional designs were developed within
the discipline of epidemiology, a field that studies the
occurrence and distribution of disease among popula-
tions. These designs examine trends over time, growth,
or sequences and patterns of change. The dimension
of time, then, becomes an important factor. Within the
field of epidemiology, the samples in time-dimensional
studies are called cohorts. Originally, cohorts were
age categories; however, the concept has been
expanded to apply to groups distinguished by many
other variables. Other means of classifying popula-
tions that have relevance in relation to time are time
of diagnosis, point of entry into a treatment protocol,
point of entry into a new lifestyle, and age at which
the subject started smoking. An understanding of tem-
poral sequencing is an important prerequisite to exam-
ining causality between variables. Thus, the results of
these designs lead to description of trends, processes,
patterns, and changes over time as well as the develop-
ment of hypotheses, and are often forerunners of
experimental designs (Fawcett & Garity, 2009).
Epidemiological studies that use time-dimensional
designs determine the risk factors or causal factors of
illness states. Cause determined in this manner is
called inferred causality (Kerlinger & Lee, 2000).
 CHAPTER 11  Selecting a Quantitative Research Design 219
Time 1 Time 2 Time 3 Time 4
measure measure measure measure
variables variables variables variables
Sample 1 Sample 1 Sample 1 Sample 1
These studies also examine trends, patterns, and
changes over time. The best-known studies in this area
are those on smoking and cancer. Because of the
strength of studies that have undergone multiple rep-
etitions, the causal link is strong. The strategy is not
as powerful as experimental designs in supporting
causality; however, in this situation, as in many
nursing contexts, one can never ethically conduct a
true experiment. A true experiment requires that there
be an experimental group (who would not smoke) and
a control group (who smoke). The participants must
be randomly assigned to one of these groups. There-
fore, without being provided a choice, some individu-
als would be required to smoke and others would be
required to abstain from smoking over a long period.
Epidemiologists use two strategies to examine
changes over time: retrospective studies and prospec-
tive studies. The norm in epidemiological studies is to
use the word cohorts to refer to groups of subjects in
prospective studies, but the term is generally not used
in retrospective studies. In retrospective studies, both
the proposed cause and the proposed effect have
already occurred. For example, the subjects could
have a specific type of cancer, and the researcher could
be searching for commonalities among subjects that
may have led to the development of that type of
cancer. In a prospective cohort study, causes may
have occurred, but the proposed effect has not.
The Framingham study is the best-known example
of a prospective study (U.S. Department of Health and
Human Services, 1968). In this study, researchers
monitored members of a community for 20 years and
examined variables such as dietary patterns, exercise,
weight, and blood lipid levels. As the subjects experi-
enced illnesses, such as heart disease, hypertension,
and lung disease, their illnesses could be related to
previously identified variables.
Prospective studies are considered more powerful
than retrospective studies in inferring causality,
because the researcher can demonstrate that the risk
factors occurred before the illness and are positively
related to the illness. Both designs are important for
use in nursing studies, because a person’s responses
to health situations are patterns that developed long
Time..n
measure
variables
Figure 11-3  Longitudinal design.
Sample 1
before the health situation occurred. These patterns
then influence the person’s responses to nursing inter-
ventions. Several designs are used to conduct time-
dimensional studies: longitudinal, cross-sectional,
trend, and event or treatment partitioning. These
designs are discussed in the following sections.
Longitudinal Designs
Longitudinal designs examine changes in the same
subjects over time. They are sometimes called panel
designs (Figure 11-3). Longitudinal designs are expen-
sive and require a long period of researcher and subject
commitment. The area to be studied, the variables, and
their measurement must be clearly identified before
data collection begins. Measurement must be carefully
planned and implemented because the measures will
be used repeatedly over time. If children are being
studied, the measures must be valid for all the ages
being studied. To use this design, researchers must be
familiar with how the construct being measured
changes and its patterns and trends over time. In addi-
tion, they need to provide a clear rationale for the
points of time they have selected for measurement.
There is often a bias in selection of subjects because
of the requirement for a long-term commitment. Indi-
viduals participating in a study conducted over long
periods might differ in some important ways from the
target population. In addition, attrition or loss of sub-
jects from the study can be high and can decrease the
validity of findings.
The sample size calculated with power analysis
needs to take into consideration the potential attrition
rate when determining the final number of subjects to
recruit. As a researcher, you must invest considerable
energy in developing effective strategies to maintain
the sample (see Chapter 15). The period during which
subjects will be recruited into the study must be care-
fully planned, and a timeline depicting data collection
points for each subject must be developed to enable
planning for the numbers and availability of data col-
lectors. If this issue is not carefully thought out, data
collectors may be confronted with the need to recruit
new subjects while they are attempting to collect data
scheduled for subjects recruited earlier. You must also
220 UNIT TWO  The Research Process

decide whether you will use a single data collector to

obtain all data from a particular subject or whether you
will use a different data collector at each point to
ensure that data are collected blindly.
Because of the large volumes of data acquired in a
longitudinal study, you must give careful attention to
strategies for managing the data. The repetition of
measures requires that data analysis be carefully
thought through. Analyses commonly used are
repeated measures analyses of variance, multivariate
analyses of variance (MANOVA), regression analysis,
cluster analysis, and time-series analysis (see Chapters
24 and 25) (Corty, 2007; Munro, 2005).
Lee, Chaboyer, and Wallis (2010) conducted a
descriptive study using a longitudinal cohort design.
This study was conducted to describe the perceptions
and physical manifestations of injury and illness of
patients with traumatic injury and to examine the
changes they experienced over time. The following
abstract from the study demonstrates the background,
longitudinal design, key results, conclusions, and
implications for practice:
Relevance to Clinical Practice: Understanding illness
representation in patients with traumatic injury may
help nurses to provide anticipatory guidance and to
design nursing interventions before and after hospital
discharge, ultimately to improve health outcomes of
those patients.” (Lee et al., 2010, p. 556)
Cross-Sectional Designs
Cross-sectional designs examine groups of subjects
in various stages of development, trends, patterns, and
changes simultaneously with the intent to describe
changes in the phenomenon across stages (see Figure
11-4). The assumption is that the stages are part of a
Time 1
measure
variables
Sample 1
Immediate Loss

“Background: Traumatic injury has attracted global Time 1

concern because it is the major reason for death and measure
disability in people under 45 years old. One model, variables
the Common Sense Model of Illness Representation
(CSMIR), has the potential to help individuals adjust Sample 2
Loss 6 mo
to changes in health status such as traumatic injury.
Design: Longitudinal study design.
Methods: This study was conducted using data col- Time 1
lected prior to hospital discharge and at three and six measure
months after hospital discharge. One individual ques- variables
tion form and the Chinese Illness Perception Ques-
tionnaire Revised (IPQ-R) (Trauma) were used to Sample 3
Loss 1 yr
collect demographic data, clinical data, and illness
representations.
Results: A total of 114 participants completed Time 1
the survey three times. The overall response rate measure
was 79.7%. Six subscales of the Chinese IPQ-R variables
(Trauma)… identity, emotional representations, con-
sequences, controllability, illness coherence, and Sample 4
causes… changed significantly over time. Two sub- Loss 2 yr
scales, Timeline (acute/chronic) and Timeline Cycli-
cal, did not change significantly. Time 1
Conclusions: Based on these findings, there may be
measure
a window of opportunity to provide appropriate variables
interventions to individuals with traumatic injury at
each time point. The results of this study have impli- Sample 5
cations for nursing practice and further nursing Loss 5 yr
research.
Figure 11-4  Cross-sectional design.
 CHAPTER 11  Selecting a Quantitative Research Design 221
process that will progress over time. Selecting subjects
at various points in the process provides important
information about the totality of the process, even
though the same subjects are not monitored through
the entire process. The processes of development
selected for the study might be related to age, position
in an educational system, growth pattern, or stages of
maturation or personal growth (if they could be clearly
enough defined that criteria could be developed for
inclusion within differentiated groups or disease
stages). Subjects are then categorized by group, and
data on the selected variables are collected at a single
point in time.
For example, suppose you wish to study grief reac-
tions at various periods after the death of a spouse.
With a cross-sectional design, you could study a group
of individuals whose spouses had died 1 week ago,
another group composed of individuals whose losses
occurred 6 months ago, and other groups whose losses
occurred 1 year, 2 years, and 5 years ago (see Figure
11-4). You could study all of these groups during one
period of time, but you could describe a pattern of
grief reactions over a 5-year period. The design is not
as strong as the longitudinal design, in which the same
participants continue in the study over time, thus elim-
inating some variance, but it allows some understand-
ing of the phenomenon over time when time allowed
for the study is limited.
Sidani et al. (2007) conducted a cross-sectional
study titled “Outcomes of Nurse Practitioners in Acute
Care: An Exploration.” The following abstract
describes the design of their study:
“The purpose of this study was to compare the out-
comes achieved by adult patients who did (n = 78)
and did not (n = 45) receive care by acute care nurse
practitioners (ACNPs), within one week following
discharge. A comparative, cross-sectional design was
used. Consenting patients completed the outcome
measures within one week following discharge. The
Time 1 Time 2 Time 3
measure measure measure
variables variables variables
Sample 1 Sample 2 Sample 3
Figure 11-5  Trend design.
outcomes included satisfaction with care, functional
status, symptom resolution, and sense of well-being,
which were measured with established instruments.
The two groups of patients were equivalent in terms
of their demographic profile and severity of condition.
The results indicated that patients who received
ACNP care, as compared to those who did not,
reported higher levels of satisfaction with care and of
physical, psychological, and social functioning. These
findings provide preliminary evidence supporting the
contribution of ACNPs to high quality care. However,
the small sample size limits the generalizability of the
study findings.” (Sidani et al., 2007, p. 1)
Trend Designs
Trend designs examine changes in the general popu-
lation in relation to a particular phenomenon (see
Figure 11-5). The researchers select different samples
of subjects from the same population at preset inter-
vals of time, and at each selected time, they collect
data from that particular sample. Researchers need to
be able to justify generalizing from the samples to the
population under study. Analysis involves strategies to
predict future trends by examining past trends. Harris,
Gordon-Larsen, Chantala, and Udry (2006, p. 74) used
a trend design to describe “longitudinal trends in race/
ethnic disparities in 20 leading health indicators from
Healthy People 2010 [U.S. Department of Health and
Human Services, 2000] across multiple domains from
adolescence to young adulthood.” These researchers
examined the study trends in an ethnically diverse,
national database, and their study design is described
in the following excerpt.
“Design, Setting, and Participants: Nationally represen-
tative data for more than 14,000 adolescents enrolled
in wave I (1994-1995) or wave II (1996) of the
National Longitudinal Study of Adolescent Health
Time 4 Time..n
measure measure
variables variables PREDICTION
Sample 4 ...n
222 UNIT TWO  The Research Process
Time 1
Key event 1
Sample 1
Time 2
Key event 2
Sample 2
Time 3
Key event 3
Sample 3
Key event 4
Figure 11-6  Cross-sectional study with treatment partitioning.
(Add Health) were followed up into adulthood (wave
III; 2001-2002). We fit longitudinal regression models
to assess and contrast the trend in health indicators
among racial/ethnic groups of adolescents as they
transition into adulthood….
Results: Diet, inactivity, obesity, health care access,
substance use, and reproductive health worsened
with age. Perceived health, mental health, and expo-
sure to violence improved with age. On most health
indicators, white and Asian subjects were at the
lowest risk and Native American subjects at the
highest risk. Although white subjects had more favor-
able health in adolescence, they experienced the
greatest declines by young adulthood. No single race/
ethnic group consistently leads or falters in health
across all indicators.” (Harris et al., 2006, p. 74)
Harris et al. (2006) found the trend design to be an
effective way to examine health indicators of Ameri-
cans over time. They noted for 15 of 20 indicators that
the health risk increased and access to health care
decreased from teen to adult years for most Ameri-
cans. The health indicators varied over time by gender
and race/ethnicity, causing the health disparities to
fluctuate over time.
Event-Partitioning Designs
A merger of the cross-sectional or longitudinal and
trend designs, the event-partitioning design, is used
Time..n
Merger of data for analysis
and interpretation of meaning
Sample..n
in some cases to enlarge sample size and to avoid the
effects of history on the validity of findings. Cook and
Campbell (1979) referred to these as cohort designs
with treatment partitioning. Figure 11-6 shows a
model of the cross-sectional study design with treat-
ment partitioning, and Figure 11-7 provides the model
of a longitudinal design with treatment partitioning.
The term treatment is used loosely here to mean a key
event that is thought to lead to change. In a descriptive
study, the researcher would not cause or manipulate
the key event but rather would clearly define it so that
when it occurred naturally, it would be recognized.
For example, you could use the event-partitioning
design to study subjects who have completed pro-
grams to stop smoking. Smoking behaviors and
incidence of smoking-related diseases might be mea-
sured at intervals of 1 year for a 5-year period.
However, the number of subjects available at one time
might be insufficient for you to adequately analyze
findings. Therefore, you could use subjects from
several programs offered at different times. You would
examine the data in terms of the relative time since
the subjects’ completion of the stop-smoking program,
not the absolute length of time. Data would be assumed
to be comparable, and a larger sample size would be
available for analysis of changes over time.
Sutton’s (2007) dissertation described a study to
examine the relationship of preoperative education for
the aging adult and anxiety. Most individuals experi-
ence anxiety in anticipation of surgery, and they expect
education and support to help alleviate their anxiety
prior to entering the operating room. Because a
growing number of surgical procedures are being
 CHAPTER 11  Selecting a Quantitative Research Design 223
Time 1 Time 2 Time 3
Key event 1
Sample 1 Sample 1 Sample 1 Sample 1
Time 1 Time 2
Key event 2
Sample 2 Sample 2 Sample 2 Sample 2
Time 1
Key event 3
Sample 3 Sample 3 Sample 3 Sample 3
Figure 11-7  Longitudinal design with treatment partitioning.
conducted on an outpatient basis (day surgery), there
is limited time for healthcare personnel to provide
psychological preparation of patients for the periop-
erative process. Sutton used an event-partitioning
design and correlational analysis to examine the rela-
tionship between education and the state anxiety expe-
rienced by aging adults in the preoperative setting.
Quantitative data were collected using the Visual
Analog Scale (VAS) to measure state anxiety in a pre-
test and post-test format, with treatment consisting of
a scripted preoperative educational presentation.
Patients scheduled for general anesthesia for a sur-
gical procedure in two acute care facilities in southern
West Virginia were asked to participate in this study
over a 4-week time frame. The sample included 52
pre-surgical patients, ages 65 to 94 years, who were
asked to participate in the study. Participants were
asked to score their state anxiety on the VAS prior to
the scripted educational presentation to establish a
baseline anxiety level. State anxiety is the emotion a
person experiences in a particular situation, versus
trait anxiety, which is the innate anxiety of a person.
Upon completion of this presentation, the participants
were given an opportunity to ask questions and receive
answers. They were then asked to indicate their state
anxiety on the VAS. The study results indicated that
75% of the study participants reported a decrease in
their state anxiety levels following the preoperative
educational presentation. This intervention requires
further testing using a quasi-experimental study design
but has the potential to reduce state anxiety in indi-
viduals experiencing day surgery (Sutton, 2007).
Time 4
Time 3 Time 4
Time 2 Time 3 Time 4
Merger of data
for analysis and
interpretation of
meaning
Case Study Designs
The case study design involves an intensive explora-
tion of a single unit of study, such as a person, family,
group, community, or institution, or of a small number
of subjects. Although the number of subjects tends to
be small, the number of variables involved is usually
large. In fact, it is important to examine all variables
that might have an impact on the situation being
studied.
Case studies were commonly used in nursing
research in the 1970s. Their use then declined, but
they are beginning to appear in the literature more
frequently today. Well-designed case studies are good
sources of descriptive information and can be used as
evidence for or against theories. Case studies can use
a mixed-methods approach, incorporating both quan-
titative and qualitative methods (Creswell, 2009;
Fawcett & Garity, 2009). Sterling and McNally (1992)
recommended single-subject case studies for examin-
ing process-based nursing practice. This strategy
allows the researcher to investigate daily observations
and interventions that are a common aspect of nursing
practice.
Case studies also can demonstrate the effectiveness
of specific therapeutic techniques. In fact, by reporting
a case study, the researcher introduces the technique
to other practitioners. The case study design also has
potential for revealing important findings that can gen-
erate new hypotheses for testing. Thus, the case study
can lead to the design of large sample studies to
examine factors identified through the case study. For
example, Sprague and Chang (2011) used a single
224 UNIT TWO  The Research Process
subject case study design to examine the effect of
acupuncture on the treatment of chronic pain. Their
study abstract identifies the key elements of their
study:
“Background: Chronic complex regional pain syn-
drome (CRPS) is a chronic pain conduction that leads
to sympathetic nervous system involvement and
trophic changes.
Objective: This study describes the use of acupunc-
ture in a case study of CRPS.
Design, Setting, and Patient: This is a single case
report of a 34-year-old patient diagnosed with CRPS.
Intervention: Acupuncture treatment including
acupoints along the Gallbladder, Liver, Spleen, Heart,
and Kidney meridians. Self-treatment plan included a
laser acupuncture pen device and disposable press
needles.
Main Outcome Measures: Beck Depression Inven-
tory (BDI), McGill Pain Questionnaire, and Sheehan
Disability Scale (SDS).
Results: The patient reported a decrease in pain
levels, depression, and an improved quality of life.
Pretreatment SDS score of 17, a 12 on the BDI, and
a 67 on the McGill Pain Questionnaire. Post-treatment
SDS decreased to 4, her BDI went to 0, and her
McGill Pain Questionnaire decreased to a 10.”
(Sprague & Change, 2011, p. 67)
Sprague and Chang (2011) noted that this patient
had dramatic improvements in depression, disability,
and pain scores over the 6 months of the study. Thus,
this case study justifies conducting further clinical
studies to determine the effectiveness of acupuncture
in the management of patients with CRPS. They rec-
ommended conducting studies with larger samples
and randomized controlled treatment designs.
How you design a case study depends on the cir-
cumstances of the case but usually includes an element
of time. History and previous behavior patterns are
usually explored in detail. As the case study proceeds,
you may become aware of components important to
the phenomenon being examined that were not origi-
nally built into the study. A case study is likely to have
both quantitative and qualitative elements; and if the
study incorporates both of these components, the
study design must clearly present this fact (Creswell,
2009). Methods used to analyze and interpret qualita-
tive data need to be carefully planned. Consultation
with a qualitative researcher can strengthen the study.
Large volumes of data are generally obtained during
a case study. Organizing the findings of a case study
into a coherent whole is a difficult but critical compo-
nent of the study (Fawcett & Garity, 2009). General-
izing study findings in the statistical sense is not
appropriate; however, generalizing the findings to
theory is appropriate and important (Crombie &
Davies, 1996; Gray, 1998; Sandelowski, 1996).
Not all case studies are research. Many of the arti-
cles referring to case studies are clinical practice arti-
cles, in which a clinical situation is reported for the
purpose of illustrating clinical practice, problems in
clinical practice, or changes that need to be made in
clinical practice. These articles do not use research
methods but, rather, describe events out of the patient
record or the writer’s personal experience.
Surveys
The term survey is used in two ways within scientific
thought. It is used in a broad sense to mean any
descriptive or correlational study; in this sense, survey
tends to mean nonexperimental (Kerlinger & Lee,
2000). In a narrower sense, the term is used to describe
a data collection technique in which the researcher
uses questionnaires (collected by email, by mail, or in
person) or personal interviews to gather data about an
identified population.
Surveys, in the narrower definition, are used to
gather data that can be acquired through self-report.
Because of this limitation in data, some researchers
view surveys as rather shallow and as contributing in
a limited way to scientific knowledge. This belief has
led to a bias in the scientific community against survey
research. In this context, the term survey is used deri-
sively. However, surveys can be an extremely impor-
tant source of data. In this text, we use the term survey
to designate a data collection technique, not a design.
Surveys can be used within many designs, including
descriptive, correlational, and quasi-experimental
studies.
Correlational Study Designs
Correlational study designs examine relationships
among variables. The examination can occur at several
levels of the independent variable. The researcher can
seek to describe a relationship, predict relationships
among variables, or test the relationships proposed by
a theoretical proposition or a model. In any correla-
tional study, a representative sample must be selected
for the study. That sample reflects the full range of
values possible on the variables being measured.
Thus, large samples are required. In correlational
designs, a large variance in the variable values is
 CHAPTER 11  Selecting a Quantitative Research Design 225
necessary to determine the existence of a relationship.
Therefore, correlational designs are unlike experimen-
tal designs, in which variance in variable scores is
controlled by controlling such design elements as the
study setting, sampling criteria, and sampling method
(Kerlinger & Lee, 2000).
In correlational designs, if the range of scores is
truncated, the obtained correlational value will be arti-
ficially depressed. Truncated means that the lowest
values and the highest values for a variable either are
not measured or are condensed and merged with less
extreme values. For example, if an attitude scale were
scored from a low score of 1 to a high score of 50,
truncated scores might indicate only scores in the
range 10 to 40. More extreme scores would be com-
bined with scores within the designated range. If trun-
cation is performed, the researcher may not find a
correlation when the variables are actually correlated.
Neophyte researchers tend to make two serious
errors with correlational studies. First, they often
attempt to establish causality by correlation, reasoning
that if two variables are related, one must cause the
other. Second, they confuse studies in which differ-
ences are examined with studies in which relation-
ships are examined. Although the existence of a
difference assumes the existence of a relationship, the
design and statistical analysis of studies examining
differences are not the same as those of studies exam-
ining relationships. If your study examines two or
more groups in terms of one or more variables, then
you are exploring differences between or among
groups as reflected in scores on the identified
variables. If your study examines a single group in
terms of two or more variables, then you are exploring
relationships between or among variables. In a
MEASUREMENT
Research
variable Description
1 of variable
Examination
of relationship
Research
Description
variable
of variable
2
Figure 11-8  Descriptive correlational design.
correlational study, the relationship examined is
between or among two or more research variables
within an identified situation. Thus, the sample is not
separated into groups. Analyses examine variable
values in the entire sample. In a correlational design,
data from the entire sample are analyzed as a single
group (Grove, 2007; Kerlinger & Lee, 2000).
Descriptive Correlational Designs
A descriptive correlational design examines the rela-
tionships that exist in a situation. Using this design
facilitates the identification of many interrelationships
in a situation in a short time. Although the descriptive
design discussed earlier may reveal relationships
among variables, the descriptive correlational design
focuses specifically on relationships among study
variables. Descriptive correlational studies may lead
to hypotheses for later studies. Figure 11-8 provides a
model of a typical descriptive correlational design
for examining a relationship between two research
variables. This design can be expanded to include
examination of relationships among several study
variables.
A descriptive correlational study may examine
variables in a situation that has already occurred or is
currently occurring. No attempt is made to control or
manipulate the situation. As with descriptive studies,
variables must be clearly identified and defined. An
example of a descriptive correlational design is the
study by Bailey, Sabbagh, Loiselle, Boileau, and
McVey (2010, p. 114) titled “Supporting families in
the ICU: A descriptive correlational study of informa-
tional support, anxiety, and satisfaction with care.”
The researchers conducted this study to describe
family members’ perceptions of informational support,
Interpretation
of meaning
Development
of hypotheses
226 UNIT TWO  The Research Process
Value of
Value of
+ Independent
Intercept
Variable 1
Figure 11-9  Predictive design.
anxiety, and satisfaction with care and to examine the
relationships among these variables. The design for
this study is described in the following excerpt from
its abstract:
“Methodology/Design: This cross-sectional descriptive
correlational pilot study collected data from a conve-
nience sample of 29 family members using self-report
questionnaires.
Setting: 22-bed medical-surgical intensive care unit
of 659-bed University affiliated teaching hospital in
Montreal, Quebec, Canada.
Results: Mean informational support, assessed with
a modified version of the CCFNI [Critical Family
Needs Inventory]…, was 55.41 (SD = 13.28; theoreti-
cal range of 20-80). Mean anxiety, assessed with the
State Anxiety Scale (Spielberger et al., 1983) was
45.41 (SD = 15.27; theoretical range 20-80). Mean
satisfaction with care… was 83.09 (SD = 15.49; theo-
retical range 24-96). A significant positive correlation
was found between informational support and satis-
faction with care (r = 0.741, p < 0.001). No significant
relationships were noted between informational
support and anxiety or between satisfaction with care
and anxiety.” (Bailey et al., 2010, p. 114)
By implementing a descriptive, correlational
research design, Bailey et al. (2010) were able to
describe their study variables and determine relation-
ships among them. The ultimate objective of the
researchers was to further refine a local informational
support initiative for families with members in the
intensive care unit. Thus, this research provided the
basis for development of an intervention that might be
tested in future quasi-experimental or experimental
studies.
This study had a descriptive correlational design,
as evidenced by the single study group, the absence
of treatment, and the use of descriptive and correla-
tional statistical techniques to analyze study data. The
study variables informational support, anxiety, and
satisfaction with care were described with means,
standard deviations, and ranges (see Chapter 22). The
Predicted
Value of
Value of
+ Independent =
Dependent
Variable 2
Variable
relationships among these three variables were deter-
mined using Pearson’s product moment correlational
coefficient (see Chapter 23).
Predictive Designs
Predictive designs are used to predict the value of one
variable on the basis of values obtained from another
variable or variables. Prediction is one approach you
can use to examine causal relationships between or
among variables. Because causal phenomena are
being examined, the terms dependent and independent
are used to describe the variables. One variable (the
one to be predicted) is classified as the dependent
variable, and all other variables (those that are predic-
tors) are classified as independent variables.
The aim of a predictive design is to predict the level
of the dependent variable from the independent vari-
ables. Figure 11-9 is a model of a predictive design
with two independent variables used to predict the
dependent variable. Independent variables most effec-
tive in prediction are highly correlated with the depen-
dent variable but not highly correlated with other
independent variables used in the study. Predictive
designs require you to develop a theory-based math-
ematical hypothesis proposing the independent vari-
ables that are expected to predict the dependent
variable effectively. You can then test the hypothesis
using regression analysis (see Chapter 24) (Corty,
2007; Munro, 2005). Predictive studies are also used
to establish the predictive validity of measurement
scales.
Mancuso (2010) conducted a correlational study
with a cross-sectional predictive correlational design
to examine the impact of health literacy and patient
trust on glycemic control in diabetic adults. The inde-
pendent variables of patient trust, health literacy,
knowledge of diabetes, performance of self-care activ-
ities, and depression were used to predict the depen-
dent variable of glycosylated hemoglobin concentration
(HbA1c). Regression analysis was conducted to deter-
mine how effective the independent variables were in
predicting the dependent variable of HbAlc. The fol-
lowing except from the abstract describes the study
design, results, and conclusions:
 CHAPTER 11  Selecting a Quantitative Research Design 227
EXOGENOUS
VARIABLES
variable
1
NONCAUSAL variable
CORRELATIONS 2 4
variable
3
Figure 11-10  Model-testing design.
“A quantitative study was conducted that examined
health literacy and patient trust as predictors of gly-
cemic control. The related factors of demographic,
socioeconomic status, diabetes knowledge, self-care
activities, and depression were also considered.
Implementing a cross-sectional, predictive design, a
convenience sample of 102 patients with diabetes was
recruited from two urban primary care clinics in the
USA. A simultaneous multiple regression was con-
ducted. The regression analysis was significant, with
patient trust and depression accounting for 28.5% of
the variance in HbA1c.” (Mancuso, 2010, p. 94)
Mancuso (2010) reported that patient trust and
depression were important factors that significantly
predicted HbA1c (R2 = 0.320; R2adj = 0.285; F(5, 96)
= 9.047; p < 0.05). However, knowledge of diabetes,
health literacy, and self-care activities was not a sig-
nificant predictor of glycemic control. She recom-
mended further research examining depression and
patient trust as predictors of glycemic control and
additional studies to explore other influences on or
barriers to glycemic control. Mancuso’s implementa-
tion of a predictive correlational design and conduct
of regression analysis were appropriate to address the
study purpose and research questions.
Model-Testing Designs
Some studies are designed specifically to test the accu-
racy of a hypothesized causal model. The model-
testing design requires that all variables relevant to
the model be measured. A large, heterogeneous sample
ENDOGENOUS
VARIABLES
CAUSAL
CORRELATIONS
variable variable
5
a b
RESIDUAL
VARIABLES
is required. All the paths expressing relationships
between concepts are identified, and a conceptual map
is developed (see Figure 11-10). The analysis deter-
mines whether or not the data are consistent with the
model. For some studies, you might set aside data
from half of the subjects and not include them in the
initial path analysis. You might use these data from
the second half of the subjects to test the fit of the
paths defined by the initial analysis.
Variables are classified into three categories: exog-
enous variables, endogenous variables, and residual
variables. Exogenous variables are within the theo-
retical model but are caused by factors outside this
model. Endogenous variables are those whose varia-
tion is explained within the theoretical model. Exog-
enous variables influence the variation of endogenous
variables. Residual variables indicate the effect of
unmeasured variables not included in the model.
These variables explain some of the variance found in
the data but not the variance within the model (Mason-
Hawkes & Holm, 1989; Norris, 2005a).
In Figure 11-10, the illustration of a model-testing
design, paths are drawn to demonstrate directions of
cause and effect. The arrows (paths) from the exoge-
nous variables 1, 2, and 3 lead to the endogenous
variable 4, indicating that variable 4 is theoretically
proposed to be caused by variables 1, 2, and 3. The
arrow (path) from endogenous variable 4 to endoge-
nous variable 5 indicates that variable 4 theoretically
causes variable 5.
To measure exogenous and endogenous variables,
you would collect data from the subjects and analyze
the accuracy of the proposed paths. Historically, these
228 UNIT TWO  The Research Process
analysis procedures were performed with a series of
regression analyses. Researchers now conduct statisti-
cal procedures that have been developed specifically
for path analysis using the computer programs LISREL
and EQS (Norris, 2005a). Structural equation model-
ing is a commonly used statistical procedure (Norris,
2005b). Path coefficients are calculated that indicate
the effect that one variable has on another. The amount
of variance explained by the model, as well as the fit
between the path coefficients and the theoretical
model, indicates the accuracy of the theory. Variance
that is not accounted for in the statistical analysis is
attributed to residual variables (variables a and b) not
included in the analyses (Mason-Hawkes & Holm,
1989; Norris, 2005a).
An example of this design is the Cummings,
Estabrooks, Midodzi, Wallin, and Hayduk (2007)
study testing a model of the influence of organiza-
tional characteristics and context on research utiliza-
tion in nursing. The following study abstract identifies
the study purpose, design, results, and conclusions:
“Background: Despite three decades of empirical
investigation into research utilization and a renewed
emphasis on evidence-based medicine and evidence-
based practice in the past decade, understanding of
factors influencing research uptake in nursing remains
limited. There is, however, increased awareness that
organizational influences are important.
Objectives: To develop and test a theoretical model
of organizational influences that predict research uti-
lization by nurses and to assess the influence of
varying degrees of context, based on the Promoting
Action on Research Implementation in Health Ser-
vices (PARIHS) framework, on research utilization
and other variables.
Methods: The study sample was drawn from a
census of registered nurses working in acute care
hospitals in Alberta, Canada, accessed through their
professional licensing body (n = 6,526 nurses; 52.8%
response rate). Three variables that measured
PARIHS dimensions of context (culture, leadership,
and evaluation) were used to sort cases into one of
four mutually exclusive data sets that reflected less
positive to more positive context. Then, a theoretical
model of hospital- and unit-level influences on
research utilization was developed and tested, using
structural equation modeling, and 300 cases were
randomly selected from each of the four data sets.
Results: Hospital characteristics that positively
influenced research utilization by nurses were staff
development, opportunity for nurse-to-nurse col-
laboration, and staffing and support services. Increased
emotional exhaustion led to less reported research
utilization and higher rates of patient and nurse
adverse events. Nurses working in contexts with
more positive culture, leadership, and evaluation also
reported significantly more research utilization, staff
development, and lower rates of patient and staff
adverse events than did nurses working in less posi-
tive contexts (i.e., those that lacked positive culture,
leadership, or evaluation).
Conclusion: The findings highlight the combined
importance of culture, leadership, and evaluation to
increase research utilization and improve patient
safety. The findings may serve to strengthen the
PARIHS framework and to suggest that, although it is
not fully developed, the framework is an appropriate
guide to implement research into practice.”
(Cummings et al., 2007, S24)
Defining Therapeutic
Nursing Interventions
In quasi-experimental and experimental studies, an
intervention (or protocol) is developed that is expected
to result in differences in posttest measures of the
treatment and control or comparison groups. This
intervention may be physiological, psychosocial, edu-
cational, or a combination of these and should be
designed to maximize the differences between the
groups. Thus, it should be the best intervention pos-
sible in the circumstances of the study and should be
expected to improve the outcomes of the experimental
group (Egan, Snyder, & Burns, 1992; Forbes, 2009;
Santacroce, Maccarelli, & Grey, 2004).
Over the last 5 years, the nursing literature has
included a growing number of publications focused
on the methodology for designing interventions
for nursing studies (Morrison et al., 2009; Wyatt,
Sikorskii, Rahbar, Victorson, & Adams, 2010; Yamada,
Stevens, Sidani, Watt-Watson, & Silva, 2010). In addi-
tion, descriptions of nursing interventions in published
studies have more detail and specificity but still not at
the level given to describing measurement instruments
(Fawcett & Garity, 2009; Waltz et al., 2010). Thus,
nurse researchers provide detailed information about
measurement but often do not provide sufficient detail
to allow a nurse to implement a nursing intervention
as it was used in a published nursing study. To some
extent, this situation may reflect the state of knowl-
edge in the nursing field regarding the provision of
 CHAPTER 11  Selecting a Quantitative Research Design 229
nursing interventions in clinical practice. Many clini-
cal nursing interventions are not well defined; thus,
each nurse may use her or his own terminology to
describe a particular intervention. In addition, an inter-
vention tends to be applied differently in each case by
a single nurse and even less consistently by different
nurses. However, the quality of nursing interventions
has been greatly enhanced with the development of
the Nursing Interventions Classification by a team of
nurses at the University of Iowa.
The Nursing Interventions Classification
The Nursing Interventions Classification (NIC) is a
standardized language used to describe interventions
or treatments performed by nurses in research and
practice. Each intervention consists of a label, a defini-
tion, and a set of activities performed by nurses car-
rying out the intervention. The NIC was initiated by
the University of Iowa in Iowa City, IA, in 1987 (NIC,
2011). The intervention labels developed over the last
20 years were derived from nursing education and
practice. The research to develop the NIC was initi-
ated in 1987 and progressed through four phases that
overlapped in time: “Phase I: Construction of the Clas-
sification (1987-1992); Phase II: Construction of the
Taxonomy (1990-1995); Phase III: Clinical Testing
and Refinement (1993-1997); and Phase IV: Use and
Maintenance (1996-ongoing)” (Bulechek, Butcher, &
Dochterman, 2008, p. 5). The research methods used
to develop the classification included content analysis,
surveys, focus groups, similarity analysis, and hierar-
chical clustering. The NIC Taxonomy contained seven
domains: Domain 1: Physiological: Basic; Domain 2:
Physiological: Complex; Domain 3: Behavioral;
Domain 4: Safety; Domain 5: Family; Domain 6:
Health System; and Domain 7: Community. There are
a total of 30 classes under the seven domains (Bowles
& Naylor, 1996; Bulechek et al., 2008).
Tripp-Reimer, Woodworth, McCloskey, and Bu-
lechek (1996), in their analysis of the structure of the
NIC interventions, identified three dimensions: inten-
sity of care, focus of care, and complexity of care.
A high intensity of care is associated with the physi-
ological illness level of the patient and the emer-
gency nature of the illness. The dimension of intensity
of care includes indicators of (1) intensity (or acuity)
and (2) whether the care is typical or novel. The
dimension of focus of care addresses (1) the target
of the intervention, ranging from the individual to
the system; (2) whether the care action is direct or
on behalf of the patient; and (3) the continuum of
practice from independent to collaborative actions.
The dimension of complexity of care encompasses a
range of knowledge, skill, and urgency of the inter-
ventions (Bulechek et  al., 2008).
The interventions in the NIC have been subjected
to multiple studies examining the effects on different
populations and the effects of varying degrees of
intensity. The 5th edition of the Nursing Interventions
Classification, developed by faculty at the University
of Iowa, included 542 research-based interventions
(Bulechek et al., 2008). NIC development continues
through the Center for Nursing Classification & Clini-
cal Effectiveness located at the University of Iowa,
and you can email them with questions (classification-
center@uIowa.edu/; see Chapter 14 for more details
on NIC).
Currently, studies are being conducted to determine
the outcomes of each intervention and to establish
links between the intervention and outcomes at varying
points in time after the intervention has been imple-
mented. Outcomes that occur immediately following
the intervention are easiest to determine. However, the
most important outcomes may be those that occur after
a client has been discharged or several weeks or
months after the intervention. Table 11-1 provides
some of the most current examples of the research
related to the NIC and the Nursing Outcomes Classi-
fication (NOC) being conducted nationally and inter-
nationally. This information is critical for ensuring the
quality of care provided by nurses and justifying
nursing actions in a cost-conscious market (Doran,
2011). For a more extensive discussion of the impor-
tance of linking interventions with outcomes mea-
sures, see Chapter 13.
Designing an Intervention for a
Nursing Study
The therapeutic nursing intervention implemented in
a nursing study needs to be carefully designed, clearly
described, and well linked to the outcomes (dependent
variables) to be measured in the study. Each of these
dimensions must be considered to develop consis-
tency in the intervention. The intervention needs to
be provided consistently to all subjects. Thus, a pub-
lished study should document intervention fidelity,
which includes the detailed description of the essen-
tial elements of the intervention and the consistent
implementation of the intervention during the study
(Forbes, 2009; Morrison et al., 2009; Santacroce
et al., 2004). In some studies, you may need to develop
a step-by-step protocol in order to ensure the detail
and control the consistency of the study intervention.
Educational treatments or educational components of
treatments might be audio or video recorded for
consistency.
230 UNIT TWO  The Research Process

TABLE 11-1  Work in Nursing Related to the NIC and the Nursing Outcomes Classification (NOC)
Year Source
2011 Lee, E., Park, H., Nam, M., & Whyte, J. (2011). Identification and comparison of interventions performed by Korean
school nurses and U.S. school nurses using the Nursing Interventions Classification (NIC). Journal of School Nursing,
27(2), 93–101.
2011 Scherb, C. A., Head, B. J., Maas, M. L., Swanson, E. A., Moorhead, S., Reed, D., & Kozel, M. (2011). Most frequent
nursing diagnoses, nursing interventions, and nursing-sensitive patient outcomes of hospitalized older adults with heart
failure: Part 1. International Journal of Nursing Terminologies & Classifications, 22(1), 13–22.
2010 de Cordova, P., Lucero, R. J., Hyun, S., Quinlan, P., Price, K., & Stone, P. W. (2010). Using the Nursing Interventions
Classification as a potential measure of nurse workload. Journal of Nursing Care Quality, 25(1), 39–45.
2010 Lunney, M., McGuire, M., Endozo, N., & McIntosh-Waddy, D. (2010). Consensus-validation study identifies relevant
nursing diagnoses, nursing interventions, and health outcomes for people with traumatic brain injuries. Rehabilitation
Nursing, 35(4), 161–166.
2010 Smith, K. J., & Craft-Rosenberg, M. (2010). Using NANDA, NIC, and NOC in an undergraduate nursing practicum.
Nurse Educator, 35(4), 162–166.
2010 Solari-Twadell, P., & Hackbarth, D. P. (2010). Evidence for a new paradigm of the ministry of parish nursing practice
using the nursing intervention classification system. Nursing Outlook, 58(2), 69–75.
2009 Scherb, C. A., & Weydt, A. P. (2009). Work complexity assessment, nursing interventions classification, and nursing
outcomes classification: Making connections. Creative Nursing, 15(1), 16–22.
2009 Schneider, J. S., & Slowik, L. H. (2009). The use of the Nursing Interventions Classification (NIC) with cardiac patients
receiving home health care. International Journal of Nursing Terminologies & Classifications, 20(3), 132–140.
2009 Wong, E. (2009). Novel nursing terminologies for the rapid response system. International Journal of Nursing
Terminologies & Classifications, 20(2), 53–63.
2009 Wong, E., Scott, L. M., Briseno, J. R., Crawford, C. L., & Hsu, J. Y. (2009). Determining critical incident nursing
interventions for the critical care setting: A pilot study. International Journal of Nursing Terminologies &
Classifications, 20(3), 110–121.
2008 Schneider, J. S., Barkauskas, V., & Keenan, G. (2008). Evaluating home health care nursing outcomes with OASIS and
NOC. Journal of Nursing Scholarship, 40(1), 76–82.
2008 Sheerin, F. K. (2008). Diagnoses and interventions pertinent to intellectual disability nursing. International Journal of
Nursing Terminologies & Classifications, 19(4), 140–149.

The first step in designing an intervention should
be a thorough review of the clinical and research
literature related to the intervention. Because of the
scarcity of information in the literature on selected
nursing interventions, you may need to rely on your
personal knowledge emerging from expertise in clin-
ical practice. The nursing actions that are included in
the intervention must be spelled out sequentially so
that other nurses are able to follow the description
and provide the intervention in a consistent manner.
The intervention implemented in a study must be
consistent in areas such as (1) content, (2) intensity,
and (3) length of time. You need to review the
interventions that are provided in the NIC text to
help you develop an intervention protocol for a
study (Bulechek et al., 2008). This text provides
details for numerous interventions that have been
implemented in previous studies. You might find an
intervention to implement in your study with a dif-
ferent population, or you might use the format to
develop a detailed new intervention to be imple-
mented in your study. Also visit the NIC website for
additional information on interventions (http://www
.nursing.uiowa.edu/excellence/nursing_knowledge/
clinical_effectiveness/nic.htm/) (NIC, 2011).
The person or persons designated to implement the
study intervention must be trained for this process.
These individuals need to be trained in a precise way
to ensure that they are 90% to 100% consistent and
accurate in their implementation of the study treat-
ment. Often during the study, the implementation of
the treatment is evaluated with a protocol checklist to
ensure consistent implementation throughout the
study. If an intervention is complex, you may need to
employ a pilot study to refine the intervention so that
it can be applied consistently (Wyatt et al., 2010;
Yamada et al., 2010).
Pierce et al. (2011) provided a detailed discussion
of the intervention they implemented in their quasi-
experimental study to raise stroke awareness among
people in rural areas. The intervention they imple-
mented was the Facts for Action to Stroke Treatment
(FAST) educational intervention program to improve
402 participants’ knowledge about stroke. The FAST
 CHAPTER 11  Selecting a Quantitative Research Design 231
intervention was chosen because of its success in
teaching people to identify stroke symptoms. The fol-
lowing excerpt from the study describes the interven-
tion content and implementation process:
“The FAST educational program consisted of a
43-slide PowerPoint presentation developed from
materials from AHA [American Heart Association],
the ASA [American Stroke Association], a review of
the literature, and the CPSS [Cincinnati Prehospital
Stroke Scale]. FAST team members presenting the
program were registered nurse faculty members or
graduate students at the School of Nursing who used
a printed script to ensure the programs were equal
in content. The CPSS test highlighted in the presenta-
tion takes 1 minute to perform and is appropriate for
teaching to the public. CPSS includes looking for
asymmetry of the face when a person is instructed to
smile, arm drift when asked to raise both arms with
eyes closed for 10 seconds, and slurred or inappropri-
ate speech when asked to repeat a familiar saying.
The saying used in this program was ‘You can’t teach
an old dog new tricks.’ To help individuals remember
the components of the CPSS, this program made the
connection to the acronym FAST by pointing out they
should look at the Face, Arm, Speech, and if there
were abnormalities in any of these areas it was Time
to call 911 (see FAST components below).
F = Face, check for droop or asymmetry
A = Arm, look for drift or lack of movement
S = Speech, use common saying / listen for
slurred or garbled speech
T = Time to call 911 if any of these signs are
abnormal.” (Pierce et al., 2011, p. 85)
Pierce et al. (2011) selected a strong intervention
(FAST Educational Program) that had been used in
previous research and implemented it with a relative
new population of rural adults. The content of the
intervention was detailed and based on current research
and national organizations’ (AHA and ASA) materials.
The intervention was implemented consistently using
PowerPoint slides and a script. The study would have
been strengthened by a description of the training of
the faculty and students who implemented the inter-
vention. The researchers concluded that the FAST-
based program was an effective intervention for
teaching rural adults to recognize stroke symptoms.
Chapter 14 provides a detailed discussion of the devel-
opment and implementation of interventions and the
conduct of intervention research.
Quasi-experimental
Study Designs
Quasi-experimental and experimental designs examine
causality. The power of the design to accomplish this
purpose depends on the extent to which the actual
effects of the experimental treatment (the independent
variable) can be detected by measuring the dependent
variable. Obtaining an understanding of the true
effects of an experimental treatment or intervention
requires action to control threats to the validity of the
findings. Threats to design validity are controlled
through selection of subjects, control of the environ-
ment, manipulation of the treatment, and reliable and
valid measurement of the dependent variables. These
design validity threats are described in Chapter 10.
Experimental study designs, with their strict
control of variance, are the most powerful method of
examining causality. For many reasons, both ethical
and practical, however, experimental designs cannot
always be used in social science research. Quasi-
experimental study designs were developed to
provide alternative means of examining causality
in situations not conducive to experimental controls.
Campbell and Stanley first described quasi-
experimental designs as a group in 1963, when only
experimental designs were considered of any worth.
Cook and Campbell expanded this description in 1979
and Shadish, Cook, and Campbell (2002) provide the
most current discussion of quasi-experimental and
experimental designs. Quasi-experimental designs
facilitate the search for knowledge and examination of
causality in situations in which complete control of a
study design is not possible. These designs have been
developed to control as many threats to validity as
possible in a situation in which at least one of the three
components of true experimental design (randomiza-
tion, comparison groups, and controlled manipulation
of the treatment) is lacking.
There are differences of opinion in nursing about
the classification of a particular study as quasi-
experimental or experimental. The experimental
designs emerged from a logical positivist perspective
with the purpose of determining cause and effect. The
focus is to determine differences between or among
groups using statistical analyses on the basis of deci-
sion theory. The true experimental design (from a
logical positivist view) requires the use of random
sampling to obtain subjects, random assignment to
control and experimental groups, rigorous control of
the treatment, a highly controlled study environment,
and designs that control threats to validity (Shadish
et al., 2002).
232 UNIT TWO  The Research Process
A less rigorous type of experimental design is
referred to as the comparative experimental design.
Researchers in both nursing and medicine are using it
for clinical situations in which the expectation of
random sampling is difficult, if not impossible, to
achieve. These studies use convenience samples with
random assignment to groups. For example, random-
ized controlled trials (RCT) usually do not use ran-
domly obtained samples but tend to be considered
experimental in nature. These studies are classified as
experimental because they have internal validity if the
two groups (experimental and control) are comparable
on variables important to the study, even though there
are biases in the original sample. However, these
designs do not address threats to statistical conclusion
validity and threats to external validity by the nonran-
dom sample. Threats to external validity have not, in
the past, been considered a serious concern because
they affect not the claim that the treatment caused a
difference but rather the ability to generalize the find-
ings. The importance of external validity, although
discounted in the past, is taking on greater importance
in the current political and health policy climate.
Chapter 13, on outcomes research, explores the con-
cerns some researchers and clinicians have about the
validity of RCTs.
Random Assignment to Groups
Random assignment to groups is a procedure used
to assign subjects to treatment or control groups ran-
domly. Random assignment is most commonly used
in nursing and medicine to assign subjects obtained
through convenience sampling methods to groups for
purposes of comparison. Random assignment used
without random sampling is purported to decrease the
risk of bias in the selection of groups. However, Otten-
bacher (1992) performed a meta-analysis to examine
the effect of random assignment versus nonrandom
assignment on outcomes. The results failed to reveal
significant differences in these two sampling tech-
niques. He suggested that previous assumptions about
design strategies should be empirically tested. The
term RCT usually means that the study used random
assignment of subjects to groups, not that the sample
was obtained through random sampling methods.
Traditional approaches to random assignment
involve using a random numbers table or flipping an
unbiased coin to determine group assignment.
However, these procedures can lead to unequal group
sizes and thus a decrease in power. Hjelm-Karlsson
(1991) suggested using what is referred to as a biased
coin design to randomly assign subjects to groups.
With this technique, selection of the group to which a
particular subject will be assigned is biased in favor
of groups that have smaller sample sizes at the point
of the assignment of that subject. This strategy is par-
ticularly useful when assignment is being made to
more than two groups. The researcher can complete
calculations for the sequencing of assignment to
groups before collecting data, thus freeing the
researcher for other activities during this critical
period. Hjelm-Karlsson (1991) suggested using cards
to make group assignments. The subject numbers and
random group assignments are written on cards. As
each subject agrees to participate in the study, the next
card is drawn from the stack, indicating that subject’s
number and group assignment. These activities could
also be accomplished with the computer.
Stout, Wirtz, Carbonari, and Del Boca (1994) sug-
gested a similar strategy they referred to as urn ran-
domization, which they described as follows:
“One would begin the study with two urns, each urn
containing a red marble and a blue marble. There is
one urn for each level of the stratifying variable; that
is, in this example there is an urn for severely ill
patients and another urn for the less severe[ly ill]
patients. When a subject is ready for randomization,
we determine whether or not he/she is severely ill
and consult the corresponding urn. From this urn (say,
for the severely ill group) we randomly select one
marble and note its color. If the marble is red we
assign the patient to Treatment A. Then we drop that
marble back into the urn and put a blue marble into
the urn as well. This leaves the ‘severely ill’ urn with
one red and two blue marbles. The next time a
severely ill patient shows up, the probability that he/
she will be assigned to Group B will be 2/3 rather than
2, thus biasing the selection process toward balance.
A similar procedure is followed every time a severely
ill subject presents for randomization. After each
subject is assigned, the marble chosen from the urn
is replaced together with a marble of the opposite
color. The urn for the less severely ill group is not
affected. If a low-severity patient presents for the
study, that patient’s probability of assignment to
either treatment is not affected by the assignment of
patients in the other stratum. To some extent, urn
randomization can be tailored to maximize balancing
or to maximize randomization.” (Stout et al., 1994,
p. 72)
Stout and colleagues (1994) also provided strate-
gies for balancing several variables simultaneously
during random assignment.
 CHAPTER 11  Selecting a Quantitative Research Design 233
Matthews, Cook, and Terada (2010) tested the
ability of urn randomization versus simple random
assignment to groups in producing balanced groups
with small sample sizes. They conducted simulated
randomizations 10 times, developing “sample size
scenarios of 20, 40, 60 (group sizes of 10, 20, and 30,
respectively), for 30 trials in total. For groups of
20-30, urn surpassed simple randomization in the
equal distribution of confounding variables between
groups, leading to effects of these variables that were
both smaller on average and more consistently close
to zero over multiple trials” (Matthews et al., 2010, p.
243). These researchers concluded that the urn method
was easy to implement and has the advantages of
unpredictability in assignment of participants to
groups and decreased potential for investigator bias.
This article includes an illustration of the urn method
to facilitate your understanding of the process for
assigning study participants to groups (see Figure 1 in
Matthews et al., 2010, p. 248).
Schlairet and Pollock (2010) used random assign-
ment of participants to groups in their study of under-
graduate nursing students’ knowledge obtained with
traditional versus simulated clinical experiences. The
following excerpt from the study describes the design
with participant group assignment:
“This intervention study used a 2x2 crossover design
and equivalence testing to explore the effects of simu-
lated clinical experiences on undergraduate students’
(n = 74) knowledge acquisition in a fundamentals of
nursing course. Following random assignment, stu-
dents participated in laboratory-based simulated clini-
cal experiences with high-fidelity human patient
simulators and traditional clinical experiences and
completed knowledge pretests and posttests.”
(Schlairet & Pollock, 2010, p. 43)
These researchers administered the knowledge
pretest to establish equivalence between the groups.
After the simulated clinical experience and traditional
clinical experience, the students were administered a
posttest. Analysis revealed a significant knowledge
gained associated with both simulated and traditional
clinical experiences, with the groups’ knowledge
scores being statistically significantly equivalent.
Thus, the researchers concluded that the simulated
clinical experience was as effective as the traditional
clinical experience in this study. The description of the
study design would have been strengthened by a dis-
cussion of how random assignment of participants to
groups was accomplished.
Each of the quasi-experimental designs described
in this section involves threats to validity owing to
constraints in controlling variance. Some achieve
greater amounts of control than others. When choos-
ing a design for a study, you must select the design
that offers the greatest amount of control possible
within your study situation (Shadish et al., 2002).
Even the first designs described in this section, which
have low power in terms of establishing causality, can
provide useful information from which to design later
studies.
Control and Comparison Groups
Quasi-experimental and experimental studies include
an experimental or intervention group that receives the
treatment or intervention and a control that receives
no treatment. Control groups, traditionally used in
experimental studies, are selected randomly from the
same population as the experimental group and receive
no treatment. Use of a control group increases the
ability of the researcher to detect differences between
groups in the real world. Thus, control groups reduce
the risk of error. Control groups are rarely used in
nursing or medical studies because of requirements
related to consent, ethical issues regarding withhold-
ing treatment, and the difficulty of acquiring sufficient
potential subjects from which to select a sample
(Shadish et al., 2002).
Comparison groups are usually selected through
the use of convenience sampling rather than random
sampling methods. There are four types of compari-
son groups: (1) groups that receive no treatment; (2)
groups that receive a placebo treatment; (3) groups
that receive the “usual treatment” or standard care; and
(4) groups that receive a second experimental treat-
ment or a different treatment dose for comparison with
the first experimental treatment (e.g., clinical trials of
drug effectiveness). As a researcher, you should clarify
the type of comparison group you are using.
When a study uses a comparison group that receives
no treatment, demonstrating statistical significance is
easier because there is less variation in the treatment
and a greater difference between the two groups.
Placebo treatments provide consistency in the com-
parison group, provide less difference between groups
than in no-treatment comparison groups, and would
be unethical in some nursing studies. “Usual treat-
ment” is the care routinely provided by the healthcare
system. However, usual treatment is uneven and thus
is often not standardized for all patients. Thus, provi-
sion of care may vary from one patient to another
depending on the availability of nursing staff and the
intensity of care demands being made on nurses at the
234 UNIT TWO  The Research Process
time the care is provided. Some patients may receive
little or no care, whereas others may receive consider-
ably more or better care. There will likely be a greater
amount of difference between patients who received
little or no care and patients in the experimental group,
and less difference between patients in the “usual care
group” who received considerably more care and the
experimental group. This wide variation reduces the
effect size of the experimental treatment, increases the
variance, and decreases the possibility of obtaining a
significant difference between groups. The researcher
should carefully spell out “usual or standard care” and
the degree of variation in the care in the facility in
which the study is being conducted.
Nonequivalent Comparison Group Designs
A comparison group is one in which the groups are
not selected by random means. Some groups are
more nonequivalent than others, and some quasi-
experimental designs involve using groups (compari-
son and treatment or intervention) that have evolved
naturally rather than being developed randomly. For
example, the treatment group might include students
registered for an 8:00 am class in a university, and the
comparison group might be students registered for a
7:00 pm class. These groups cannot be considered
equivalent because the individuals in the comparison
group may be different from individuals in the treat-
ment group. Allowing individuals to select the group
they will be in (treatment or comparison) rather than
being randomly assigned by the researcher also
increases the threat to design validity. For example,
allowing subjects to select either the treatment group
that receives an exercise program or the comparison
group that has no exercise program threatens the valid-
ity of the study. The subjects selecting the treatment
of the exercise program are different from those select-
ing no exercise program.
The approach to statistical analysis is problematic
in quasi-experimental designs. Although many
researchers use the same approaches to analysis as are
used for experimental studies, the selection bias inher-
ent in nonequivalent comparison groups makes this
practice questionable. Reichardt (1979) recommended
using additional statistical analyses to examine the
data from various perspectives and to compare levels
of significance obtained from each analysis. As a
researcher, you must carefully assess the potential
threats to validity in interpreting statistical results,
because statistical analysis cannot control for threats
to validity (Munro, 2005; Shadish et al., 2002). The
following sections describe examples of nonequiva-
lent comparison group designs.
One-Group Posttest-Only Design
The one-group posttest-only design is referred to as
pre-experimental rather than quasi-experimental
because of its weaknesses and the numerous threats to
validity it involves. It is inadequate for making causal
inferences (see Figure 11-11). Usually in this design,
no attempt is made to control the selection of subjects
who receive the treatment (the experimental group). It
is difficult to justify generalizing findings beyond
those tested. The group is not pretested; therefore,
there is no direct way to measure change. The
researcher cannot claim that posttest scores were a
consequence (effect) of the treatment if scores before
the treatment are unknown. Because there is no com-
parison group, one does not know whether groups not
receiving the treatment would have similar scores on
the dependent variable. The one-group posttest-only
design is used more commonly in evaluation than in
research.
Cook and Campbell (1979) suggested situations in
which the one-group posttest-only design can be
appropriate and adequate for inferring causality. For
example, the design could be used to determine that a
single factory’s use of vinyl chloride is causing an
increase in the rate of neighborhood and employee
cancers. The incidence of cancer in the community at
large is known. The fact that vinyl chloride causes
cancer and the types of cancer it causes are also
known. These norms would then take the place of the
pretest and the comparison group. Thus, to use this
design intelligently, one must know a great deal about
the causal factors interacting within the situation
(Shadish et al., 2002). This is not the usual situation
in nursing studies.
Posttest-Only Design with a Comparison Group
Although the posttest-only design with a compari-
son group offers an improvement on the previous
design because of the addition of a nonequivalent com-
parison group, it is still referred to as pre-experimental
(see Figure 11-12). The addition of a comparison
group can lead to a false confidence in the validity of
the findings. Selection threats are a problem with both
groups. The lack of a pretest remains a serious impedi-
ment to defining change. Differences in posttest scores
between groups may be caused by the treatment or by
differential selection processes (Shadish et al., 2002).
One-Group Pretest-Posttest Design
Another pre-experimental design, the one-group
pretest-posttest design, is one of the more commonly
used designs. However, it has such serious
weaknesses that findings are often uninterruptable
 CHAPTER 11  Selecting a Quantitative Research Design 235

Manipulation of Measurement of
independent variable dependent variable(s)
TREATMENT POSTTEST

Treatment—often ex post facto

Experimental group—those who receive the treatment and the posttest

Pretest—inferred—norms of measures of dependent variable(s) of population

from which experimental group taken

Comparison group—implied—norms of measures of dependent variable(s) of

population from which experimental group taken

Approach to analysis: • comparison of posttest scores with inferred norms

• confident inferences about change

Uncontrolled threats to validity: • no link between treatment and change

• no comparison group
• maturation
• undetected confounding variables
• inability to access threats to validity

Figure 11-11  One-group posttest-only design.

Manipulation of Measurement of
independent variable dependent variable(s)

Experimental group TREATMENT POSTTEST

Nonequivalent comparison group POSTTEST

Treatment—often ex post facto

may not be well defined

Experimental group—those who receive the treatment and the posttest

Pretest—inferred—norms of measures of dependent variable(s) of population

from which experimental group taken

Comparison group—not randomly selected—tend to be those who naturally in

the situation do not receive the
treatment

Approach to analysis: • comparison of posttest scores of experimental and

comparison group
• comparison of posttest scores with norms

Uncontrolled threats to validity: • no link between treatment and change

• no pretest
• selection

Figure 11-12  Posttest-only design with a comparison group.

236 UNIT TWO  The Research Process

Measurement of Manipulation of Measurement of

dependent variable(s) independent variable dependent variable(s)
Experimental POSTTEST
PRETEST TREATMENT
group group

Treatment—greater research control

usually not ex post facto

Experimental group—greater researcher control

expected to serve as comparison group

Comparison group—pretest scores of treatment group expected to serve as control

Approach to analysis: • comparison of pretest and post-test scores

Uncontrolled threats to validity: • history

• statistical regression
• maturation
• testing
• instrumentation

Figure 11-13  One-group pretest-posttest design.

(see Figure 11-13). Pretest scores cannot adequately

serve the same function as a comparison group. Events
can occur between the pretest and posttest that alter
responses to the posttest. These events then serve as
alternative hypotheses to the proposal that the change
in posttest scores is due to the treatment. Posttest
scores might be altered by (1) maturation processes;
(2) administration of the pretest; and (3) changes in
instrumentation. Additionally, subjects in many studies
using this design are selected on the basis of high or
low scores on the pretest. Thus, there is an additional
threat that changes in the posttest may be due to
regression toward the mean. Lim, Chiu, Dohrmann,
and Tan (2010) implemented a one-group pretest-
posttest design to study medication management by
registered nurses (RNs). The following study excerpt
describes their design:
“This exploratory study used a non-randomized
pre- and post-test one group quasi-experimental
design without comparators. It comprised a 23-item
knowledge-based test questionnaire, one-hour teach-
ing session, and a self-directed learning package. The
volunteer sample was RNs from residential aged
care facilities, involved in medication management.
Participants sat a pre-test immediately before the
education, and post-test 4 weeks later (same
questionnaire)…. Pre-test sample n = 58, post-
test n = 40, attrition rate of 31%…. Descriptive sta-
tistical data analysis of overall pre- and post-test
incorrect responses showed: pre-test proportion
of incorrect responses = 0.40; post-test proportion
of incorrect responses = 0.27; Z-test comparing pre-
and post-tests scores of incorrect responses = 6.55
and one-sided p-value = 2.8E-11 (p < 0.001).” (Lim
et al., 2010, p. 98)
Lim et al. (2010) concluded that the pretest identi-
fied knowledge deficits in medication management
and adverse drug reactions in the elderly and the post-
test indicated statistically significant improvement in
the RNs’ knowledge. The researchers recognized the
limitations of their study design by calling it explor-
atory and also identified the fairly high attrition rate,
31%. The use of the same questionnaire as a pretest
and posttest could also threaten the study validity
because the change in the posttest scores might be due
to memory of questionnaire items in addition to the
effect of the educational treatment. Based on these
limitations, Lim et al. recommended further studies
with larger samples and stronger design to determine
the impact of the educational treatment on the RNs’
knowledge of medication management. The addition
of a nonequivalent comparison group, as described in
the next design, can greatly strengthen the validity of
the findings.
 CHAPTER 11  Selecting a Quantitative Research Design 237

Measurement of Manipulation of Measurement of

dependent variable(s) independent variable dependent variable(s)

Experimental Pretest Treatment Posttest

group

Nonequivalent Pretest Posttest

comparison group

Treatment—experimental group
comparison group not treated or receives standard or routine care

Comparison group—not randomly selected

Approach to analysis: • Examine difference between comparison and experimental pretest

• Examine difference between pretest and posttest
• Examine difference between comparison and experimental posttest

Uncontrolled threats to validity: • Selection-maturation

• Instrumentation
• Differential statistical regression
• Interaction of selection and history

Figure 11-14  Pretest and posttest design with a comparison group.

Pretest and Posttest Design with a

Comparison Group
The pretest and posttest design with a comparison
group is the most commonly used design in
social science research (see Figure 11-14). This quasi-
experimental design is the first design discussed here
that is generally interpretable. The uncontrolled threats
to validity are primarily due to the absence of random-
ization and, in some studies, the inability of the
researcher to manipulate the treatment. Cook and
Campbell (1979) offered a detailed discussion of
the effects of these threats on interpreting study
findings.
Variations in this design include the use of (1)
proxy pretest measures (a different pretest that corre-
lates with the posttest); (2) separate pretest and post-
test samples; and (3) pretest measures at more than
one time interval. The first two variations weaken the
design, but the last variation greatly strengthens it. In
some studies, the comparison group consists of
patients cared for before a new treatment was initiated.
Data on this comparison group are obtained through
chart audit or from electronic databases owned by the
facility. Obviously there is no opportunity to control
the quality of the data obtained through chart audit.
Thus, this strategy weakens the design.
Costanzo, Walker, Yates, McCabe, and Berg (2006)
used a pretest-posttest comparison group design in
their study of physical activity counseling for older
women. They described their design as follows:
“Physical inactivity is a major factor in increasing
women’s risk for chronic disease, disability, and pre-
mature mortality. This study compared the effective-
ness of five behavioral counseling (BC) sessions with
a comparison group receiving one BC session based
on the five A’s (ask, advise, assist, arrange, and agree)
to increase moderate-intensity physical activity,
muscle strengthening, and stretching activity. The
health promotion model provided the framework for
the intervention. A pretest/posttest comparison
group design was used, with random assignment of
46 women recruited from an urban Midwestern com-
munity. A significant group interaction was found only
for cardiorespiratory fitness (p < 0.001). Significant
time effects were found (p < 0.001) for both groups
in increasing handgrip, leg strength, and flexibility. BC
is a promising intervention to achieve physical activity
behavior change with older women.” (Costanzo
et al., 2006, p. 786)
Pretest and Posttest Design with Two
Comparison Treatments
The two-treatment design is used when two experi-
mental treatments are being compared to determine
which is the more effective. In most cases, this design
is used when one treatment is the currently identified
treatment of choice and the researcher has identified a
treatment that might lead to even better outcomes
238 UNIT TWO  The Research Process
Measurement of
dependent variable(s)
Experimental Pretest
group 1
Experimental Pretest
group 2
T1 = Treatment #1
T2 = Treatment #2
Figure 11-15  Pretest and posttest design with two comparison treatments.
(see Figure 11-15). This design is strengthened by the
addition of one or more of the following: a no-treatment
group, a placebo-treatment group, or a routine or stan-
dard care group (see Figure 11-16).
Dudgeon et al. (2010) conducted a quasi-experiment
study of HIV-infected men using a pretest and posttest
design with two comparison treatment groups and a
control group (see Figure 11-16). The two treatment
groups participated in moderate- and low-intensity
exercise interventions, and the control group partici-
pated in no activity. This study had a strong design,
and the moderate-intensity exercise was found to be
the most effective in improving circulating hormones
and cytokines in the HIV-infected men. The research-
ers provided a clear discussion of their study design
that is presented in the following excerpt:
“Exercise has the potential to impact disease by alter-
ing circulating anabolic and catabolic factors. It was
the goal of this study to determine if two different
regimes of low-intensity and moderate-intensity
exercise affected circulating levels of anabolic and
catabolic factors in HIV-infected men. Exercise-naïve,
HIV-infected men, medically cleared for study partici-
pation, were randomized into one of the following
groups: a moderate-intensity group (MOD), who
completed 30 minutes of moderate-intensity aerobic
training followed by 30 minutes of moderate-intensity
resistance training; a low-intensity group (LOW), who
completed 60 minutes of treadmill walking; or a
control group (CON), who attended the clinic but
participated in no activity. Blood and saliva samples
were collected at selected time points before, during,
and after each of the 3 required sessions. Compared
with baseline, the MOD group (n = 14) had a 135%
Manipulation of Measurement of
independent variable dependent variable(s)
T1 Posttest
T2 Posttest
increase in growth hormone (GH) (p < 0.05) and a
34% decrease in cortisol (CORT) (p < 0.05) at the
post time point, a 31% increase in interleukin-6 (IL-6)
(p < 0.05) at 30 minutes post exercise, a 23% increase
in IL-6 (p < 0.05), and a 13% decrease in soluble
tumor necrosis factor receptor 2 (sTNFrII) (p < 0.05)
at 60 minutes post exercise. The LOW (n = 11) group
had a 3.5% decrease in sTNFrII (p < 0.05) at 30
minutes post exercise compared with baseline and
49% decrease (p < 0.05) in GH at 60-minutes post
exercise. The CON group (n = 13) had a decrease in
GH at 30-minutes (62%, p < 0.05) and 60-minute
(61%, p < 0.05) post exercise [group participated in
no exercise] compared with baseline. The increase in
GH from baseline to post was greater in the MOD
group (p < 0.05) than the other groups and the
decrease in CORT from pre to post was greater in
the MOD group (p < 0.05) than in the other groups.
These data suggest that individual sessions of both
low-intensity and moderate-intensity exercise can
alter circulating anabolic and catabolic factors in HIV-
infected men. The changes in the MOD group present
potential mechanisms for the increases in lean tissue
mass seen with resistance exercise training.”
(Dudgeon et al., 2010, p. 560)
Pretest and Posttest Design with a
Removed Treatment
In some cases, gaining access to a comparison group
is not possible. The removed-treatment design with
pretest and posttest creates conditions that approxi-
mate the conceptual requirements of a control group
receiving no treatment. The design is basically a one-
group pretest-posttest design. However, after a delay,
a third measure of the dependent variable is taken,
followed by an interval in which the treatment is
 CHAPTER 11  Selecting a Quantitative Research Design 239

Measurement of Manipulation of Measurement of

dependent variable(s) independent variable dependent variable(s)

Experimental Pretest T1 Posttest

group 1

Experimental Pretest T2 Posttest

group 2

Comparison Pretest Usual Posttest

group or routine care

T1 = Treatment #1
T2 = Treatment #2

Figure 11-16  Pretest and posttest design with two comparison treatments and a usual or routine care group used as a comparison group.

Sample Experimental Dissipation of Controlled condition

condition treatment effect (treatment withheld)

#1 M(1) T M(2) M(3) M(4)

Equivalent time intervals

M = measurement of dependent variable(s)

T = treatment–manipulation of the independent variable

Approach to analysis: • Comparison of changes in dependent variable

• Scores between measurement points

Uncontrolled threats to validity: • Statistical conclusion validity

• Construct validity of the cause
• Selection

Figure 11-17  Pretest and posttest design with a removed treatment. M(1), pretest; M(2), posttest; M(3), pretest of controlled condition; M(4), posttest
of controlled condition.

removed, followed by a fourth measure of the depen-

dent variable (see Figure 11-17). The periods between
measures must be equivalent. In nursing situations, the
researcher must consider the ethics of removing an
effective treatment. Even if doing so is ethically
acceptable, the response of subjects to the removal
may make interpreting changes difficult.
It is difficult in CINAHL (Cumulative Index to
Nursing and Allied Health Literature) and MEDLINE
to locate examples of studies using removed-treatment
designs because of the search process required. A
search in PsychInfo located one study: Schneider
(1998) described a study of the effects of virtual reality
on symptom distress in children receiving cancer che-
motherapy, as shown in the following excerpt:
“An interrupted time series design with removed
treatment was used to answer the following research
questions: (1) Is virtual reality an effective distraction
intervention for reducing chemotherapy related
symptom distress in children? And (2) Does virtual
reality in children have a lasting effect? Hypotheses:
(1) There will be differences in measures of symptom
distress in a single group of children with cancer who
receive a virtual reality distraction intervention during
the second chemotherapy treatment and who receive
no virtual reality intervention during the first and third
chemotherapy treatments. The convenience sample
consisted of 11 children receiving outpatient
240 UNIT TWO  The Research Process

causal variable must be rigorously defined to allow

chemotherapy at a clinical cancer center. Measures of
symptom distress were obtained at nine time points
during three consecutive chemotherapy treatments.
Four indicators were used to measure the dependent
variable of symptom distress. The Symptom Distress
Scale (SDS) (McCorkle & Young, 1978) was consid-
ered a general indicator. Specific indicators of
symptom distress included the State-Trait Anxiety
Inventory for Children (STAIC C-1) (Spielberger
et al., 1978) and single item indicators for nausea and
vomiting.” (Schneider, 1998, p. 2126)
differential predictions of directions of effect. To be
maximally interpretable, the following two groups
must be added: (1) a placebo control group in which
the treatment is not expected to affect the dependent
variable and (2) a no-treatment control group to
provide a baseline. This design is not commonly used
in nursing, so an older study is presented as an
example.
McConnell (1976) used a reversed-treatment
design to test how knowledge of the results affected a
subject’s attitude toward a motor learning task, as
described in the following excerpt:

Schneider (1998) found that the use of a virtual

reality distraction intervention did decrease symptom
distress in children receiving outpatient chemother-
apy. However, the study had a very small sample, and
the researcher recommended replication with a larger
sample using additional clinical cancer centers.
Pretest and Posttest Design with a
Reversed Treatment
The reversed-treatment nonequivalent control
group design with pretest and posttest introduces two
independent variables—one expected to produce a
positive effect and one expected to produce a negative
effect (see Figure 11-18). There are two experimental
groups, each exposed to one of the treatments. The
design tests differences in response to the two treat-
ments. This design, because of its high construct valid-
ity of the cause, is more useful for theory testing than
the no-treatment control group design. This means that
there are strong theoretical sources proposing that spe-
cific treatments cause specific effects. The theoretical
The study “tested the hypotheses that a group which
has the greatest number of gains in performance in
successive trial scores of a motor task will develop a
more positive attitude toward the task, and that a
group which has the greatest number of gains in per-
formance in successive trial scores will show the
greatest change in an already formed attitude. Twelve
male and 12 female physical education majors were
randomly divided into 2 groups. Each subject per-
formed 20 trials of 15 seconds each on a rotary
pursuit task, read the directions for the completion
of the attitude measuring instrument, and then com-
pleted the instrument. This series of activities was
repeated a 2nd time. The difference in the treatment
of the 2 groups occurred in the knowledge of results
(KR: i.e., time on target). The 1st group received its
KR during the 1st 20 trials to the full second; during
the 2nd 20 trials, this group received its KR to .01
second. The other group received the reverse

Measurement of Manipulation of Measurement of

dependent variable(s) independent variable dependent variable(s)

Experimental Pretest Proposed positive Posttest

group 1 effective treatment

Experimental Pretest Proposed negative Posttest

group 2 effective treatment

Approach to analysis: • Comparison of group 1 and group 2 pretest

• Comparison of group 1 and group 2 posttest
• Comparison of changes between pretest and
posttest between groups

Uncontrolled threats to validity: • Statistical conclusion validity

Figure 11-18  Pretest and posttest design with a reversed treatment.

 CHAPTER 11  Selecting a Quantitative Research Design 241

Some threats, however, are particularly problem-

treatment. The difference in treatment caused the
subjects in the group being given KR to .01th of a
second to achieve more gains in performance than
those whose KR was to the full second. Further analy-
ses supported both hypotheses.” (McConnell, 1976,
p. 394)
Interrupted Time-Series Designs
The interrupted time-series designs are similar to
descriptive time designs except that a treatment is
applied at some point in the observations. Time-series
analyses have some advantages over other quasi-
experimental designs. First, repeated pretest observa-
tions can assess trends in maturation before the
treatment. Second, the repeated pretest observations
allow measures of trends in scores before the treat-
ment, decreasing the risk of statistical regression,
which would lead to misinterpretation of findings. If
you keep records of events that could influence sub-
jects in your study, you can determine whether histori-
cal factors that could modify responses to the treatment
were in operation between the last pretest and the first
posttest.
atic in time-series designs. Record-keeping proce-
dures and definitions of constructs used for data
collection tend to change over time. Thus, maintaining
consistency can be a problem. The treatment can result
in attrition so that the sample before treatment may be
different in important ways from the post-treatment
group. Seasonal variation or other cyclical influences
can be interpreted as treatment effects. Therefore,
identifying cyclical patterns and controlling for them
are critical to the analysis of study findings (Shadish
et al., 2002).
McCain and McCleary (1979) have suggested
using the autoregressive integrated moving average
(ARIMA) statistical model to analyze time-series
data. The ARIMA is a statistical model that has some
distinct advantages over regression analysis tech-
niques. For adequate statistical analysis, at least 50
measurement points are needed; however, Cook and
Campbell (1979) believe that ARIMA with even small
numbers of measurement points can provide better
information than that obtained in cross-sectional
studies. The numbers of measurements of the depen-
dent variables (M) shown in the designs illustrated in
Figures 11-19 through 11-21 are limited by space.

Measurement of Manipulation of Measurement of

dependent variable(s) independent variable dependent variable(s)

Pretests Treatment Posttests

Experimental
group M(1) M(2) M(3) M(4) T M(5) M(6) M(7) M(8)

Subjects: May remain the same across the study or different individuals may be
selected at each measurement period (as in the descriptive
time-dimensional designs)

Pretest dependent variable measures: In addition to researcher-initiated tools,

measures may be obtained from archival data sources such as patient
records, scores available in student records, or employee records.

Treatment: May be a key event occurring at one specific time that is expected to modify
subject responses to the dependent variable; that is, curriculum change,
administrative change, change in nursing care of a particular type of patient.
The treatment tends to continue after implementation rather than occurring and
then being withdrawn.
Approach to analysis: • Changes in trends of scores before and after treatment

Uncontrolled threats to validity: • History

• Seasonal trends
• Instrumentation
• Selection
• Mono-operation bias
• Cyclical influences interpreted as treatment effects

Figure 11-19  Simple interrupted time-series design. M(1) through M(8), Measurement of dependent variable(s).
242 UNIT TWO  The Research Process

Measurement of Manipulation of Measurement of

dependent variable(s) independent variable dependent variable(s)

GROUPS Pretests Treatment Posttests

Experimental group M(1) M(2) M(3) M(4) T M(5) M(6) M(7) M(8)
Comparison group M(1) M(2) M(3) M(4) M(5) M(6) M(7) M(8)

Subjects: May remain the same across the study or may be part of a group in which
some individuals may change (e.g., class, work group, patient teaching,
or patient care group).

Pretest dependent variable measures: In addition to researcher-initiated tools,

measures may be obtained from archival data sources.

Treatment: May be a key event occurring at one specific time that is expected to modify
experiment subject’s responses to dependent variable measures. Treatment
tends to continue after implementation rather than occurring and then
being withdrawn.

Approach to analysis: • Changes in trends of scores before and after treatment

• Comparison of trends in experimental and comparison groups
• Temporal persistence of treatment effects

Uncontrolled threats to validity: • Selection and history interaction

• As experimental and comparison groups increase in
noncompatibility, threats to validity increase
• Interaction of populations and treatment
• Cyclical influences interpreted as treatment effects

Figure 11-20  Interrupted time-series design with a no-treatment comparison group.

Pretest T Posttest Pretest T Posttest etc.

Single sample
M(1) M(2) T M(3) M(4) T M(5) M(6) T M(7) M(8) T M(9) M(10)

Treatment: • Provided repeatedly

T= • Effects must dissipate rapidly
• Should be scheduled randomly

Approach to analysis: • Powerful for inferring casual effects

• Differences in pretest and posttest scores must be in
opposite directions to be interpretable

Figure 11-21  Interrupted time-series design with multiple treatment replications. M(1) through M(8), Measurement of dependent variable(s).

They are not meant to suggest limiting measures to
the numbers shown.
Simple Interrupted Time-Series Design
The simple interrupted time-series design is similar
to the descriptive time-series study, with the addition
of a treatment that occurs or is applied (interrupts the
time series) at a given point in time (see Figure 11-19).
The treatment, which in some cases is not completely
under the control of the researcher, must be clearly
defined. There is no control or comparison group in
this design. The use of multiple methods to measure
the dependent variable greatly strengthens the design.
Threats that are well controlled by this design are
maturation and statistical regression. van Doormaal
et al. (2009) implemented an interrupted time-series
 CHAPTER 11  Selecting a Quantitative Research Design 243
design to determine the effect of a Computerized
Physician Order Entry system with basic Clinical
Decision Support (CPOE/CDSS) on the incidence of
medication errors (MEs) and preventable adverse drug
events (PADEs). The outcome measurements included
percentage of medication orders with one or more
MEs and the percentage of patients with one or more
PADEs. The following study excerpt describes the
study design:
Design
“The study was set up as an interrupted time series
that is characterized by a series of measurements
over time interrupted by an intervention. In this study
the intervention was the implementation of a Com-
puterized Physician Order Entry system in combina-
tion with a basic Clinical Decision Support System
(CPOE/CDSS). Data collection took place during a
5-month pre-implementation period (during which
the hand-written medication order system continued
to be used) and during a 5-month post-implementations
period (when the CPOE/CDSS system continued to
be used.) The post-implementation data collection
period started 8 weeks after finishing the implemen-
tation process in order to make sure that initial prob-
lems were solved.” (van Doormaal et al., 2009,
p. 817)
These researchers found that 55% of the medica-
tion orders contained at least one ME prior to the
implementation of the CPOE/CDSS, and 17% post
intervention. Thus, the implementation of the CPOE/
CDSS led to a significant immediate reduction of
40.3% in medication orders with one or more errors.
Pre-implementation, the mean percentage of admitted
patients experiencing at least one PADE was 15.5%,
as opposed to 7.3% post implementation. However,
when the use of interrupted time-series design is con-
sidered, the immediate change was not significant
(−0.42%; 95% CI: −15.52%, 14.68%) because of the
observed underlying negative trend and could not be
attributed to the introduction of CPOE/CDSS. van
Doormaal et al. (2009, p. 816) concluded that the
“CPOE/CDSS reduces the incidence of medication
errors. However, a direct effect on actual patient harm
(PADEs) was not demonstrated.”
Interrupted Time-Series Design with a
No-Treatment Comparison Group
The addition of a comparison group to the interrupted
time-series design greatly strengthens the validity of
the findings. The comparison group allows the
researcher to examine the differences in trends between
groups after the treatment and the persistence of treat-
ment effects over time (see Figure 11-20). Although
the treatment may continue (e.g., a change in nursing
management practices or patient teaching strategies),
the initial response to the change may differ from later
responses.
Zhang, Adams, Ross-Degnan, Zhang, and Soume-
rai (2009, p. 520) implemented an interrupted time-
series design with a comparison group to determine
the impact of a prior-authorization policy in Maine on
the ordering of second-generation antipsychotic and
anticonvulsant for patients with bipolar disorder, how
often they were discontinued, and pharmacy costs
among Medicaid beneficiaries.” The excerpt from the
study abstract describes the study design, key results,
and conclusions:
“Methods: Using Medicaid and Medicare utilization
data for 2001-2004, the authors identified 5,336
patients with bipolar disorder in Maine (study group)
and 1,376 in New Hampshire (comparison group).
With an interrupted time-series and comparison
group design, longitudinal changes were measured in
second-generation antipsychotic and anticonvulsant
use; survival analysis was used to examine treatment
discontinuations and rates of switching medications.
Results: The prior-authorization policy resulted in
an 8-percentage point reduction in the prevalence of
use of nonpreferred second-generation antipsychotic
and anticonvulsant medications (those requiring prior
authorization) but did not increase use of preferred
agents (no prior authorization) or rates of switching.
The prior-authorization policy reduced total phar-
macy reimbursements for bipolar disorder by $27 per
patient during the eight-month policy period.
However, the hazard rate of treatment discontinua-
tion (all bipolar drugs) while the policy was in effect
was 2.28 (95% confidence interval = 1.36-4.33)
higher than during the pre-policy period, with adjust-
ment for trends in the comparison state.
Conclusion: The small reduction in pharmacy
spending for bipolar treatment per patient after the
policy was implemented may have resulted from
higher rates of medication discontinuation rather than
switching. The findings indicate that the prior-
authorization policy may have increased patient risk
without appreciable cost savings.” (Zhang et al.,
2009, p. 520)
 244 UNIT TWO  The Research Process

Interrupted Time-Series Design with Multiple
Treatment Replications
The interrupted time-series design with multiple
treatment replications is a powerful design for infer-
ring causality (see Figure 11-21). It requires greater
researcher control than is usually possible in social
science research outside closed institutional settings,
such as laboratories or research units. The studies that
led researchers to adopt behavior modification tech-
niques used this design. For significant differences to
be interpretable, the pretest and posttest scores must
be in different directions with the introduction and
removal of the treatment. Within this design, treat-
ments can be modified by substituting one treatment
for another or combining two treatments and examin-
ing interaction effects (Shadish et al., 2002).
Experimental Study Designs
Experimental study designs provide the greatest
amount of control possible to examine causality more
closely. To examine cause, one must eliminate all
factors influencing the dependent variable other than
the cause (independent variable) being studied. Other
factors are eliminated by being controlled. The study
is designed to prevent any other element from intrud-
ing into observation of the specific cause and effect
that the researcher wishes to examine.
The three essential elements of experimental
research are (1) randomization; (2) researcher-
controlled manipulation of the independent variable;
and (3) researcher control of the experimental situa-
tion, including a control or comparison group. Experi-
mental designs exert much effort to control variance.
Sample criteria are explicit, the independent variable
or intervention is provided in a precisely defined way,
the dependent variables are carefully operationalized,
and the situation in which the study is conducted is
highly controlled to prevent the interference of unstud-
ied factors from modifying the dynamics of the process
being studied (Shadish et al., 2002).
Table 11-2 was developed so you might compare
the four major types of designs: descriptive, correla-
tional, quasi-experimental, and experimental. The
key focus of each type of design is identified with
an example study. Interventions are implemented
only in quasi-experimental and experimental study
designs.

TABLE 11-2  Comparison of Four Major Types of Design

Type of
Design Key Focus Sample Purpose Statement Intervention?
Descriptive Describes “what is” The purpose of this study was to (a) “determine practice and No
differences in practices between registered nurses and
respiratory therapists in managing patients receiving mechanical
ventilation” (Kjonegaard, Fields, & King, 2010, p. 168).
Correlational Examines relationships “The purpose of this study was to examine Jordanian mental health No
among study nurses’ experiences of providing mental health care, their
variables work-related stress, and organizational support received
“(Hamdan-Mansour, Al-Gamal, Puskar, Yacoub, & Marini, 2011,
p. 86).
oasis-ebl|Rsalles|1475959166

Quasi- Tests causality with “The aim of the study was to investigate the outcome of nursing Yes
experimental suboptimal control assessment, pain assessment and nurse-initiated intravenous
opioid analgesic compared to standard procedure for patients
seeking emergency care for abdominal pain. Outcome measures
were: (a) pain intensity, (b) frequency of received analgesic, (c)
time to analgesic, (d) transit time, and (e) patients’ perceptions
of the quality of care in pain management” (Muntlin, Carlsson,
Safwenberg, & Gunningberg, 2011, p. 13).
Experimental Tests causality with The purpose of this “experimental study was to examine the effects Yes
optimal control of a moderate-intensity aerobic exercise program on pain-like
behavior and neurotrophin-3 (NT-3) in female mice. The
rationale for conducting this study was that the literature and
clinical practice have supported the use of aerobic exercise in
reducing pain and improving function in people with chronic
pain but the molecular basis for these positive actions are poorly
understood” (Sharma, Ryals, Gajewski, & Wright, 2010, p. 714).
 CHAPTER 11  Selecting a Quantitative Research Design 245

Measurement of Manipulation of Measurement of

dependent variable(s) independent variable dependent variable(s)

Randomized Pretest Treatment Posttest

experimental
group
Randomized Pretest Posttest
comparison or
control group
Treatment: Under control of researcher

Approach to analysis: • Comparison of pretest and posttest scores

• Comparison of comparison and experimental groups
• Comparison of pretest/posttest differences between samples

Uncontrolled threats to validity: • Testing

• Instrumentation
• Mortality
• Restricted generalizability as control increases

Figure 11-22  The classic experimental design; pretest-posttest control group design.

Classic Experimental Design levels of the treatments, such as differing frequency,

The original, or classic, experimental design, or
pretest-posttest control group design, is still the
most commonly used experimental design (see Figure
11-22). There are two randomized groups, one receiv-
ing the experimental treatment and one receiving no
treatment, a placebo treatment, or the routine or stan-
dard care. By comparing pretest scores and the groups’
demographic characteristics, one can evaluate the
effectiveness of randomization in providing equivalent
groups. The researcher implements the treatment or
independent variable under very controlled conditions.
The dependent variable is measured twice, before and
after the manipulation of the independent variable
(Shadish et  al., 2002). As in all well-designed studies,
the dependent and independent variables are conceptu-
ally linked, conceptually defined, and operationalized.
Instruments used to measure the dependent variable
clearly reflect the conceptual meaning of the variable
and have good evidence of reliability and validity.
Often, more than one means of measuring the depen-
dent variable is advisable to avoid mono-operation
and mono-method biases (Waltz et  al., 2010).
Most other experimental designs are variations of
the classic experimental design. Multiple groups (both
experimental and comparison) can be used to great
advantage in the pretest-posttest design and the
posttest-only design. For example, the researcher
could withhold treatment from one comparison group
and treat another comparison group with a placebo.
Multiple experimental groups could receive varying
intensity, or duration of nursing care interventions.
These additions greatly increase the generalizability
of the study findings.
Malm, Karlsson, and Fridlund (2007) conducted an
experimental study of the effects of a self-care program
on health-related quality of life (HRQoL) for pace-
maker patients. The abstract describes their study as
follows:
“An experimental, multi-centre, randomized study
with a nurse-led intervention was conducted with the
aim of evaluating the effects on HRQoL of a 10-month
self-care program for pacemaker patients. In the
present study, there were no significant differences in
HRQoL when comparisons were made between the
experimental group and the control group. Results
show two main findings for patients in the self-care
program (n = 97; mean age 71 years): a significantly
better HRQoL in terms of experiencing the symp-
toms that were the reason for pacemaker implanta-
tion, as having decreased or disappeared, and a higher
level of perceived exertion in a 1 1/2-minute stair test
compared with patients who had standard checkups
(n = 115; mean age 73 years). It is important to
actively include pacemaker patients in a self-care
program while still in the acute phase in the hospital.
Healthcare professionals should support the patient
in a kind and professional manner by providing clear,
246 UNIT TWO  The Research Process
relevant information, and planning a self-care program
based on the nurse’s assessment of the patient’s
needs. To enable patients to manage their life situa-
tions, training and continued education for healthcare
professionals is necessary so that their efforts are
based on a holistic approach to nursing care and rec-
ognition of the patient perspective, with emphasis on
developing education and counseling for women,
patients with atrial fibrillation/sick sinus disease, and
patients whose pacemakers have ventricular pacing.”
(Malm et al., 2007, p. 15)
Experimental Posttest-Only Comparison
Group Design
In some studies, the dependent variable cannot be
measured before the treatment. For example, before
the beginning of treatment, it is not possible to
measure, in a meaningful way, a subject’s responses
to interventions designed to control nausea from che-
motherapy or postoperative pain. Additionally, in
some cases, subjects’ responses to the posttest can be
due, in part, to learning from or having a subjective
reaction to the pretest (pretest sensitization). If this
issue is a concern in your study, you may eliminate
the pretest and use an experimental posttest-only
design with a comparison group (see Figure 11-23).
However, you then will not be able to use many pow-
erful statistical analysis techniques within the study.
Additionally, the effectiveness of randomization in
obtaining equivalent experimental and comparison
groups cannot be evaluated in terms of the study vari-
ables. Nevertheless, the groups can be evaluated in
Manipulation of
independent variable
Randomized Treatment
experimental
group
Randomized
comparison or control group
Treatment: Under control of researcher
Approach to analysis: • Comparison of comparison and experimental groups
Uncontrolled threats to validity: • Instrumentation
• Mortality
• Restricted generalizability as control increases
Figure 11-23  Experimental posttest-only comparison group design.
terms of sample characteristics and other relevant
variables.
Randomized Blocking Designs
The randomized blocking design uses the two-group
pretest-posttest pattern or the two-group posttest
pattern with one addition: a blocking variable. The
blocking variable, if uncontrolled, is expected to con-
found the findings of the study. To prevent this confu-
sion, the subjects are rank ordered in relation to the
blocking variable.
For example, if effectiveness of a nursing interven-
tion to relieve post-chemotherapy nausea was the
independent variable in your study, severity of nausea
could confound the findings. Subjects could be ranked
according to severity of nausea. You would then iden-
tify and randomly assign the two subjects with the
most severe nausea, one to the experimental group and
one to the comparison group. You then would identify
and randomly assign the two subjects next in rank.
You would follow this pattern until the entire sample
was randomly assigned as matched pairs. This proce-
dure ensures that the experimental group and the com-
parison group are equal in relation to the potentially
confounding variable.
The effect of blocking can also be accomplished
statistically (through the use of analysis of covariance)
without categorizing the confounding variable into
discrete components. However, for this analysis to be
accurate, one must be careful not to violate the assump-
tions of the statistical procedure (Shadish et al., 2002;
Spector, 1981).
Santana-Sosa, Barriopedro, Lopez-Mojares, Perez,
and Lucia (2009) implemented a randomized block
Measurement of
dependent variable
Posttest
Posttest
 CHAPTER 11  Selecting a Quantitative Research Design 247
design in their study of the effect of exercise training
on patients with Alzheimer’s disease (AD). The
purpose of this study was to determine the effects of
a 12-week training program for patients with AD
using the Senior Fitness test on their (1) overall func-
tional capacity and (2) ability to perform activities of
daily living. With use of a randomized block design,
16 patients (8 control; 8 experimental) were assigned
to a training group (mean age 76 ± 4 years) or a control
group (73 ± 4 years). The results showed significant
improvements after training (p < 0.05) in upper and
lower body muscle strength and flexibility; agility and
dynamic balance; and endurance, fitness, gait, and
balance abilities in performing activities of daily
living (ADLs) independently. No changes were found
in the control group over the 12-week period. The
researchers indicated that exercise training should be
included in the nursing care for patients with AD.
Factorial Design
In a factorial design, two or more different character-
istics, treatments, or events are independently varied
within a single study. This design is a logical approach
to examining multicausality. The simplest arrange-
ment is one in which two treatments or factors are
involved and, within each factor, two levels are
manipulated (for example, the presence or absence of
the treatment); this is referred to as a 2 × 2 factorial
design. This design is illustrated in Figure 11-24, in
which the two independent variables are relaxation
and distraction as means of relieving pain.
A 2 × 2 factorial design produces a study with four
cells (A through D). Each cell must contain an approx-
imately equivalent number of subjects. Cells B and C
allow the researcher to examine each intervention
separately. Cell D subjects receive no treatment and
serve as a control group. Cell A allows the researcher
to examine the interaction between the two indepen-
dent variables. This design can be used, as in the
randomized block design, to control for confounding
variables. The confounding variable is included as an
independent variable, and interactions between it and
Level of Distraction
Level of
Relaxation Distraction No Distraction
Relaxation A B
No Relaxation C D
Figure 11-24  Example of factorial design.
the other independent variable are examined (Shadish
et al., 2002; Spector, 1981).
Extensions of the factorial design to more than two
levels of variables are referred to as M × N factorial
designs. Within this design, independent variables can
have any number of levels within practical limits.
Note that a 3 × 3 design involves 9 cells and requires
a much larger sample size. A 4 × 4 design would
require 16 cells. A 4 × 4 design would allow relaxation
to be provided at four levels of intensity, such as no
relaxation, relaxation for 10 minutes twice a day,
relaxation for 15 minutes three times a day, and relax-
ation for 20 minutes four times a day. Distraction
would be provided at similar levels.
Factorial designs are not limited to two indepen-
dent variables; however, interpretation of larger
numbers becomes more complex and requires greater
knowledge of statistical analysis. Factorial designs do
allow the examination of theoretically proposed inter-
relationships between multiple independent variables.
However, very large samples are required (Shadish
et al., 2002).
Winzer et al. (2010) conducted a clinical trial using
a factorial design to examine the effects of radiother-
apy and tamoxifen on the breast cancer recurrence
rates of patients after breast-conserving surgery (BCS).
The following excerpt from this study’s abstract
describes the design and key outcomes of the study.
“Between 1991 and 1998, 361 patients… were ran-
domised to radiotherapy (yes/no) and tamoxifen for
2 years (yes/no) in a 2 x 2-factorial design; the exclu-
sion of 7 centers (14 patients) left 347 patients for the
analysis. First results after a median follow-up of 5.9
years were published. Herein we present updated
results after a median follow-up of about 10 years.
One hundred and eleven events concerning event-
free survivals (EFS) have been observed. Since a
strong interactive effect between radiotherapy and
tamoxifen has been established, the results are pre-
sented in terms of the treatment effects for all four
treatment groups separately. Mainly due to the pres-
ence of local recurrences, the event rate was much
higher in the group with BCS only than in the other
three groups. No significant difference could be
established between the four treatment groups for
distant disease-free survival rates (DDFS). Updated
results give further evidence that even in patients with
a favorable prognosis, the avoidance of radiotherapy
and tamoxifen after BCS increases the rate of local
248 UNIT TWO  The Research Process

Pain Control Primary Nursing Care

Management
Primary Care No Primary Care
Unit Unit Unit Unit Unit Unit Unit Unit
A B C D E F G H
Unit
A
Unit
Traditional Care
B
PRN Medication Unit
C
Figure 11-25  Nested design.
Unit
D

Unit
E
Unit
New Approach F
“around the clock”
medication Unit
G
Unit
H

considered a subject and would be randomly assigned

recurrences substantially. Rates are about three times
higher in the BCS only group. For the two outcomes,
EFS and DDFS, no important difference could be seen
between the three groups with an additional treat-
ment. However, because of the limited sample size
with corresponding low power, the strength of evi-
dence for such a comparison is weak.” (Winzer et al.,
2010, p. 95)
to a treatment.
Sawyer, Deatrick, Kuna, and Weaver (2010)
explored patients with obstructive sleep apnea and
their perceptions of their disease and treatment with
continuous positive airway pressure by implementing
a nested design. A picture of this study design is pre-
sented in Figure 11-26. The following study excerpts
describe its design and key findings:

Nested Designs Background: “Obstructive sleep apnea (OSA) patients’

In some experimental situations, you may wish to
consider the effect of variables that are found only at
some levels of the independent variables being studied.
Variables found only at certain levels of the indepen-
dent variable are called nested variables and are best
investigated with nested designs (Shadish et al.,
2002). Possible nested variables are gender, race,
socioeconomic status, and education. A nested vari-
able may also be the patients who are cared for on
specific nursing units or at different hospitals; the sta-
tistical analysis in this case would be conducted as
though the unit or hospital were the subject rather than
the individual patient. Figure 11-25 illustrates the
nested design. In actual practice, nursing units used in
this manner would have to be much larger in number
than those illustrated, because each unit would be
consistent use of continuous positive airway pressure
(CPAP) therapy is critical to realizing improved func-
tional outcomes and reducing untoward health risks
associated with OSA.
Methods: We conducted a mixed methods, con-
current, nested study to explore OSA patients’
beliefs and perceptions of the diagnosis and CPAP
treatment that differentiate adherent from nonad-
herent patients prior to and after the first week of
treatment, when the pattern of CPAP use is estab-
lished [see Figure 11-26]. Guided by social cognitive
theory, themes were derived from 30 interviews
conducted post-diagnosis and after 1 week of CPAP
use. Directed content analysis, followed by categori-
zation of participants as adherent/nonadherent from
 CHAPTER 11  Selecting a Quantitative Research Design 249
Initial Clinical Evaluation at
Sleep Center & Recruitment (n = 19)
Informed Consent &
Enrollment (n = 18)
Demographic Survey
Standardized OSA and
CPAP Education:
20-minute video & brochure
Diagnostic Sleep Study (n = 16)
Failure to attend sleep study
excluded
Interview 1
Postdiagnosis (n = 16)
CPAP Sleep Study
(n = 16)
Interview 2
Post-CPAP Treatment (n = 16)
CPAP Adherence Data
Collection (n = 15)
Refused CPAP treatment excluded
Figure 11-26  Nested study design. (From Sawyer, A. M., Deatrick, J. A.,
Kuna, S. T., & Weaver, T. E. [2010]. Differences in perceptions of the
diagnosis and treatment of obstructive sleep apnea and continuous
positive airway pressure therapy among adherers and nonadherers.
Qualitative Health Research, 20[7], 876.)
objectively measured CPAP use, preceded across-
case analysis among 15 participants with severe OSA.
Beliefs and perceptions that differed between adher-
ers and nonadherers included OSA risk perception,
symptom recognition, self-efficacy, outcome expec-
tations, treatment goals, and treatment facilitators/
barriers. Our findings suggest opportunities for
developing and testing tailored interventions to
promote CPAP use.” (Sawyer et al., 2010, p. 373)
Crossover or Counterbalanced Designs
In some studies, more than one treatment is adminis-
tered to each subject. The treatments are provided
sequentially rather than concurrently. Comparisons
are then made of the effects of the different treatments
on the same subject. For example, two different
methods known to achieve relaxation might be used
as the two treatments. One difficulty with this type of
study is that exposure to one treatment may result in
effects (called carryover effects) that persist and
influence responses of the subject to later treatments.
Also, subjects can improve as they become more
familiar with the experimental protocol, a situation
called a practice effect. They may become tired or
bored with the study, a development called a fatigue
effect. The direct interaction of one treatment with
another, such as the use of two drugs, can confound
differences in the two treatments.
Crossover, or counterbalancing, is a strategy
designed to guard against possible erroneous conclu-
sions resulting from carryover effects. With crossover
or counterbalancing design, subjects are randomly
assigned to a specific sequencing of treatment condi-
tions. This approach distributes the carryover effects
equally throughout all the conditions of the study, thus
canceling them out. To prevent an effect related to
time, the same amount of time must be allotted to each
treatment, and the crossover point must be related to
time, not to the condition of the subject (Shadish et al.,
2002).
In addition, the design must allow for an adequate
interval between treatments to dissipate the effects
of the first treatment; this interval is referred to as a
washout period. For example, the design would
specify that each treatment would last 6 days and
that on the eighth day, each subject would cross over
to the alternative treatment after a 2-day washout
period.
The researcher also must be alert to the possibil-
ity that changes may be due to factors such as
disease progression, the healing process, or the
effects of treatment of the disease rather than the
study treatment. The process of counterbalancing
can become complicated when more than two treat-
ments are involved. Counterbalancing is effective
only if the carryover effect is essentially the same
from treatment A to treatment B as it is from treat-
ment B to treatment A. If one treatment is more
fatiguing than the other or more likely to modify
response to the other treatment, counterbalancing
will not be effective. You can use the crossover
design to control variance in your study and thus
allow the sample size to be smaller. The sample size
250 UNIT TWO  The Research Process

required to detect a significant effect is considerably

smaller because the subjects serve as their own con- “Until very recently, the genesis and use of new treat-
trols. Because the data collection period is longer, ments came about by means having little to do with
however, the rate of subject dropout may increase the scientific method. For millennia, the majority of
(Beck, 1989). therapies appear to have evolved by one of three
An example of this design is the Chang, Lin, Lin, methods: accidental discovery of treatments with
and Lin (2007) study of feeding premature infants unmistakable efficacy; the use of hypotheses alone,
using either single-hole or cross-cut nipple units. They without any experimentation; or the utilization of
described their study as follows: experimentation without controls, randomization,
blinding,… or adequate sample sizes. Treatments
originating by one of the latter two routes frequently
persisted for a very long time despite a lack of unbi-
“The purpose of this study was to compare the
ased evidence of their efficacy. Bloodletting, purging,
amount of total milk intake, feeding time, sucking
and the use of homeopathic dosages of drugs are
efficiency, heart rate (HR), respiratory rate (RR), and
examples. Failure of a treatment in any particular case
oxygen saturation (SpO2) of premature infants when
was usually attributed by its practitioners to its
fed with either single-hole or cross-cut nipple units.
misuse, to poor diagnosis, or to complicating factors.”
Twenty stable infants admitted to a level II nursery in
(Wooding, 1994, p. 26)
a tertiary care center with gestational ages averaging
32.2 +/- 3.2 wks were enrolled. Subjects had an
average postmenstrual age of 34.1 +/- 1.6 wks, and
average body weight of 1996 +/- 112 gm. A crossover
Over the years, the methodology for clinical trials
design was used and infants were observed for two
has evolved in medicine, resulting in less bias in sam-
consecutive meals separated by a four-hour interval.
pling, treatment implementation, and data collection
They were bottle fed with equal feeding amounts
(Meinert & Tonascia, 1986; Piantadosi, 1997; Pocock,
using a single-hole and cross-cut nipple administered
1996; Whitehead, 1992; Wooding, 1994). Meinert and
in random order. Results showed that infants fed with
Tonascia (1986) defined a clinical trial as a planned
single-hole nipple units took more milk (57.5 +/-
experiment designed to determine the efficacy of a
8.3 ml vs. 51.6 +/- 9.5 ml, p = 0.011), had a shorter
treatment or independent variable in a study in which
feeding time per meal (11.5 +/- 4.9 min vs. 20.9 +/-
the experimental group receives the treatment and the
5.0 min, p < 0.001), and sucked more efficiently (5.8
control group does not. The experimental group and
+/- 2.5 ml/min vs. 2.7 +/- 1.0 ml/min, p < 0.001) com-
the control group are usually established through
pared to those fed through cross-cut nipples. Infants
random assignment, and the study outcomes or depen-
using cross-cut nipple units had a higher RR (44.4
dent variables are compared for the two groups. In
+/- 4.6 breaths/minutes vs. 40.8 +/- 4.9 breaths/
clinical trials, the patients in both groups are enrolled,
minutes, p = 0.002) and SpO2 (96.1 +/- 1.4% vs. 94.6
treated, and followed over the same time. The outcome
+/- 3.2%, p = 0.044) than those using single-hole
measures are usually clinical events, laboratory tests,
nipples. Oxygen desaturation (SpO2 < 90% and
or mortality.
lasting for longer than 20 sec) and bradycardia were
The phase I, II, III, and IV clinical trial categories
not recorded in either group of infants during feeding.
were developed specifically for testing experimental
Compared to using cross-cut nipple units, premature
drug therapy (Meinert & Tonascia, 1986; Whitehead,
infants using single-hole nipple units take more milk
1992). Phase I, the initial testing of a new drug,
and tend to tolerate feedings better. A single-hole
focuses on determining the best drug dose and identi-
nipple may be a choice for physiologically stable
fying safety effects. Phase II trials seek preliminary
bottle-fed premature infants.” (Chang et al., 2007,
evidence of efficacy and side effects of the drug dose
p. 215)
determined by the phase I trial. Sometimes in Phase
II, the experimental group receiving the drug is com-
pared with a placebo group.
Clinical Trials Phase III trials are comparative definitive studies in
Clinical trials have been used in medicine since 1945. which the new drug’s effects are compared with those
Wooding (1994) described the strategies that were of the drug considered standard therapy. Phase III
used to introduce new medical therapies before that trials are sometimes referred to as “full-scale defini-
time: tive clinical trials,” suggesting that a decision is made
 CHAPTER 11  Selecting a Quantitative Research Design 251
on the basis of the findings as to whether the experi-
mental drug is more effective than standard treatment.
In some phase III clinical trials, the sample size is not
determined before initiation of data collection. Rather,
data are analyzed at intervals to test for significant
differences between groups. If a significant difference
is found, data collection may be discontinued. Other-
wise, the data collection will continue and retesting is
initiated after accrual of additional subjects (Meinert
& Tonascia, 1986; Whitehead, 1992). Phase IV trials
occur after regulatory approval of the drug and are
designed to monitor patients over time to determine
drug safety, uncommon side effects, and long-term
consequences in a larger population. Phase IV trials
might also focus on testing marketing strategies and
examining cost-effectiveness for the drug (Piantadosi,
1997; Wooding, 1994).
Piantadosi (1997) recommended redefining these
stages to be broader and applicable to more types of
trials. He suggested using the following terminology:
early development, middle development, comparative
studies, and late development. In early development
trials, researchers would develop and test the treat-
ment mechanism (thus, they could also be called TM
trials). Middle development studies would focus on
clinical outcomes and treatment “tolerability.” Toler-
ability would have three components: feasibility,
safety, and efficacy; thus, Piantadosi (1997) referred
to middle development studies as safety and efficacy
trials, or SE trials. In this phase, the researcher would
estimate the probability that patients would benefit
from the treatment (or experience side effects from it).
Performance criteria such as success rate might
be used.
Comparative studies, according to Piantadosi
(1997), would have defined clinical end points and
would address comparative treatment efficacy (so
could be called CTE trials). These studies would
include a concurrent control group that receive the
standard treatment and an experimental group that
receives the experimental treatment. These studies
might also include a placebo group, which receives a
faux treatment to determine the psychological response
to receiving a treatment. Late development studies
would be designed to identify uncommon side effects,
interactions with other treatments, and unusual com-
plications. They would be developed as expanded
safety trials, or ES trials, conducted overtime with
larger samples. Over the years, researchers have
increased the control in the development and imple-
mentation of healthcare clinical trials and these trials
are usually referred to as randomized controlled trials
(RCTs) in the research literature.
Randomized Controlled Trials
Currently in medicine and nursing, the randomized
controlled trial (RCT) design is noted to be the
strongest methodology for testing the effectiveness of
a treatment because of the elements of the design that
limit the potential for bias. Subjects are randomly
assigned to the treatment and control groups to
reduce selection bias (Shadish et al., 2002). Biases
developing from participant attrition from studies are
managed with intention-to-treat analyses (Polit, Gil-
lespie, & Griffin, 2011). In addition, blinding or
withholding of study information from data collec-
tors, participants, and their healthcare providers can
reduce the potential for bias. Thus, RCTs when
appropriately conducted are considered the gold stan-
dard for determining the effectiveness of healthcare
interventions.
However, there were serious criticisms of the
inconsistencies and biases identified in the clinical
trials conducted. Thus, in 1993, a panel of 30 experts—
clinical trial researchers, medical journal editors, epi-
demiologists, and methodologists—met in Ottawa,
Canada, to develop a scale to assess the quality of
RCTs reports. This group initiated the Standardized
Reporting of Trials (SORT) statement (CONSORT,
2011). This statement included a checklist and flow
diagram that investigators were encouraged to follow
when conducting and reporting RCTs. The initial work
of this group was revised in 2001 and became
the Consolidated Standards for Reporting Trials
(CONSORT). This guideline was updated with the
CONSORT 2010 Statement published by Schultz,
Altman, and Moher (2010) as representatives of the
CONSORT Group. Figure 11-27 provides a flow
diagram of the progression through the phases of an
RCT— enrollment, intervention allocation, follow-up,
and data analysis—for two randomized parallel
groups. This diagram was included in the CONSORT
2010 Statement to facilitate the conduct of quality
RCTs nationally and internationally (Schulz et al.,
2010). The CONSORT 2010 Statement also offers
a checklist of information that researchers need
to supply when reporting a RCT; it can be found
in Chapter 19, the Schulz et al. (2010) publication,
or online (http://www.consort-statement.org/consort-
statement/) (CONSORT, 2012).
Only in the past 10 years has the term RCT been
used to describe studies conducted in nursing. When
implementing these clinical trials in nursing, the meth-
odology needs to be redefined to fit the knowledge-
building needs of nursing and also conform to the
CONSORT standards. Thus, an RCT conducted in
nursing needs to meet the following expectations:
252 UNIT TWO  The Research Process

Assessed for eligibility (n = )

Enrollment

Excluded (n = )
Not meeting inclusion criteria (n = )
Declined to participate (n = )
Other reasons (n = )

Randomized (n = )
Allocation

Allocated to intervention (n = ) Allocated to intervention (n = )

Received allocated intervention (n = ) Received allocated intervention (n = )
Did not receive allocated intervention (give Did not receive allocated intervention (give
reasons) (n = ) reasons) (n = )
Follow-Up

Lost to follow-up (give reasons) (n = ) Lost to follow-up (give reasons) (n = )

Discontinued intervention (give reasons) (n = ) Discontinued intervention (give reasons) (n = )
Analysis

Analyzed (n = ) Analyzed (n = )
Excluded from analysis (give reasons) (n = ) Excluded from analysis (give reasons) (n = )

Figure 11-27  2010 Statement flow diagram of the progress through the phases of a parallel randomized trial of two groups (that is, enrollment,
intervention allocation, follow-up, and data analysis). (From CONSORT. [2012]. The CONSORT Statement. Retrieved from http://www.consort-
statement.org/consort-statement/; Schulz, K. F., Altman, D. G., Moher, D., for the CONSORT Group [2010]. CONSORT 2010 Statement: Updated
guidelines for reporting parallel group randomised trials. British Medical Journal, 340, c332.)

1. The study is designed to be a definitive test of the
hypothesis that the intervention causes the defined
effects or outcomes.
2. Previous studies have provided evidence that the
intervention causes the desired outcome.
3. The intervention is clearly defined, and a protocol
has been established for its clinical application
ensuring intervention fidelity (Santacroce et al.,
2004; Yamada et al., 2010).
4. The study is conducted in a clinical setting, not in
a laboratory.
5. The design meets the criteria of an experimental
study (Schulz et al., 2010).
6. Subjects are drawn from a reference population
through the use of clearly defined criteria. Base-
line states are comparable in all groups included
in the study. Selected subjects are then randomly
assigned to treatment and comparison groups—
thus, the term randomized clinical trial
(CONSORT, 2012; Schulz et al., 2010).
7. Subjects are accrued individually over time as
they enter the clinical area, are identified as
 CHAPTER 11  Selecting a Quantitative Research Design 253
meeting the study criteria, and agree to participate
in the study.
8. The study has high internal validity. The design
is rigorous and involves a high level of control of
potential sources of bias that will rule out possible
alternative causes of the effect (Shadish et al.,
2002). The design may include blinding to accom-
plish this purpose. Blinding means that the
patient, those providing care to the patient, and/or
the data collectors do not know whether the
patient is in the experimental group or the control
group. The CONSORT 2010 Statement recom-
mends that the researchers clearly designate who
is blinded in the research plan, conduct the study
as planned, and designate who was blinded in the
final report (Polit et al., 2011; Schulz et al., 2010).
9. The treatment or independent variable is equal
and consistently applied to all subjects in the
experimental group (CONSORT, 2012).
10. Dependent variables or outcomes are measured
consistently with quality measurement methods
(Waltz et al., 2010).
11. The proposed study has been externally reviewed
by expert researchers who have approved the
design.
12. The study has received external funding sufficient
to allow a rigorous design with a sample size
adequate to provide a definitive test of the
intervention.
13. The research report covers the items on
the CONSORT 2010 Statement checklist
(CONSORT, 2012).
14. If the clinical trial results indicate a significant
effect of the intervention, the evidence is suffi-
cient to warrant application of the findings in
clinical practice (Melnyk & Fineout-Overholt,
2011).
15. The intervention is defined in sufficient detail so
that clinical application can be achieved (Bulechek
et al., 2008; Polit et al., 2011; Schulz et al., 2010).
Elwood (1998) suggested that clinical trial meth-
odology could be used for prevention intervention
studies as well as testing treatments. Murray (1998)
proposed methods of randomizing groups rather than
subjects in prevention studies and explored issues
related to community-based trials such as sample
mortality.
Clinical trials may be carried out simultaneously in
multiple geographical locations to increase sample
size and resources and to obtain a more representative
sample (Schulz et al., 2010). In this case, the primary
researcher must coordinate activities at all the sites.
Meinert and Tonascia (1986) indicated that the costs
per patient per year of the study are less for multi-
center studies than for single-center trials. If you plan
to use this technique in your research, you must con-
front several problems. Coordination of a project of
this type requires much time and effort. Keeping up
with subjects is critical but may be difficult. Commu-
nication with and cooperation of staff assisting with
the study in the various geographical locations are
essential but sometimes challenging. You may encoun-
ter attempts to ignore the protocol and provide tradi-
tional care (Fetter et al., 1989; Gilliss & Kulkin, 1991;
Schulz et al., 2010; Tyzenhouse, 1981). Meinert and
Tonascia (1986) recommended the development of a
coordinating center for multisite clinical trials that will
be responsible for receiving, editing, processing, ana-
lyzing, and storing data generated in the study. Nurse
researchers need to follow the CONSORT 2010 State-
ment recommendations in the conduct of RCTs and in
their reporting (CONSORT, 2012; Schulz et al., 2010).
Jones, Duffy, and Flanagan (2011) conducted an
RCT to test the efficacy of a nurse-coached interven-
tion (NCI) on the outcomes of patients undergoing
ambulatory arthroscopy surgery. The NCI was devel-
oped to improve the postoperative experiences of
patients and families following ambulatory surgery.
This study was funded by the National Institute of
Nursing Research. The following study excerpt identi-
fies the study hypothesis, design and major findings:
“This study was conducted to test the hypothesis that
ambulatory arthroscopic surgery patients who receive
a nurse-coached telephone intervention will have sig-
nificantly less symptom distress and better functional
health status than a comparable group who receive
usual practice….
The study sample in this randomized clinical trial
with repeated measures was 102 participants (52 in
the intervention group and 50 in the usual practice
group) drawn from a large academic medical center
in the Northeast United States. Symptom distress
was measured using the Symptom Distress Scale, and
functional health was measured using the Medical
Outcomes Study 36-Item Short-Form Health Survey
General Health Perceptions and Mental Health sub-
scales.” (Jones et al., 2011, p. 92)
Jones et al. (2011) detailed their study intervention
and the steps taken by the researchers to promote fidel-
ity in the implementation of the NCI. The nurses were
trained in the delivery of the NCI by the study team
using a video. Each nurse coach was given a packet
254 UNIT TWO  The Research Process
of guidelines for management of the participants’
symptoms:
“The guidelines addressed common patient problems
associated with postoperative recovery after arthros-
copy with general anesthesia (e.g., nausea, vomiting,
pain, immobility). The guidelines contained five areas
to evaluate: (a) assessment (self-report), (b) current
management of symptoms, (c) evaluation by the
coach of the adequacy of the intervention, (d) addi-
tional intervention strategies to address the present-
ing symptoms, and (e) proposed outcome (self-
report).” Jones et al., 2011, p. 95)
The sample criteria were detailed to identify the
target population. Once the patients consented to be in
the study, they were randomly assigned to the NCI
group or the usual care group with use of the sealed-
envelope method (Maxwell & Delaney, 2004). The
researchers found that the NCI delivered by telephone
postoperatively to patients undergoing arthroscopic
surgery significantly reduced their symptom distress
and improved their physical and mental health.
The steps of this study followed the steps outlined
in the CONSORT flow diagram in Figure 11-27.
More details for conducting RCTs can be found in
Chapter 14.
Studies That Do Not Use
Traditional Research Designs
In some approaches to research, the research designs
described in this chapter cannot be used. These studies
tend to be in highly specialized areas that require
unique design strategies to accomplish their purposes.
Designs for primary prevention and health promotion,
secondary analysis, and methodological studies are
described here.
Primary Prevention and Health
Promotion Studies
To study primary prevention and health promotion as
a nurse researcher, you must apply a treatment of
primary prevention (the cause) and then attempt to
measure the effect (an event that does not occur if the
treatment was effective). Primary prevention studies,
then, attempt to measure things that do not happen.
One cannot select a sample to study, apply a treatment,
and then measure an effect. The sample must be the
community. The design involves examining changes
in the community, and the variables are called indica-
tors. A change in an identified indicator is inferred to
be a consequence of the effectiveness of the preven-
tion program (treatment).
Specific indicators would depend on the focus of
prevention. For example, nurses in Canada identified
oral mucositis as a recurring issue in oncology clinical
practice and developed an oral care guide. They used
the University Health Network Nursing Research Uti-
lization Model and the Neuman Systems Model as
conceptual frameworks (Salvador, 2006). The follow-
ing excerpt describes their study:
“A flowchart was developed to ensure a coordinated
and continuous provision of oral care. Educational
presentations were conducted to familiarize nurses
and members of the multidisciplinary team of the
practice changes. The introduction of the oral care
regimen as primary prevention, plus systematic oral
assessment and monitoring had the potential to
reduce the occurrence and severity of oral mucositis
in patients undergoing autologous stem cell transplan-
tation.” (Salvador, 2006 p. 18)
How might you study the effectiveness of this
primary prevention strategy? Because one indicator
alone would be insufficient to infer effect, multiple
indicators and statistical analyses appropriate for these
indicators must be used. For example, you might
measure the color of the oral mucosa, moistness in the
mouth, severity of oral mucositis, and amount of pain
expressed by the patient when eating.
Secondary Analyses
Secondary analysis design involves studying data
previously collected in another study. Data are reex-
amined with the use of different subsets of the data or
variables and different statistical analyses from those
previously used. The design involves analyzing data
to validate the reported findings, examine dimensions
previously unexamined, or redirect the focus of the
data to allow comparison with data from other studies
(Gleit & Graham, 1989; Windle, 2010).
Secondary analysis studies may use data collected
by state and national governments. For example,
Chertok, Luo, and Anderson (2011) analyzed over 17
years of West Virginia’s birth certificate data to answer
a research question about associations between
changes in prenatal smoking habits and subsequent
infant birth weights. Cipher, Hooker, and Guerra
(2006) compared the prescribing trends of nurse
 CHAPTER 11  Selecting a Quantitative Research Design 255
practitioners, physician assistants, and physicians
using data in the National Ambulatory Medical Care
Survey (NAMCS) database. They found that for 70%
of the visits, the three clinician types were likely to
write at least one prescription, and that PAs were more
likely than NPs or physicians to prescribe controlled
substances.
Because data collected for some national databases
may have used oversampling of underrepresented
groups, researchers using these databases must care-
fully examine the inclusion and exclusion sampling
criteria used by the original research team. Providing
substantive information about the challenges pre-
sented by large database research is beyond the scope
of this book. Researchers interested in using large
databases for secondary analysis should consult a ref-
erence book on the topic (Kiecolt & Nathan, 1985;
Trzesniewski, Donnellan, & Lucas, 2011; Vartanian,
2011).
Data accumulated during the research programs of
groups of faculty provide opportunities for secondary
data analysis as well. This approach allows the inves-
tigators to examine questions related to the data that
were not originally posed. These data sets may provide
opportunities for junior faculty members or graduate
students to become involved in a research program.
This type of secondary data analysis requires thought-
ful, honest consideration to avoid violation of profes-
sional standards, as Aaronson (1994) recommends:
“Fundamentally, each paper written from the same
study or the same dataset must make a distinct and
significant scientific contribution. Presumably this is
not only the major overriding criterion used by
reviewers, but also the author’s intent when writing
the paper. When a particular paper is one of several
oasis-ebl|Rsalles|1475959179
from the same study, project, or dataset, the author’s
responsibility to identify the source of the data is that
much greater. To lead readers to think a report is
from a new study or a different dataset than that used
in the authors’ previous work is dishonest, particu-
larly if the second paper purports to substantiate find-
ings of the first one…. Apart from the overriding
concern about ‘milking the data,’ the most common
objection to multiple articles from a single study is
concern about the age of the data…. Concerns in
nursing about the number of papers generated from
a single study may reflect the emerging status of sec-
ondary analysis as a legitimate approach to nursing
research….
All of the reasons offered for using secondary
analysis—answering new questions with existing data,
applying new methods to answer old questions, the
real exigencies of cost and feasibility—serve equally
to justify the continued use of data collected years
ago, by the original investigator of a large project, as
well as by others…. The issue remains one of sound
science. The question that must be asked is: Does this
particular paper make a meaningful and distinct con-
tribution to the scientific literature?” (Aaronson,
1994, pp. 61-62)
An example of secondary analysis is Koci and
Strickland’s (2007) study of the relationship of ado-
lescent physical and sexual abuse with perimenstrual
syndrome (PMS) in adulthood. Data analyzed in this
study were from a longitudinal study of a community
sample of 568 women in a database called Nursing
Assessment of PMS: Neurometric Indices. A study
such as this yields an enormous amount of data that
are not examined in the original study. Koci and
Strickland (2007) posed a different research question
that could be examined using the data from the longi-
tudinal study. They found that “a history of both ado-
lescent physical abuse and sexual abuse was
significantly associated with PMS in adulthood.
Women with a history of adolescent physical and
sexual abuse had significantly more severe PMS pat-
terns with more dysphoria than women without abuse”
(Koci & Strickland, 2007, p. 75).
Methodological Designs
Methodological designs are used to develop the
validity and reliability of instruments to measure con-
structs used as variables in research. The process is
lengthy and complex. The average researcher time
required to develop a research tool to the point of
appropriate use in a study is 5 years (Waltz et al.,
2010). Methodological studies include assessment of
content validity, evaluation of the conceptual structure
of a scale, construct validity, and assessment of
reliability.
Simons, Stark, and DeMarco (2011) conducted a
methodological study to develop a new, four-item
instrument to measure bullying of nurses in the work-
place. These researchers described their process for
developing a valid and reliable questionnaire for mea-
suring workplace bullying and provided direction for
further research to support continued development of
this instrument. The key ideas of this study are pre-
sented in the following abstract:
256 UNIT TWO  The Research Process
“Studies on workplace bullying either in the U.S. or
internationally rarely include nurses. We tested the con-
current validity of the Negative Acts Questionnaire-
Revised (22 items) with a sample of nurses. Five hundred
eleven registered nurses (RNs) responded to a mailed
survey. Factor, reliability, and regression analyses tested
dimensionality, reliability, and construct and criterion
validity. Workplace bullying is best seen as a one-
dimensional construct. A subset of four items was found
Algorithms for Selecting
Research Designs
To select a research design, the investigator must
follow paths of logical reasoning. You need a calculat-
ing mind to explore all the possible consequences of
using a particular design in a study. In some ways,
selecting a design is like thinking through the moves
in a chess game. You must carefully think through the
consequences of each option. The research design
organizes all the components of the study in a way that
is most likely to lead to valid answers to the questions
that have been posed. Studies with high-quality
research designs implemented over time have a strong
influence on the research evidence generated for prac-
tice (Brown, 2009; Fawcett & Garity, 2009; Melnyk
& Fineout-Overholt, 2011).
To help you select the most appropriate design for
your study, we encourage you to use the algorithms,
or decision trees, provided at the end of this chapter.
The first algorithm (see Figure 11-28) will help you
identify the type of study you plan to conduct or to
determine the type of study in a publication. The next
four algorithms (see Figures 11-29 through 11-32) will
assist you in identifying designs in published studies
and in selecting a specific design for a study you plan
to conduct. Figure 11-29 is an algorithm of commonly
conducted descriptive study designs, and Figure 11-30
includes common correlational study designs. Figure
11-31 is an algorithm for identifying common quasi-
experimental designs, and Figure 11-32 includes
experimental designs. Most of the designs identified
in these figures have been discussed in this chapter.
Selecting a design is not a rigid, rule-guided task. As
a researcher, you have considerable flexibility in
choosing a design. The pathways within the algo-
rithms are not absolute and are to be used as guides.
Sometimes researchers combine different types of
designs to address their study purpose and their
research objectives, questions, or hypotheses.
to be both valid and reliable in measuring bullying in this
sample. Findings support use of a one-dimensional, four-
item questionnaire to measure perceived bullying in
nursing populations. Using a four-item questionnaire
decreases participant and researcher burden and makes
available an outcome measure for future descriptive and
predictive interventional research.” (Simons et al.,
2011, p. 132)
KEY POINTS
• Researchers have developed designs to meet unique
research needs as they emerge.
• At present, most nurse researchers are using designs
developed by other disciplines, which are a useful
starting point, but nurse scientists need to go
beyond them to develop designs that will more
appropriately meet the needs of the knowledge base
in nursing.
• Descriptive studies are designed to gain more infor-
mation about variables within a particular field of
study.
• Correlational studies examine relationships
between variables but do not provide information
on causality.
• Quasi-experimental and experimental designs
examine causality. The power of the design to
accomplish this purpose depends on the degree to
which the actual effects of the experimental treat-
ment (the independent variable) can be detected by
measuring the dependent or outcome variable.
• Obtaining an understanding of the true effects of
an experimental treatment requires action to control
threats to the validity of the findings.
• Threats to validity are controlled through selection
of subjects, manipulation of the treatment, and
measurement of variables.
• Currently in medicine and nursing, the randomized
controlled trial (RCT) design is noted to be the
strongest methodology for testing the effectiveness
of a treatment because of the elements of the design
that limit the potential for bias. The CONSORT
2010 Statement clarifies the steps for conducting
and reporting an RCT.
• Studying primary prevention and health promotion
involves applying a treatment of primary preven-
tion (the cause) and then attempting to measure the
effect (an event that does not occur if the treatment
was effective).
Text continued on p. 261
 CHAPTER 11  Selecting a Quantitative Research Design 257

Is there a treatment?

No Yes

Is the primary purpose Is the treatment tightly

examination of relationships? controlled by the researcher?

No Yes No Yes

Descriptive Will the sample be Quasi-experimental Will a randomly

design studied as a study assigned comparison
single group? or control group
be used?

No Yes No Yes

Correlational Is the original

design sample randomly
selected?

No Yes

Experimental
study

Figure 11-28  Type of study.

258 UNIT TWO  The Research Process

Examining sequences across time?

No Yes

One group? Following same subjects

across time?

No Yes
No Yes

Comparative Descriptive
descriptive design Data collected Single unit
design across time? of study?

No Yes No Yes

Cross- Study events Longitudinal Case

sectional partitioned study study
design across time?

No Yes

Trend Repeated
analysis measures of
each subject?

Yes No

Cross-sectional design Longitudinal design

with treatment with treatment
partitioning partitioning

Figure 11-29  Descriptive studies.

Describe relationships Predict relationships Test theoretically

between/among between/among proposed
variables? variables? relationships?

Figure 11-30  Correlational studies.

Descriptive correlational Predictive correlational Model testing

design design design
 CHAPTER 11  Selecting a Quantitative Research Design 259

No Comparison group? Yes

Pretest? Pretest?

No Yes No Yes

One-group Repeated Posttest Repeated

posttest only measures? only with measures?
design nonequivalent
group

Comparison
with No Yes No Yes
population
values?
Strategy for Compare Compare Interrupted
comparison? experimental and treatments? time series
comparison conditions with
nonequivalent
No Yes comparison
group

Suggest No Yes No Yes

reevaluating
design
Compare Treatment No Yes
Suggest Compare variables? replications?
reevaluating treatment
design and
comparison Untreated Reversed-treatment
conditions comparison nonequivalent
group with comparison group
One-group No Yes pretest and with pretest
posttest only posttest and posttest
design
Simple Interrupted
interrupted time series
time with
series nonequivalent
dependent
No Yes variables No Yes

Nonequivalent Repeated Interrupted Interrupted

dependent treatment time series time series
variables design with with
design removed repeated
treatment replications

Figure 11-31  Quasi-experimental studies.

260 UNIT TWO  The Research Process

No Comparison group? Yes

Posttest Repeated
only comparison measures?
group design

No Yes

Examine effects Repeated

of confounding measures design
variables?

No Yes

Multiple sites? Blocking?

No Yes No Yes

Pretest/posttest Randomized Comparison of Randomized

comparison or controlled multiple levels block design
control group design trials (RCTs) of treatment

No Yes

Examination of Nested
complex relationships designs
among variables in
relation to treatment

No Yes

Examination Multivariate
of multiple designs
causality?

No Yes

Solomon Factorial
four-group designs
design

Figure 11-32  Experimental studies.

 CHAPTER 11  Selecting a Quantitative Research Design 261
• Secondary analysis is the study of data previously
collected in another study or for a non-research
purpose.
• Methodological studies are designed to develop the
validity and reliability of instruments to measure
constructs used as variables in research.
• Algorithms for design identification and selection
are provided in Figures 11-29 to 11-32.
REFERENCES
Aaronson, L. S. (1994). Milking data or meeting commitments:
How many papers from one study? Nursing Research, 43(1),
60–62.
Aberson, C. L. (2010). Applied power analysis for the behavioral
sciences. New York, NY: Routledge Taylor & Francis Group.
Bailey J. J., Sabbagh M., Loiselle C. G., Boileau J., & McVey L.
(2010). Supporting families in the ICU: A descriptive correla-
tional study of informational support, anxiety, and satisfaction
with care. Intensive & Critical Care Nursing, 26(2), 114–122.
Beck, S.L. (1989). The crossover design in clinical nursing research.
Nursing Research, 38(5), 291–293.
Bialocerkowski, A., Klupp, N., & Bragge, P. (2010). Research
methodology series: How to read and critically appraise a reli-
ability article. International Journal of Therapy and Rehabilita-
tion, 17(3), 114–120.
Bowles, K.H., & Naylor, M.D. (1996). Nursing intervention
classification systems. Journal of Nursing Scholarship, 28(4),
303–308.
Brown, S.J. (2009). Evidence-based nursing: The research-practice
connection. Boston, MA: Jones and Bartlett.
Bulechek, G. M., Butcher, H. K., & Dochterman, J. M. (2008).
Nursing Interventions Classification (NIC) (5th ed.). St. Louis,
MO: Mosby Elsevier.
Campbell, D. T., & Stanley, J. C. (1963). Experimental and quasi-
experimental designs for research. Chicago, IL: Rand McNally.
Chang, Y. J., Lin, C. P., Lin, Y. J., & Lin, C. H. (2007). Effects of
single-hole and cross-cut nipple units on feeding efficiency and
physiological parameters in premature infants. Journal of Nursing
Research, 15(3), 215–223.
Chertok, I. R. A., Luo, J., & Anderson, R. H. (2011). Association
between changes in smoking habits in subsequent pregnancy and
infant birth weight in West Virginia. Maternal and Child Health
Journal, 15(2), 249–254.
Cipher, D. J., Hooker, R. S., & Guerra, P. (2006). Prescribing trends
by nurse practitioners and physician assistants in the United
States. Journal of the American Academy of Nurse Practitioners,
18(6), 291–291.
CONSORT. (2011). How CONSORT began. Retrieved, from http://
www.consort-statement.org/about-consort/history/.
CONSORT. (2012). The CONSORT Statement. Retrieved from
http://www.consort-statement.org/consort-statement/.
Cook, T. D., & Campbell, D. T. (1979). Quasi-experimentation:
Design and analysis issues for field settings. Chicago, IL: Rand
McNally.
Corty, E. W. (2007). Using and interpreting statistics: A practical
text for the health, behavioral, and social sciences. St. Louis, MO:
Mosby Elsevier.
Costanzo, C., Walker, S. M., Yates, B. C., McCabe, B., & Berg, K.
(2006). Physical activity counseling for older women. Western
Journal of Nursing Research, 28(7), 786–810.
Cramer, M. E., Chen, L. W., Roberts, S., & Clute, D. (2007). Evalu-
ating the social and economic impact of community-based prena-
tal care. Public Health Nursing, 24(4), 329–336.
Creswell, J. W. (2009). Research design: Qualitative, quantitative,
and mixed methods approaches (3rd ed.). Los Angeles, CA:
Sage.
Crombie, I. K., & Davies, H. T. O. (1996). Research in health care:
Design, conduct and interpretation of health services research.
New York, NY: Wiley.
Cummings, G. G., Estabrooks, C. A., Midodzi, W. K., Wallin, L.,
& Hayduk, L. (2007). Influence of organizational characteristics
and context on research utilization. Nursing Research, 56(4
Suppl), S24–S39.
DeVon, H. A., Block, M. E., Moyle-Wright, P., Ernst, D. M.,
Hayden, S. J., Lazzara, D. J., et al. (2007). A psychometric toolbox
for testing validity and reliability. Journal of Nursing Scholarship,
39(2), 155–164.
Doran, D. M. (2011). Nursing outcomes: The state of the science
(2nd ed.). Sudbury, MA: Jones & Bartlett Learning.
Dudgeon, W. D., Phillips, K. D., Durstine, J. L., Burgess, S. E.,
Lyerly, G. L. W., Davis, J. M., et al. (2010). Individual exercise
sessions alter circulating hormones and cytokines in HIV-infected
men. Applied Physiology, Nutrition & Metabolism, 35(4),
560–568.
Egan, E. C., Snyder, M., & Burns, K. R. (1992). Intervention studies
in nursing: Is the effect due to the independent variable? Nursing
Outlook, 40(4), 187–190.
Elwood, J. M. (1998). Critical appraisal of epidemiological studies
and clinical trials. New York: Oxford University Press.
Fawcett, J., & Garity, J. (2009). Evaluating research for evidence-
based nursing practice. Philadelphia, PA: F. A. Davis.
Fetter, M. S., Fettham, S. L., D’Apolito, K., Chaze, B. A., Fink, A.,
Frink, B. B., et al. (1989). Randomized clinical trials: Issues for
researchers. Nursing Research, 38(2), 117–120.
Fisher, R. A. (1935). The design of experiments. New York, NY:
Hafner.
Forbes, A. (2009). Clinical intervention research in nursing. Inter-
national Journal of Nursing Studies, 46(4), 557–568.
Gilliss, C. L., & Kulkin, I. L. (1991). Monitoring nursing interven-
tions and data collection in a randomized clinical trial. Western
Journal of Nursing Research, 13(3), 416–422.
Gleit, C., & Graham, B. (1989). Secondary data analysis: A valuable
resource. Nursing Research, 38(6), 380–381.
Gray, M. (1998). Introducing single case study research design: An
overview. Nurse Researcher, 5(4), 15–24.
Grove, S. K. (2007). Statistics for health care research: A practical
workbook. St. Louis, MO: Saunders Elsevier.
Hamdan-Mansour, A. M., Al-Gamal, E., Puskar, K., Yacoub, M.,
& Marini, A. (2011). Mental health nursing in Jordan: An
investigation into experience, work stress and organizational
support. International Journal of Mental Health Nursing, 20(2),
86–94.
Harris, K. M., Gordon-Larsen, P., Chantala, K., & Udry, J. R., R.
(2006). Longitudinal trends in race/ethnic disparities in leading
health indicators from adolescence to young adulthood. Archives
of Pediatric & Adolescent Medicine, 160(4), 74–81.
262 UNIT TWO  The Research Process
Hjelm-Karlsson, K. (1991). Using the biased coin design for ran-
domization in health care research. Western Journal of Nursing
Research, 13(2), 284–288.
Jones, D., Duffy, M. E., & Flanagan, J. (2011). Randomized clinical
trial testing efficacy of a nurse-coached intervention in arthros-
copy patients. Nursing Research, 60(2), 92–99.
Kerlinger, F. N., & Lee, H. B. (2000). Foundations of behav-
ioral research (4th ed.). Fort Worth, TX: Harcourt College
Publishers.
Kiecolt, J. K., & Nathan, L. E. (1985). Secondary data analysis of
survey data. Beverly Hills, CA: Sage Publications.
Kjonegaard, R., Fields, R., & King, M. L. (2010). Current practice
in airway management: A descriptive evaluation. American
Journal of Critical Care, 19(2), 168–174.
Kline, G.A., & Edwards, A. (2007). Antepartum and intra-partum
insulin management of type 1 and type 2 diabetic women: Impact
on clinically significant neonatal hypoglycemia. Diabetes
Research & Clinical Practice, 7(22), 223–230.
Koci, A., & Strickland, O. (2007). Relationship of adolescent physi-
cal and sexual abuse to perimenstrual symptoms (PMS) in adult-
hood. Issues in Mental Health Nursing, 28(1), 75–87.
Lee, B. O., Chaboyer, W., & Wallis, M. (2010). Illness representa-
tions in patients with traumatic injury: A longitudinal study.
Journal of Clinical Nursing, 19(3-4), 556–563.
Lim, L. M., Chiu, L. H., Dohrmann, J., & Tan, K. L. (2010). Reg-
istered nurses’ medication management of the elderly in aged care
facilities. International Nursing Review, 57(1), 98–106.
Malm, D., Karlsson, J. E., & Fridlund, B. (2007). Effects of a self-
care program on the health-related quality of life of pacemaker
patients: A nursing intervention study. Canadian Journal of Car-
diovascular Nursing, 17(1), 15–26.
Mancuso, J. M. (2010). Impact of health literacy and patient trust
on glycemic control in an urban USA population. Nursing and
Health Sciences, 12(1), 94–104.
Mason-Hawkes, J., & Holm, K. (1989). Casual modeling: A com-
parison of path analysis and LISREL. Nursing Research, 38(5),
312–314.
Matthews, E. E., Cook, P. F., & Terada, M. (2010). Randomizing
research participants: Promoting balance and concealment in
small samples. Research in Nursing & Health, 33(3), 243–253.
Maxwell, S. E., & Delaney, H. D. (2004). Designing experiments
and analyzing data: A model comparison perspective (2nd ed.).
Mahway, NJ: Lawrence Erlbaum Associates.
McCain, L.J., & McCleary, R. (1979). The statistical analysis of the
simple interrupted time-series quasi-experiment. In T. D. Cook &
D. T. Campbell (Eds.), Quasi-experimentation: Design and analy-
sis issues for field settings (pp. 233–293). Chicago, IL: Rand
McNally.
McConnell, A. (1976). Effect of knowledge of results on attitude
formed toward a motor learning task. Research Quarterly in Exer-
cise & Sport, 47(3), 394–399.
McCorkle, R., & Young, K. (1978). Development of a symptom
distress scale. Cancer Nursing, 1(5). 373–378.
Meghani, S. H., & Keane, A. (2007). Preference for analgesic treat-
ment for cancer pain among African Americans. Journal of Pain
& Symptom Management, 34(2), 136–147.
Meinert, C. L., & Tonascia, S. (1986). Clinical trials: Design,
conduct, and analysis. New York, NY: Oxford University
Press.
Melnyk, B. M., & Fineout-Overholt, E. (2011). Evidence-based
practice in nursing & healthcare: A guide to best practice (2nd
ed.). Philadelphia, PA: Lippincott Williams, & Wilkins.
Morrison, D. M., Hoppe, M. J., Gillmore, M. R., Kluver, C., Higa,
D., & Wells, E. A. (2009). Replicating an intervention: The
tension between fidelity and adaptation. AIDS Education and Pre-
vention, 21(2), 128–140.
Munro, B. H. (2005). Statistical methods for health care research
(5th ed.). Philadelphia, PA: Lippincott Williams & Wilkins.
Muntlin, Å., Carlsson, M., Säfwenberg, U., & Gunningberg, L.
(2011). Outcomes of a nurse-initiated intravenous analgesic pro-
tocol for abdominal pain in an emergency department: A quasi-
experimental study. International Journal of Nursing Studies,
48(1), 13–23.
Murray, D. M. (1998). Design and analysis of group-randomized
trials. New York, NY: Oxford University Press.
Norris, A. E. (2005a). Path analysis. In B. H. Munro (Ed.), Statisti-
cal methods for health care research (5th ed.) (pp. 377–403).
Philadelphia, PA: Lippincott Williams, & Wilkins.
Norris, A. E. (2005b). Structural equation modeling. In B. H. Munro
(Ed.), Statistical methods for health care research (5th ed.) (pp.
405–434). Philadelphia, PA: Lippincott Williams, & Wilkins.
Nursing Interventions Classification (NIC, 2011). Overview:
Nursing interventions classification (NIC). Retrieved from http://
www.nursing.uiowa.edu/cncce/nursing-interventions-classification-
overview/.
Ottenbacher, K. (1992). Impact of random assignment on study
outcome: An empirical examination. Controlled Clinical Trials,
13(1), 50–61.
Peterson, M. J., Schwab, W., van Oostrom, J. H., Gravenstein, N.,
& Caruso, L. J. (2010). Effects of turning on skin-bed interface
pressures in healthy adults. Journal of Advanced Nursing, 66(7),
1556–1564.
Piantadosi, S. (1997). Clinical trials: A methodologic perspective.
New York, NY: Wiley.
Pierce, C., Fahs, P. S., Dura, A., Fronczek, A., Morgan, L. L.,
Leppert, T., et al. (2011). Raising stroke awareness among rural
dwellers with a Facts for Action to Stroke Treatment-based Edu-
cational Program. Applied Nursing Research, 24(2), 82–87.
Pocock, S. J. (1996). Clinical trials: A practical approach. New
York, NY: Wiley.
Polit, D. F., Gillespie, B. M., & Griffin, R. (2011). Deliberate igno-
rance: A systematic review of blinding in nursing clinical trials.
Nursing Research, 60(1), 9–16.
Reichardt, C. S. (1979). The statistical analysis of data from non-
equivalent group designs. In T. D. Cook & D. T. Campbell (Eds.),
Quasi-experimentation: Design and analysis issues for field set-
tings (pp. 147–206). Chicago, IL: Rand McNally.
Ryan-Wenger, N. A. (2010). Evaluation of measurement precision,
accuracy, and error in biophysical data for clinical research and
practice. In C. F. Waltz, O. L. Strickland, & E. R. Lenz (Eds.),
Measurement in nursing and health research (4th ed.) (pp. 371–
383). New York, NY: Springer Publishing Company.
Salvador, P. T. (2006). Development of an oral care guide for
patients undergoing autologous stem cell transplantation. Cana-
dian Oncology Nursing Journal, 16(1), 18–20.
Sandelowski, M. (1996). One is the liveliest number: The case
orientation of qualitative research. Research in Nursing & Health,
19(6), 525–529.
 CHAPTER 11  Selecting a Quantitative Research Design 263
Santacroce, S. J., Maccarelli, L. M., & Grey, M. (2004). Methods:
Intervention fidelity. Nursing Research, 53(1), 63–66.
Santana-Sosa, E., Barriopedro, M. I., López-Mojares, L. M., Pérez,
M., & Lucia, A. (2009). Exercise training is beneficial for
Alzheimer’s patients. International Journal of Sports Medicine,
29(10), 845–850.
Sawyer, A. M., Deatrick, J. A., Kuna, S. T., & Weaver, T. E. (2010).
Differences in perceptions of the diagnosis and treatment of
obstructive sleep apnea and continuous positive airway pressure
therapy among adherers and nonadherers. Qualitative Health
Research, 20(7), 873–892.
Schlairet, M. C., & Pollock, J. W. (2010). Equivalence testing of
traditional and simulated clinical experiences: Undergraduate
nursing students’ knowledge acquisition. Journal of Nursing Edu-
cation, 49(1), 43–47.
Schneider, S. M. (1998). Effects of virtual reality on symptom
distress in children receiving cancer chemotherapy. Dissertation
Abstracts International: Section B: The Sciences & Engineering,
59(5-B), 2126.
Schulz, K. F., Altman, D. G., & Moher, D. (2010). CONSORT
2010 Statement: Updated guidelines for reporting parallel group
randomized trials. Annals of Internal Medicine, 152(11),
726–733.
Shadish, W. R., Cook, T. D., & Campbell, D. T. (2002). Experimen-
tal and quasi-experimental designs for generalized causal infer-
ence. Boston, MA: Houghton Mifflin.
Sharma, N. K., Ryals, J. M., Gajewski, B. J., & Wrights, D. E.
(2010). Aerobic exercise alters analgesia and neurotrophin-3 syn-
thesis in an animal model of chronic widespread pain. Physical
Therapy, 90(5), 714–725.
Sidani, S., Doran, D., Porter, H., LeFort, S., O’Brien-Pallas, L. L.,
Zahn, C., et al. (2007). Outcomes of nurse practitioners in acute
care: An exploration. Internet Journal of Advanced Nursing Prac-
tice, 8(1), 15.
Simons, S. R., Stark, R. B., & DeMarco, R. F. (2011). A new, four-
item instrument to measure workplace bullying. Research in
Nursing & Health, 34(2), 132–140.
Spector, P. E. (1981). Research designs. Beverly Hills, CA: Sage.
Spielberger, C. D., Edwards, C. D., Lushene, R. E., et al. (1978).
Manual for the State-Trait Anxiety Inventory. Palo Alto, CA: Con-
sulting Psychologist Press.
Spielberger, C. D., Gorsuch, R. L., Lushene, R., Vagg, P. R., &
Jacobs, G. A. (1983). Manual for the State-Trait Anxiety Inventory
for Children. Palo Alto, CA: Consulting Psychologist Press.
Sprague, M., & Chang, J. C. (2011). Integrative approach focusing
on acupuncture in the treatment of chronic complex regional pain
syndrome. The Journal of Alternative and Complementary Medi-
cine, 17(1), 67–70.
Sterling, Y. M., & McNally, J. A. (1992). Single-subject research
for nursing practice. Clinical Nurse Specialist, 6(1), 21–26.
Stout, R. L., Wirtz, P. W., Carbonari, J. P., & Del Boca, F. K. (1994).
Ensuring balanced distribution of prognostic factors in treatment
outcome research. Journal of Studies in Alcoholism, 12(Suppl.),
70–75.
Strand, T. & Lindgren, M. (2010). Knowledge, attitudes and barriers
towards prevention of pressure ulcers in intensive care units: A
descriptive cross-sectional study. Intensive and Critical Care
Nursing, 26(6), 335–342.
Sutton, T. R. (2007). Anxiety and preoperative education in the
senescent patient (Doctoral dissertation, Mountain State Univer-
sity). Retrieved from ProQuest Dissertations and Theses database.
(Document ID 1446895).
Thompson, S. K. (2002). Sampling (2nd ed.). New York, NY: John
Wiley & Sons.
Tripp-Reimer, T., Woodworth, G., McCloskey, J. C., & Bulechek,
G. (1996). The dimensional structure of nursing interventions.
Nursing Research, 45(1), 10–17.
Trzesniewski, K. H., Donnellan, M. B., & Lucas, R. E. (Eds.)
(2011). Secondary data analysis: An introduction for psycholo-
gists. Washington, DC: American Psychological Association.
Tyzenhouse, P.S. (1981). Technical notes: The nursing clinical trial.
Western Journal of Nursing Research, 3(1), 102–109.
U.S. Department of Health and Human Services (U.S. DHHS)
(1968). The Framingham study: An epidemiological investigation
of cardiovascular disease (USDHHS Publication No. RC667F813).
Bethesda, MD: Author.
U.S. Department of Health and Human Services (U.S. DHHS)
(2000). Healthy People 2010: With understanding and improving
health and objectives for improving health. Retrieved from
www.healthypeople.gov/Publications/.
van Doormaal J. E., van den Bemt P. M., Zaal R. J., Egberts A. C.,
Lenderink B. W., Kosterink J. G., et al. (2009). The influence that
electronic prescribing has on medication errors and preventable
adverse drug events: An interrupted time-series study. Journal of
the American Medical Informatics Association, 16(6), 816–825.
Vartanian, J. P. (2011). Secondary data analysis. New York, NY:
Oxford University Press.
Waltz, C. F., Strickland, O. L., & Lenz, E. R. (2010). Measurement
in nursing and health research (4th ed.). New York, NY: Springer.
Whitehead, J. (1992). The design and analysis of sequential clinical
trials. New York, NY: Ellis Horwood.
Windle, P. E. (2010). Secondary data analysis: Is it useful and valid?
Journal of PeriAnesthesia Nursing, 25 (5), 323–324.
Winzer, K., Sauerbrei, W., Braun, M., Liersch, T., Dunst, J., Guski,
H., & Schumacher, M. (2010). Radiation therapy and tamoxifen
after breast-conserving surgery: Updated results of a 2 x 2 ran-
domized clinical trial in patients with low risk of recurrence.
German Breast Cancer Study Group (GBSG). European Journal
of Cancer, 46(1), 95–101.
Wooding, W. M. (1994). Planning pharmaceutical clinical trials:
Basic statistical principles. New York, NY: Wiley.
Wyatt, G., Sikorskii, A., Rahbar, M. H., Victorson, D., & Adams,
L. (2010). Intervention fidelity: Aspects of complementary and
alternative medicine research. Cancer Nursing, 33(5), 331–342.
Yamada, J., Stevens, B., Sidani, S., Watt-Watson, J., & De Silva, N.
(2010). Content validity of a process evaluation checklist to
measure intervention implementation fidelity of the EPIC Inter-
vention. Worldviews of Evidence-Based Nursing, 7(3), 158–164.
Zhang Y., Adams A. S., Ross-Degnan D., Zhang F., & Soumerai
S.B. (2009). Effects of prior authorization on medication discon-
tinuation among Medicaid beneficiaries with bipolar disorder.
Psychiatric Services, 60(4), 520–527.

12
CHAPTER
Qualitative Research Methodology
Q
ualitative researchers begin the research
process with reviewing the literature and iden-
tifying a research problem. Depending on the
research problem and type of knowledge that is
needed, the qualitative researcher develops objectives
or questions and determines which philosophical
approach will be used to guide the study (see Chapter
4). The early steps of the qualitative research process,
which is similar to the quantitative research process,
are explored in Chapters 5 and 6. Other steps in the
research process are implemented differently in or are
unique to qualitative studies. In this chapter, informa-
tion about qualitative methodology will be provided
so that you can understand the process and envision
what the experience will be like if you conduct a
qualitative study.
Qualitative analysis techniques use words rather
than numbers as the basis of analysis. In qualitative
analysis, reasoning flows from the images, documents,
or words provided by the participant toward more
abstract concepts and themes. This reasoning process,
inductive thinking, guides the organizing, reducing,
and clustering of the data. To achieve the goal of
describing and understanding participant perspectives,
qualitative methods of sampling, data gathering, and
analyses allow for more flexibility than the methods
of the quantitative paradigm. Because data analysis
begins as data are gathered, insights from early data
may suggest additional questions that need to be asked
or other modifications to the study methods. For
example, suppose a researcher was conducting a
grounded theory study about personal identity after
losing a limb during military combat. During the inter-
views, a participant mentions feeling he is an imposter
when others treat him as a war hero. Although the
planned interview questions did not include a question
about feelings related to the reactions of others to
military service, the researcher may choose to add such
a question for subsequent interviews. The researcher
264
  http://evolve.elsevier.com/Grove/practice/
may adapt the data collection or analysis strategies
during the study; however, changes are not impulsive
and must be supported with clear rationale. These
changes are documented as part of maintaining rigor.
Maintaining rigor in the context of flexibility can
be difficult. Therefore, we suggest that you seek
primary and additional sources of guidance for under-
standing the philosophical base you plan to use, as
well as the process of collecting and analyzing qualita-
tive data. A research mentor, especially a researcher
with more experience with the methods you are using,
can be invaluable. By sharing their personal experi-
ences with the mentees, research mentors can guide
less experienced researchers in application of research
principles in qualitative studies (Kostovich, Saban, &
Collins, 2010).
This chapter provides examples of qualitative
methods used to gather, analyze, and interpret data.
Literature reviews, theoretical frameworks, study
purposes, and research questions or objectives are
described in the context of qualitative methods,
because these are steps in the research process that are
implemented somewhat differently in qualitative
studies. The chapter also includes qualitative sampling
and the data collection methods of observation, inter-
views, focus groups, and electronically mediated data.
Data analysis strategies are described, and examples
are provided. The chapter ends with a presentation of
methods specific to different philosophical approaches.
Clinical Context and
Research Problems
Qualitative studies are motivated by the desire to
know more about a phenomenon, a social process, or
a culture from the perspectives of the people who are
experiencing the phenomenon, involved in the social
process, or living in the culture (Creswell, 2007). The
 CHAPTER 12  Qualitative Research Methodology 265
motivation may be that nurses realize that patient
teaching is not effective with a specific group. A new
project may be planned for low-income teenage
mothers, but all those implementing the project are
more than 40 years of age and have above average
incomes. A hurricane ravages a community, and disas-
ter relief efforts are not well received by the commu-
nity. Persons with sickle cell anemia are living past
age 60 years, and previous studies are focused on
younger persons who are being diagnosed with the
disease. Any of these situations may indicate the need
for understanding the insider’s perspective that could
be addressed by a qualitative study.
For example, Lesniak (2010) established the need
for her phenomenological study by defining the phe-
nomenon of self-injury from the literature, indicating
the problem of uncaring treatment, and identifying the
nurses who could most benefit from the information:
“Self-injury behavior is the nonsuicidal, deliberate
infliction of a wound to oneself in an attempt to seek
expression. Self-injury is more prevalent in the
adolescent population, and often emergency and
advanced practice nurses are the health professionals
who encounter this phenomenon.… [M]any of these
adolescents have reported receiving less than caring
treatment in those emergency departments. Emer-
gency nurses are on the frontline in many emergen-
cies involving adolescents; therefore, a working
knowledge of self-injury behavior would impact their
practice. It would benefit them to be provided infor-
mation concerning the characteristics of adolescents
who are self-injuring, how the behavior is expressive
in nature, and the repeating emotional patterns of
those who self-injure (Lesniak, 2008).” (Lesniak,
2010, pp. 137-138)
Swanlund (2010) provided the rationale for her
exploratory descriptive study by describing the
research findings related to medication management
in older adults. She noted that other researchers had
focused on medication management for a single
illness, instead of the more realistic situation of taking
medications for multiple chronic illnesses. “Addition-
ally, the research questions for these studies were
derived from the healthcare providers’ perspectives.
These studies have left many questions unanswered
regarding the reasons for nonadherence from the per-
spective of the older adult” (Swanlund, 2010, p. 23).
On the basis of this reasoning, Swanlund asked 27
adults older than 74 years to describe situations that
helped or hindered their ability to manage their car-
diovascular medications.
Literature Review for
Qualitative Studies
Both Lesniak (2010) and Swanlund (2010) conducted
a review of the literature and referred to the conclu-
sions of their review, which confirmed a lack of evi-
dence about their topics. However, some qualitative
researchers defer the literature review until after data
collection and analysis to avoid biasing their analysis
and interpretation of the data (Walls, Parahoo, &
Fleming, 2010). Most often, qualitative researchers
review the literature at the beginning of the process to
establish the need for the study and to provide guid-
ance for the development of data collection methods.
A more thorough review of published research find-
ings and theories may occur during data analysis and
interpretation to “place the findings of the study in the
context of what is already known about the phenom-
ena” (Speziale & Carpenter, 2007).
Theoretical Frameworks
Most qualitative researchers do not identify specific
theoretical frameworks during the design of their
studies, as is expected for quantitative studies. The
concern of most qualitative researchers is that design-
ing a study in the context of a theory will influence
the researcher’s thinking and result in findings that are
meaningful in the theoretical context but may not be
true to the participants’ perspectives on the topic.
However, the philosophical bases for the various
approaches to qualitative studies provide theoretical
grounding for qualitative studies without predisposing
the data analysis to a single interpretation.
Theory is an explicit component in some qualita-
tive research designs. The expectation of grounded
theory researchers is that qualitative research is induc-
tive in nature and that theory will emerge from the
data (Speziale & Carpenter, 2007; Walls et al., 2010).
Parse’s (2001) research methodology begins with
selecting phenomena to study that are universal human
experiences, move through analysis using Parse’s
theoretical principles as a guide, and end with evaluat-
ing the findings for their contribution to her theory.
Researchers who used Parse’s research methodology
usually present their findings using language from
Parse’s theory. Chen (2010) used Parse’s research
method to study the lived experiences of persons
with spinal cord injuries. From the stories told by the
266 UNIT TWO  The Research Process
participants, Chen stated the structure of the experi-
ence in one sentence: “The lived experience of moving
forward is confronting difficulties, going on and
finding self-value and confidence to affirm oneself
while co-creating successes amid opportunities and
restrictions” (p.1138). Chen further transformed the
structure through conceptual integration into the con-
clusion that “moving forward is transforming the
valuing of originating ‘enabling-limiting’ ” (p. 1138).
Exploratory qualitative study design may benefit
from making explicit the researcher’s theoretical per-
spective on the research problem. Scott and Wilson
(2011) used the ecology theory for their qualitative
study of African Americans living in the South in the
United States. The ecology theory was consistent with
their research objective, “identifying potential social
determinants of health among African Americans in a
rural community in the Deep South, from the perspec-
tives of African American community members”
(p. 3). In Singapore, researchers Cheng, Qin, and Tee
(2009) interviewed patients with hematological disor-
ders about their experiences in isolation. They noted
that Maslow’s (1987) hierarchy of needs theory and
Roy’s Adaptation Theory (1976) guided the develop-
ment of the interview questions and the data analysis.
Qualitative researchers who use frameworks during
study development must maintain intellectual honesty
to prevent the theoretical perspective from obscuring
the perspectives of the participants. Your decision
about whether to identify a theoretical perspective
should be consistent with the research approach you
have chosen. If a theoretical perspective has shaped
your views of a research problem, you should acknowl-
edge that influence and indicate explicitly the compo-
nents of the study guided by the theory.
Purposes
The purpose should clearly identify the goal or aim of
the study that has emerged from the research problem
and literature review. The purpose of qualitative
studies should include the phenomenon of interest, the
population, and the setting (see Chapter 5). Ask your-
self, “Can I achieve this purpose with a qualitative
study?” Study purposes such as testing an intervention
and determining the effectiveness of a program are not
consistent with qualitative approaches. However, a
qualitative researcher could address the participants’
experiences with the intervention or their perceptions
about a program. The purpose of qualitative studies
will vary slightly depending on the qualitative
approach that is being used. For example, note in
Table 12-1 that the phenomenological study focused
TABLE 12-1  Selected Examples of Purpose
Statements in Qualitative Studies
Qualitative
Approach Purpose Statement
Phenomenological “[E]xplore patients’ lifeworld and way
research of managing life with advanced PD
[Parkinson’s disease] prior to DBS
[deep brain stimulation] and to
illuminate what they expect from
life following DBS” (Haahr,
Kirkevold, Hall, & Ostergaard,
2010, p. 410).
Grounded theory “[D]escribe the processes that patients
research in palliative home care use to
maintain hope” (Ollson, Ostlund,
Strang, Grassman, & Friedrichsen,
2010, p. 607).
Ethnographic “[U]nderstand the culture care value
research and beliefs of women during
addiction and recovery” (Lange,
2007, p. 74).
Exploratory “[D]etermine the situations or variables
qualitative that influence successful
research cardiovascular medication
management, as perceived by
community-dwelling adults over age
74” (Swanlund, 2010, p. 23).
Historical “[C]harts the international travels of
research four especially mobile nurses… tells
these nurses’ stories and analyzes
their ideologies of development and
progress” (Irwin, 2011, p. 78).
on life-world of the participants and the grounded
theory study focused on the processes used to maintain
hope. The purposes listed in Table 12-1 are consistent
with the study’s identified philosophical approach.
Research Objectives
or Questions
Hypotheses are not appropriate for qualitative studies
because hypotheses specify outcomes of studies and
variables that are to be manipulated or measured,
actions that are not consistent with the philosophical
orientation of qualitative research. Qualitative
researchers may identify research objectives or ques-
tions to connect the purpose of the study to the plan
for data collection and analysis. Because qualitative
research is more open-ended and the focus is on the
participants’ perspectives, qualitative researchers may
not specify research objectives or research questions
 CHAPTER 12  Qualitative Research Methodology 267
in order to avoid prematurely narrowing the topic. In
their study of the occupational stress of children’s pal-
liative care nurses, McCloskey and Taggart (2010)
identified two research objectives:
“Two objectives were set. Firstly, to explore the
experiences of occupational stress within nurses
working in three distinct young people’s palliative
care services (i.e., hospice care, hospital care and
community care) and secondly, to examine the con-
sequences of such stress upon the nurses’ lives.”
(McCloskey & Taggart, 2010, p. 234)
These objectives clearly stated and did not limit
what the researcher might find. These are ideal char-
acteristics of objectives for a qualitative study. McClo-
skey and Taggart (2010) found four themes that
described the experiences of the nurses: work demands,
relationships, maintaining control, and support and
roles. The theme of relationships was seen as particu-
larly challenging because of emotional demands and
ethical conflicts.
A research question establishes “the perimeters of
the project and suggest the methods to be used for data
gathering and analysis” (Corbin & Strauss, 2008, p.
19). Research questions may be broadly stated or may
be based on a theoretical framework if the researcher
has identified a framework. Schumacher (2010) iden-
tified Leininger’s (1985) theory of culture care diver-
sity and universality as the conceptual guide for her
ethnographic study with persons in rural Dominican
Republic. Schumacher identified four comprehensive
research questions based on the theory. Questions 2
and 4 are provided as examples:
“Research Question 2: In what ways do technological,
religious, philosophical, kinship and social, cultural,
political and legal, economic, and educational factors
influence care meanings, beliefs, and practices of
Dominicans?…
“Research Question 4: In what ways can Leini-
nger’s care modes of preservation/maintenance,
accommodation/negotiation, and repatterning/
restructuring be used to plan nursing care that is
culturally congruent for rural Dominicans?” (Schu­
macher, 2010, p. 94)
Schumacher (2010) developed research questions
that were congruent with Leininger’s theory. When a
researcher identifies a conceptual framework or
guiding theory for a study, the research questions
should be congruent with the theory. Schumacher
clearly stated appropriately broad questions that
included the study’s sample (rural Dominicans) and
desired outcome (plan culturally congruent nursing
care). The following three themes were identified,
each consistent with Leininger’s theory: “(a) family
presence is essential for meaningful care experiences
and care practices, (b) respect and attention are central
to the meaning of care, and (c) rural Dominicans both
value and use generic (folk) and professional care
practices” (Schumacher, 2010, p. 97).
Obtaining Research Participants
The goal of sampling for quantitative studies is to
obtain data from a subgroup of a population that is
statistically representative of the population with the
intent of being able to generalize the findings to the
population (see Chapter 15). Qualitative researchers
seek participants who have experienced the phenom-
enon of interest (Speziale & Carpenter, 2007) and are
thus deemed to be “information-rich cases for in-depth
study” (Liamputtong & Ezzy, 2005, p. 46). For ethno-
graphic studies, the participants may also include key
informants who are knowledgeable of the culture being
studied. The selection of participants is nonrandom and
may not be totally specified before the study begins.
Depending on the research question and the aims
of the study, the researcher may use more than one
sampling strategy during the study. For example, a
researcher who is studying the experience of reacting
to a diagnosis of breast cancer may choose to select
only women who have not previously been diagnosed
with cancer, have not had a family member die from
breast cancer, and have been diagnosed within the last
6 weeks. This approach to sampling is called criterion
sampling. Similar logic can be applied to identify par-
ticipants for a focus group, when it is desirable to have
participants who can identify with each other’s experi-
ences. When used for focus groups, this sampling
method is called homogeneous group sampling (Liam-
puttong & Ezzy, 2005). Table 12-2 provides defini-
tions and references for sampling strategies that are
frequently used by qualitative researchers. These sam-
pling strategies are not mutually exclusive and may be
labeled differently by various researchers.
The sample for a rigorous qualitative study is not
as large as the sample for a rigorous quantitative study.
The researcher stops collecting data when enough
rich, meaningful data have been obtained to achieve
the study aims. For new researchers, this answer to
“How big should my sample be?” is totally unsatisfac-
tory. When applying for human subjects’ approval, the
268 UNIT TWO  The Research Process
TABLE 12-2  Sampling Strategies used by
Qualitative Researchers
Sampling Definition
Convenience Inviting participants from a location or group
sampling because of ease and efficiency.
Snowball Researcher asks participants to refer others
sampling who have had similar experiences to
participate in the study; also called chain
sampling or network sampling.
Historical Exhaustive search for all relevant, surviving
sampling primary and secondary sources about an
event or phenomenon that occurred in the
past (Lundy, 2012).
Purposive Recruitment of participants as sources of data
sampling that can provide and expand upon the data
needed to achieve the study aims.
Theoretical Recruitment of participants who are
sampling* considered to be best sources of data related
to the study (Wuest, 2012); additional
participants may be recruited to validate or
expand upon emerging concepts; associated
with grounded theory approaches.
Criterion Recruitment of participants who do or do not
sampling* have certain characteristics deemed to affect
the phenomena being studied (Liamputtong
& Ezzy, 2005).
Maximum Recruitment of participants who represent
variation potentially different experiences related to
sampling* the domain of interest (Seidman, 2006).
Critical case Recruitment of participants whose experiences
sampling* with the research topic are expected to be
very different.
Deviant case Recruitment of participants who may be
sampling* outliers or represent extreme cases of the
domain of interest.
*Considered by some authors to be subtypes of purposive sampling.
researcher will be asked the maximum sample size.
Giving a generous range of 12 to 25 can be a way to
answer this question but will depend on the study
design. Researchers who use focus groups often have
larger samples. The actual number depends on when
data saturation is achieved. Data saturation is the
point when new data begins to repeat what has already
been found. Patterns emerge in the data. The researcher
has the data needed to answer the research question
and remain true to the principles of the study design.
Marshall and Rossman (2011) indicate that a better
term is theoretical sufficiency, because one can never
completely know all there is to know about a topic.
Chapter 15 also provides some detailed ideas about
sampling methods and sample size in qualitative
studies.
Researcher-Participant Relationships
One of the important differences between quantitative
research and qualitative research is the nature of rela-
tionships between the researcher and the individuals
being studied. The nature of these relationships has an
impact on the data collected and their interpretation.
In varying degrees, the researcher influences the indi-
viduals being studied and, in turn, is influenced by
them. The mere presence of the researcher may alter
behavior in the setting. The researcher desires to
connect at the human level with the participant (Corbin
& Strauss, 2008). Although this involvement is con-
sidered a source of bias in quantitative research, quali-
tative researchers consider it to be a natural and
necessary element of the research process. Shorter and
Stayt (2010, p. 161) noted that “the research findings
are thus a product of a co-constructive process between
researcher and participant.”
The researcher’s personality is a key factor in
qualitative research, in which skills in empathy and
intuition are cultivated. You will need to become
closely involved in the subject’s experience to inter-
pret it. Participants need to feel safe and to be able to
trust the researcher prior to sharing their deepest
experiences with the researcher. It is necessary for
you to be open to the perceptions of the participants
rather than attaching your own meaning to the experi-
ence. To do this, you need to be aware of personal
experiences and potential biases related to the phe-
nomenon being studied. It is helpful to document
these experiences and potential biases before and
during the study in a reflective journal, to be able to
set them aside or bracket them. For example, a
researcher who plans to interview women undergoing
irradiation for breast cancer would need to acknowl-
edge that his/her own mother died from complica-
tions of breast cancer. This awareness and ability to
be involved with the participants and yet be able to
analyze the data abstractly with intellectual honesty is
called reflexivity. Reflexivity consists of the ability to
be aware of your biases and past experiences that
might influence how you would respond to a partici-
pant or interpret the data. This ability is critical in
qualitative studies because data emerge from a rela-
tionship with the participant and are analyzed in the
mind of the researcher, rather than through a statisti-
cal program (Wolf, 2012).
Data Collection Methods
Because data collection occurs simultaneously with
data analysis in qualitative studies, the process is
complex. Collecting data is not a mechanical process
 CHAPTER 12  Qualitative Research Methodology 269
that can be completely planned before it is initiated.
The researcher as a whole person is completely
involved—perceiving, reacting, interacting, reflecting,
attaching meaning, and recording. For a particular
study, the researcher may need to address data collec-
tion issues related to relationships between the
researcher and the participants, reflect on the mean-
ings obtained from the data, and organize management
and reduction of large volumes of data. Qualitative
researchers are not limited to a single type of data or
collection method during a study. For example, one
research team used interviews and focus groups in
their study of fall prevention for older people in
the community (Dickinson, Machen, Horton, Jain, &
Maddex, 2011). Lopez (2009) used data from semi-
structured interviews, nonparticipant observation, and
informal conversational interviews to describe the
processes used to make decisions about the care of
nursing home residents who become acutely ill. Quali-
tative data collection may also be combined in a study
with the collection of quantitative data. These mixed-
methods studies are described in detail in Chapter 10.
Observations, interviews, and focus groups are the
most common methods of gathering qualitative data,
and each is described in detail, followed by an example
from the literature. Electronic means of qualitative
data collection, such as photographs, videos, and
blogs, are briefly described as well. Following the
general types of data collection, methods specific to
each qualitative approach are discussed.
Observations
In many qualitative studies, the researcher observes
social behavior and may participate in social interac-
tions with those being studied. Observation is collect-
ing data through listening, smelling, touching, and
seeing, with an emphasis on what is seen. Even when
other data collection methods are being used, such as
interviews, you want to stay aware of your surround-
ings and attend to the nonverbal communication that
occurs between the participant and others in the imme-
diate surroundings (Marshall & Rossman, 2011).
Unstructured observation involves spontaneously
observing and recording what one sees. Although
unstructured observations give the observer freedom,
there is a risk that the observer may lose objectivity
or may not remember all of the details of the event.
The most complete way to collect observation data is
to video-record the situation being studied, but doing
so may alter the behavior of those being observed or
may not be possible because of confidentiality con-
cerns. If video recording is not possible, then the
researcher may take notes during the observation
periods. If taking notes is a problem, then the researcher
needs to record the observations as soon as possible
afterwards.
Collecting data through unstructured observation
may evolve into more structured observations. The
researcher may begin with few predetermined ideas
about what will be observed. As the study progresses,
the researcher clarifies the situations or areas of focus
that are most relevant to the research questions and
begins to structure the observation. A researcher
observing parent behavior in an ambulatory pediatric
care clinic may initially focus on the interaction of
parents with their children in the waiting area and in
the room with the provider. During data collection,
the researcher begins to notice common nurturing
behaviors of the parents and, from these observations,
develops a checklist to use while observing. In this
way, the researcher has structured the observations
that might be the focus of future studies. Other
researchers may enter the setting with a checklist or
tool for documenting observations and revise the tool
as needed. The notes made during and immediately
following the observations are called field notes
(Speziale & Carpenter, 2007). Recording observa-
tions can be as simple as using a pad and writing
utensil in a public place or as sophisticated as pro-
ducing an electronic diagram of the locations of
nurses by having them wear positioning devices.
Observations may be supplemented by taking photo-
graphs in the setting or video recording an observa-
tion. After the observation, the diagrams of the
nurses’ positions, the photographs, or the videos may
serve to remind the observer of specific elements of
the situation. In addition, the researcher may analyze
the video by viewing short segments and making
notes about each. By reflecting on the photographs
and videos, the researcher may identify details that
were not captured in the actual observation. The field
notes, notes about the electronic records made during
the observation, and the researcher’s memories of
and reflections about being in the setting are the data
that are analyzed.
The researcher, by virtue of being in the setting,
becomes a participant to some degree. The balance
between participation and observation has been
described in four ways. The first is complete partici-
pation. The people in the situation may not be aware
that the participant is a researcher (Speziale &
Carpenter, 2007). In public settings, a researcher can
ethically observe people and interactions without
obtaining permission (Liamputtong & Ezzy, 2005). In
less public settings, the researcher may observe others
who learn later that he or she is a researcher. When
270 UNIT TWO  The Research Process
the researcher’s role is unknown to the study partici-
pants, they need to have consented to incomplete dis-
closure before the study is conducted. After the study,
they must be debriefed regarding the undisclosed
aspects of the study (see Chapter 9). The participants
have the option as to whether the data the researcher
collected about them is included or not in the study.
When the researcher is in the participant as observer
role, participants are aware of the dual roles of the
researcher from the beginning of the study. Full
engagement in the situation may interfere with the
researcher’s ability to note important details and
move within the setting to follow an evolving situa-
tion. In these situations, the role of observer as par-
ticipant may be more appropriate. As the term
indicates, the researcher’s observer role takes priority
and is the focus of the data collection. Complete
observation occurs when the researcher remains
passive and has no direct social interaction in the
situation.
When the researcher is observing for the purposes
of data collection, multiple types of information
should be noted. The observer may gather data on the
physical setting, the people in the setting, and their
activities and interactions. Activities and interactions
can be described in terms of their frequency, duration,
precipitating factors, and organization. Hall, Pedersen,
and Fairley (2010) observed the activities and interac-
tions of nurses during their work on medical and surgi-
cal units in three hospitals. The data from 480 hours
of observations were analyzed to produce a frame-
work to describe interruptions by source, type, cause,
the activity that was interrupted, and the outcomes of
the interruptions. In addition to interviewing nurses in
a burn intensive care unit, Zengerle-Levy (2006)
observed them for 134 hours as they cared for pediat-
ric patients. Through observation, the researcher was
able to “uncover common practices that were not
articulated” about how they cared for children in the
unit in which parents were absent (p.227). Hall et al.
(2010) used observation as the primary source of data.
Zengerle-Levy (2006) used observations to supple-
ment the data collected during the interview. In both
cases, the researchers clearly described the purpose of
the observations and details such as total time spent
and types of activities observed. This information
indicates that observations were effectively used and
the study was rigorous.
Example Study Using Observation
Lauzon Clabo (2008) studied pain assessment on two
postoperative units in a teaching hospital. During the
first phase of the study, her primary method of data
collection was observation. Note the emphasis placed
on gaining access to the setting.
“[T]he focus was on gaining entry to the hospital and
simultaneously to each unit, establishing the research-
er’s role, developing relationships with the nursing
staff, becoming familiar with the general routines and
rhythms of practice, as well as beginning to map out
the social context of each unit. During this phase, the
primary method was participant observation. The
researcher (LLC) [one of the authors] was present on
the unit through the course of normal care primarily
for the purpose of observation, but also interacted
informally with staff in hallway conversations. Data
were recorded and analysed on an ongoing basis using
field notes that included a daily event log and a series
of observational, theoretical and methodological
notes.” (Lauzon Clabo, 2008, p. 533)
Lauzon Clabo (2008) followed the initial phase of
data collection with observations of individual nurses
on the day shift. Each of these observations was fol-
lowed by an interview with the nurse.
“The observations provided an overall sense of how
any one nurse went about conducting a pain assess-
ment, while the semi-structured interviews captured
more detail regarding the nurse’s actual thinking,
especially in regard to the approaches used to assess
each client’s pain.” (Lauzon Clabo, 2008, p. 533)
The final phase of the study was a focus group on
each unit, during which Lauzon Clabo (2008) pre-
sented a rough draft of the findings for that unit. On
Unit A, the nurses had similar expectations for level
and type of pain for specific surgical procedures. They
found caring for patients whose pain did not follow
the usual trajectory to be challenging because of the
need to contact the physician frequently and the desire
to keep the physician happy. On Unit B, the nurses
focused on the patient’s description of his or her pain
and supplemented it with objective and subjective
data, such as the use of pain medications. Lauzon
Clabo (2008) concluded that pain assessment was a
synthesis of the patient’s narrative, evidence from
behaviors and medication usage, and expectations
based on experience. This study had rigorous method-
ology as evidenced by the researcher’s going back to
the nurses as individuals (interviews) and as a group
 CHAPTER 12  Qualitative Research Methodology 271
(focus group) to supplement and validate the initial
findings. This rigor gives the reader of the study con-
fidence to view the results as credible. The clinical
implications are that nurses must allow the patient’s
narrative to take precedence over behaviors, medica-
tion usage, and the nurses’ expectations in the assess-
ment of postoperative pain.
Interviews
Interviews are interactions between the participant
and the qualitative researcher that produce data as
words. The researcher as an interviewer seeks infor-
mation from a number of individuals, whereas the
focus group strategy is designed to obtain the perspec-
tive of the normative group, not individual perspec-
tives. Interviews may also be conducted in quantitative
studies to assist subjects in the completion of a survey
or questionnaire. This assistance may include reading
the questions to subjects with limited literacy and
documenting their responses to the questions in person
or over the phone. The focus of this section is inter-
viewing in qualitative studies.
Depending on the research question, the qualitative
researcher may conduct multiple interviews with each
participant or may follow an initial interview with
another during which the participant can review the
researcher’s description of the first interview. Seidman
(2006) recommends that the researcher interview each
participant three times for phenomenological studies.
The first interview is focused on a life history, the
second on details of the phenomenon, and the third on
reflection on the experience. Using multiple inter-
views allows the relationship between the researcher
and the participant to develop. Over time, the partici-
pant may learn to trust the researcher more and reveal
insights about his or her experiences that contribute to
the study’s findings. Follow-up interviews may be
used to share the results of the ongoing data analysis
with participants and ask additional questions for
clarification. Multiple interviews may be required to
study an ongoing process. For a grounded theory
study of women undergoing coronary artery bypass
graft surgery, Banner (2010a) interviewed the women
preoperatively and at 6 weeks and 6 months after
surgery.
In addition to determining how many times each
participant will be interviewed, the researcher will
need to plan the interview location, format, and
method of documenting the interview. Interviews
might be conducted in a room in a public library, a
fast-food restaurant at an off-peak time, an exam
room in a clinic, or the participant’s home. The loca-
tion should be selected as a neutral place that has
private areas, is convenient for the participant
(Speziale & Carpenter, 2007), and considers the
safety of those involved. Accessibility and confiden-
tiality should also be considerations. An exam room
may not be a neutral site for a study exploring the
patient-provider relationship. During a community-
based study, the researcher’s appearance may become
associated with a stigmatized topic, such as HIV
infection, substance use, or domestic violence. A
public place may not protect the participant’s identity
and confidentiality. A participant’s home may not be
safe for the researcher to come to at certain times of
day. A participant’s home, however, can offer a sense
of comfort and familiarity for the participant and
provide the researcher insight into the participant’s
experience. Gharaibeh and Owels (2009) conducted
a qualitative study of Jordanian women who stayed
in abusive relationships (N = 28). They allowed par-
ticipants to “choose the place of the interview in
order to make them feel comfortable and secure” (p.
379). All participants chose to be interviewed in their
homes when the husband or other relatives were not
present.
The format of the interview can be unstructured,
semistructured, or structured. Unstructured inter-
views are informal and conversational and may be
useful during an ethnographical study or the early
stages of a study. Most qualitative interviews are sem-
istructured, or organized around a set of open-ended
questions. Some experts call these topical or guided
interviews (Marshall & Rossman, 2011). The degree
of guidance may be as minimal as having one initial
abstract question or prompt or as structured as multi-
ple predefined questions to narrow the interview
to specific aspects of the phenomenon being studied.
In either case, the researcher remains open to how
the participant responds and carefully words follow-
up questions or prompts to allow the emic view,
the participant’s perspective, to emerge. Structured
interviews are organized with narrower questions in
a specific order. The questions may be asked without
follow-up questions, and the researcher responses may
be scripted in a structured interview (Marshall &
Rossman, 2011). Having this level of structure may
decrease the anxiety of less experienced interviewers
but may result in findings that reflect the etic, or out-
sider’s, view.
The words spoken and the nonverbal communica-
tion during an interview are the data. Some research-
ers prefer to audio-record or video-record the interview
and focus on the interaction and relationship with the
participant during the interview (Banner, 2010b). A
recording of the interview results in a “transportable,
272 UNIT TWO  The Research Process
repeatable resource that allows multiple hearings or
viewings as well as access to other readers” (Nikander,
2008, p. 229). The participant must be aware that the
interview is being electronically recorded, but the less
obtrusive the equipment, the more quickly the partici-
pant will forget its presence, relax, and speak more
freely. Logistically, the researcher needs to plan ahead
to have the power cords or batteries needed for the
recording device (Banner, 2010b). Using batteries
may make the device less obtrusive. A sensitive micro-
phone will allow you to pick up even faint or distorted
voices, thereby increasing your ability to make an
accurate transcription later. If tapes are being used,
ensure that the lengths of the tapes are adequate to
record the entire interview with few interruptions to
change the tape. More likely, the recording will be
made on a digital device that can be saved on a com-
puter for transcription. In some situations, recording
devices may not be appropriate or the participant may
prefer that the interview not be recorded. During the
unrecorded interviews, the researcher may take notes
and set aside time immediately following the inter-
view to document the interview with as much detail
as possible.
Learning to Interview
Preparing to interview is critical because interviewing
is a skill that directly affects the quality of the data
produced (Marshall & Rossman, 2011). Researchers
must give themselves the opportunity to develop this
skill before they start interviewing study participants.
A skilled interviewer can elicit higher-quality data
than an inexperienced interviewer by allowing for
silence or asking a probing follow-up question without
alienating the participant. Unskilled interviewers may
not know how or when to intervene, when to encour-
age the participant to continue to elaborate, or when
to divert to another subject. The interviewer must
know how to handle intrusive questions. For practice,
conduct interviews with colleagues with experience in
interviewing (Munhall, 2012). These rehearsals will
help you identify problems before initiating the study
(Banner, 2010b). You may want to conduct one or
more trial interviews with individuals who meet the
sampling criteria to allow you to try out the proposed
questions. Practice sessions and pilot interviews also
allow you to determine a realistic time estimate for the
interviews. Researchers often underestimate the time
needed for an interview. Allow yourself enough time
so that you can conduct the interview without feeling
rushed. Be sensitive to time-related concerns of the
participants, however, and offer the option of stopping
if an interview is going longer than expected.
Participants may need to be able to catch a bus to get
home, pick up children from child care, or stop to take
a dose of medication.
Establishing a Positive Environment
for an Interview
When preparing for an interview, establish an environ-
ment that encourages an open, relaxed conversation.
Be sensitive to the physical surroundings. Sit in com-
fortable chairs so that neither you nor the participant
is facing windows with direct sunlight. Sitting at a
table may be more comfortable and provides a surface
for the participant to sign the consent form or com-
plete a demographic form. You may want to offer
water or other beverage as a way to provide time for
a social connection prior to beginning the interview.
When dressing for an interview, the researcher needs
to consider how the participant is likely to be dressed.
Dressing in formal business attire or a nursing uniform
may emphasize the power differences in the relation-
ship. Dressing too casually may be viewed as an indi-
cation that the interaction is not important to the
researcher. Power issues may affect the effectiveness
of the interview.
Conducting an Effective Interview
As the researcher, you have the power to shape the
interview agenda. Participants have the power to
choose the level of responses they will provide. You
might begin the interview with a broad request such
as “Describe for me your experience with…” or “Tell
me about.…” Ideally, the participant will respond as
though she or he is telling a story. You respond non-
verbally with a nod or eye contact to convey your
interest in what is being said. Try to avoid agreeing or
disagreeing with what the participant is saying. Being
nonjudgmental allows the participants to share their
experiences more freely. When it seems appropriate,
encourage your subject to elaborate further on a par-
ticular dimension of the topic. Use of nonthreatening
but thought-provoking questions is often called
probing. Participants may need validation that they
are providing the needed information. Some partici-
pants may give short answers, so you may have to
encourage them to elaborate. When the participant
stops talking, ask a follow-up question that reflects
back on what you have heard. Interviewer responses
should be encouraging and supportive without being
leading. Listening more and talking less is a key prin-
ciple of effective interviews (Seidman, 2006). That
includes tolerating silence. If the participant is not
talking but seems to be thinking or considering the
topic, stay quiet. Silence can be a powerful invitation
 CHAPTER 12  Qualitative Research Methodology 273
that allows the participant to show deeper emotions
and thoughts.
Problems during Interviews
Difficulties can occur during interviews. Common
problems include interruptions such as telephone calls
or text messages, “stage fright” that often arises when
the participant realizes he or she is being recorded,
failure to establish a rapport with your subject, verbose
participants, and those who tend to wander off the
subject. Turn off or silence your cell phone at the
beginning of the interview, and ask the participant to
do the same. If a participant seems paralyzed by the
presence of the recording device, move the device out
of his or her line of sight if possible. Ask demographic
questions or factual questions to ease into the inter-
view. When the participant moves to a subject that you
think is unrelated to the focus of the study, you may
want to ask how this new subject is related to previous
comments on the topic of interest. You may be sur-
prised to learn that what you perceived to be unrelated
is associated with the topic in the participant’s per-
spective. You may also need to tactfully guide the
interview back to the topic. Remind participants that
they can decline answering any question and can end
the interview at any time.
The physical, mental, and emotional condition of
the participant may cause difficulties during the inter-
view. The data you obtain are affected by characteris-
tics of the person being interviewed. These include
age, ethnicity, gender, professional background, edu-
cational level, and relative status of interviewer and
interviewee as well as impairments in vision or
hearing, speech impediments, fatigue, pain, poor
memory, disorientation, emotional state, and language
difficulties. Although institutional review boards tend
to view interviews as noninvasive, interviews are an
invasion of the psyche. For some participants, the
experience may be therapeutic. However, an interview
is capable of producing risks to the health of the par-
ticipant. Therefore, the interviewer must always avoid
inflicting unnecessary harm upon the participant.
Participants with fatigue or pain related to illness or
treatments should be offered the opportunity to stop,
take a break, or schedule a second interview for
another day.
In a phenomenological study of persons hospital-
ized for depression, Moyle (2002) noted challenges
with obtaining rich data due to the flattened emotions
and slowed cognition that may result from severe
depression. During the study, some of the participants
underwent electroconvulsive therapy that led to some
memory loss. The participants in Moyle’s study had
health professionals readily available because they
were hospitalized. In other settings, the researcher
needs to prepare to refer the participant if necessary.
Emotional expression during an interview may be
expected, depending on the topic. When the partici-
pant becomes distressed or overcome with emotion,
however, you may choose to turn off the recording
device and stop the interview completely or for a few
minutes. You may be able to continue if the participant
is able to become composed. Stay with the individual,
offer a tissue, and move closer to the participant. Rec-
ognize topics that are more likely to be distressing,
and have a plan developed for emergency assistance
if needed or a list of mental professionals available if
support or a referral is needed. For example, you
might schedule interviews in collaboration with the
chaplain or psychiatric mental health nurse practitio-
ner to ensure that one of them is available for consulta-
tion when you will be interviewing family members
whose spouses are receiving hospice care. Recognize
that you, the researcher, may also need emotional and
psychological support. The researcher may be strongly
affected by the stories of the participants. Arrange to
have a mentor or trusted friend available to talk with
before or after interviews. The researcher may need to
rest following an interview, because the experience
requires heightened focus for an extended period,
which can result in fatigue (Creswell, 2007).
Example Study Using Interviews
Greco, Nail, Kendall, Cartwright, and Messecar
(2010), in the report of their grounded theory study on
mammography decision making among older women
(N = 23), limited the sample to women who were 55
years or older and had at least one relative who had
been diagnosed with breast cancer. They clearly
described the rationale for the study, the selection of
participants, and collection of data in the following
excerpt:
“[D]ecision-making studies of women with a breast
cancer family history have largely focused on younger
women who have undergone genetic testing for
breast cancer disposition.… In this study, older
woman are defined as 55 years of age and older since
65% of the breast cancers occur in this age group.…
Women who met the study criteria and agreed to
participate were interviewed in person at a private
place of their choosing. Recruitment was discontin-
ued when the analysis provided no new information
contributing to the development of the theory. The
274 UNIT TWO  The Research Process
researcher conducted open-ended interviews using a
semi-structured interview guide. For the purpose of
describing the participants, demographic data were
collected at the end of the interview. Field notes were
maintained to describe the environmental and emo-
tional context of the interviews. Interviews were
recorded and transcribed verbatim, with all person-
ally identifying information deleted.” (Greco et al.,
2010, p. 349-350)
Greco et al. (2010) used a collaborative process to
come to consensus during the data collection and
analysis process. In addition, some participants were
interviewed a second time to assist in refining the
theory. These actions strengthened the data collection
and analysis process and improved the credibility of
the findings. The researchers produced a theory they
called “guarding against cancer” (p. 351). Study limi-
tations included an all-white sample and the use of
self-report for mammograms. Implications for prac-
tice are that health professionals recommending that a
woman obtain a mammogram was a strong motivator.
Another was that older women may believe they no
longer need to be screened and so may stop obtaining
mammograms.
Focus Groups
Focus groups were designed to obtain the partici-
pants’ perceptions in a focused area in a setting that is
permissive and nonthreatening. One of the assump-
tions underlying the use of focus groups is that interac-
tions among people can help them to express and
clarify their views in ways that are less likely to occur
in a one-on-one interview (Gray, 2009). People in a
focus group who are alike in some characteristic may
feel safer or less anxious when talking about difficult
experiences. Many different communication forms
occur in focus groups, including teasing, arguing,
joking, anecdotes, and nonverbal clues, such as gestur-
ing, facial expressions, and other body language.
Focus groups as a means of data collection serve a
variety of purposes in nursing research. Focus groups
have been used to understand the experiences of
people who are receiving care or may need care.
Researchers have used focus groups to explore the
fatigue of patients following a stroke (Finn & Stube,
2010) and the social support needs of older African
Americans who have survived cancer (Hamilton,
Moore, Powe, Agarwai, & Martin, 2010). The use of
prescription medications without medical care of
Latino immigrants (Coffman, Shobe, & O’Connell,
2008) and the excessive drinking of college students
(Dodd, Glassman, Arthur, Webb, & Miller, 2010) have
also been studied by means of focus groups. Using
focus groups, researchers have explored nurses’ expe-
riences such as the stress of providing palliative care
for children (McCloskey & Taggart, 2010) and the
strategies used in critical care units to prevent and
correct medical errors (Henneman et al., 2010). Focus
groups of nurses have also been the data collection
method for studies of the barriers school nurses
encounter when talking to parents about their chil-
dren’s weight (Steele et al., 2011) and teamwork as
experienced by nurses in the neonatal intensive care
units and emergency departments (Simmons & Sher-
wood, 2010).
Instrument development and refinement are fre-
quently based on the data collected during focus
groups. An example of instrument development was
the study conducted by Anatchkova and Bjorner
(2010). They conducted focus groups to identify key
elements of role functioning as a part of health. The
roles identified were to be used to develop items for
research instruments measuring quality of life. An
example of instrument refinement was the study con-
ducted by Owen-Smith, Sterk, McCarty, Hankeson-
Dyson, and DiClemente (2010). Their sample consisted
of persons diagnosed with AIDS who used comple-
mentary and alternative medicine (CAM) therapies.
During two focus groups, the researchers provided
the copy of an existing measure of CAM use and
asked each participant to complete the instrument. Fol-
lowing completion of the instrument, the participants
were asked to identify the strengths and limitations of
the instrument. The third and fourth focus groups were
not given the CAM instrument but were asked to
describe what CAM meant to them. By combining the
data from the focus groups, the researchers developed
a revised CAM instrument. They validated their find-
ings by having participants in a fifth focus group com-
plete a computer-based version of the revised
instrument and provide feedback. This rigorous process
produced a revised instrument that will be appropriate
for use with the persons diagnosed with AIDS.
The effective use of focus groups requires careful
planning. The location needs to be carefully selected
to ensure privacy, comfort, and safety. Meeting rooms
in public facilities such as schools, libraries, or
churches may be appropriate community locations for
focus groups, depending on the research question and
the study aims. For focus groups with specific popula-
tions, the facility used for support services may have
a quiet room that is accessible and known to
 CHAPTER 12  Qualitative Research Methodology 275
participants. Nurses or other health professionals may
participate in focus groups in the healthcare facility
but might be more forthcoming in a location away
from the facility. For a focus group with Hispanic
persons who were infected with HIV, a hospital
meeting room was selected. When the research team
approached the reception desk in a hospital waiting
area to ask the location of the room, the staff person
stated the location of the “AIDS focus group” in a loud
voice heard by the waiting visitors (Gray, 2009). In
this case, the researchers’ schedule for travel to the
out-of-town focus group did not allow for previewing
the location so they did not have the opportunity to
inform the reception staff of the name being used for
the focus group meeting. If a focus group is planned
for a sensitive topic, indicate on the invitation and any
materials the name by which the group will be identi-
fied. For example, instead of identifying the group as
the “Testicular Cancer Study,” a better name might be
the “Men’s Health Study.”
Other logistics include the expected length of
the meeting, recruiting subjects, and recording the
group interaction. Focus groups typically last from 45
minutes to 1 1 2 hours. Liamputtong and Ezzy (2005)
indicate that a group might go as long as 3 hours but
that participants will begin to tire before 2 hours have
passed. Be clear on the recruitment materials about
the expected duration of the focus group. Allow for
the time it will take to complete consent and demo-
graphic forms in determining the length of the data
collection process. Provide a reasonable estimate of
the time needed, recognizing that whether people
attend may be affected by how long the group is
expected to last.
Recruiting appropriate participants for each of the
focus groups is critical, because recruitment is the
most common source of failure. Each focus group
should consist of 4 to 12 participants (Marshall &
Rossman, 2011). If there are fewer participants, the
discussion tends to be inadequate. In most cases, par-
ticipants are expected to be unknown to one another.
However, for a focus group targeting professional
groups such as clinical nurses or nurse educators, such
anonymity usually is not possible. You may use pur-
posive sampling to seek out individuals known to have
the desired expertise (see Chapter 15). In other cases,
you may look for participants through the media,
posters, or advertisements. A single contact with an
individual who agrees to attend a focus group does not
ensure that this person will attend the group session.
You will need to make repeated phone calls and
remind the candidates by mail. You may need to offer
compensation for their time and effort in the form of
cash, phone card, gift card, or bus token incentives.
Cash payments are, of course, the most effective if the
resources are available through funding. Other incen-
tives include refreshments at the focus group meeting,
T-shirts, coffee mugs, gift certificates, and coupons.
Over-recruiting may be necessary; a good rule is to
invite two more potential participants than you need
for the group.
Recruiting participants with common social and
cultural experiences creates more homogeneous
groups (Liamputtong & Ezzy, 2005). Selecting par-
ticipants who are similar to one another in lifestyle or
experiences, views, and characteristics is believed to
facilitate open discussion and interaction. These char-
acteristics might be age, gender, social class, income
level, ethnicity, culture, lifestyle, or health status. For
example, for a study of barriers to implementing HIV/
AIDS clinical trials in low-income minority commu-
nities, focus groups might be organized by race/
ethnicity and gender. In heterogeneous groups, com-
munication patterns, roles, relationships, and tradi-
tions might interfere with the interactions within the
focus group. Be cautious about bringing together par-
ticipants with considerable variation in social stand-
ing, education, or authority, because some group
members may hesitate to participate fully, whereas
others may discount the input of those with perceived
lower standing.
Establish a setting for your focus group that is
relaxed. There should be space for participants to
sit comfortably in a circle or U shape and maintain
eye contact with one another. Ensure that the acous-
tics of the room will allow you to obtain a quality
audio-recording of the sessions. As with the one-
on-one interview discussed earlier, place your audio
or video recorders unobtrusively. Use a highly sensi-
tive microphone. Hiring a court reporter to do a
real-time transcription may have advantages over
recording the interaction for transcription later (Scott
et al., 2009). Inaudible voices on the recording or
overlapping voices can pose challenges to later
transcription.
The facilitator is critical to the success of a focus
group. Select a facilitator when possible who reflects
the age, gender, and race/ethnicity of the group. The
researcher may be the facilitator of the group or may
train another person for the role. Training of the facili-
tator should be thorough and allow time for practice
(Gray, 2009). The facilitator, sometime called the
moderator, needs to deeply understand the aims of
the focus groups and to communicate these aims to
the participants before the group session. Instruct par-
ticipants that all points of view are valid and helpful
276 UNIT TWO  The Research Process
and that speakers should not be asked to defend their
positions. Make clear to the group that the facilitator’s
role is to moderate the discussion, not to contribute.
In addition to the moderator, you may want to have an
observer or assistant moderator who takes notes, espe-
cially of facial expressions or interactions not captured
by the recording (Liamputtong & Ezzy, 2005). Making
notes on the dynamics of the group is also useful. Note
how group members interact with one another.
Carefully plan the questions that are to be asked
during the focus group and, if time permits, pilot-test
them. Limit the number of questions to those most
essential to allow adequate time for discussion. You
may elect to give participants some of the questions
before the group meeting to enable them to give
careful thought to their responses. Questions should
be posed in such a way that group members can build
on the responses of others in the group, raise their own
questions, and question one another. Probes can be
used to elicit richer details, by means of questions
such as “How would that make a difference?” or
responses such as “Tell us more about that situation.”
Avoid pushing participants toward taking a stand and
defending it. Once rapport has been established, you
may be able to question or challenge ideas and increase
group interaction.
The researcher and/or facilitator may come to the
focus groups with preconceived ideas about the topic.
Early in the session, provide opportunities for partici-
pants to express their views on the topic of discus-
sion. Use probes or questions if the discussion
wanders too far from the focus of the study. A good
facilitator will weave questions into the discussion
naturally. The facilitator’s role is to clarify, para-
phrase, and reflect back what group members have
said. These discussions tend to express group norms,
and individual voices of dissent may be stifled.
However, when a sensitive topic is being discussed,
the group may actively facilitate the discussion
because less inhibited members break the ice for
those who are more reticent. Participants may also
provide group support for expressing feelings, opin-
ions, or experiences. Late in the session, the facilita-
tor may encourage group members to go beyond the
current discussion or debate and reflect on inconsis-
tencies among the views of participants and within
their own thinking.
Example Study Using Focus Groups
Taylor and Burch (2011) conducted focus groups to
evaluate an enhanced recovery protocol (ERP) that
was developed for patients undergoing colorectal
surgery. The following excerpt gives further details:
“Patients who had entered onto the ERP between 1
September 2008 and 28 February 2009, and lived
locally (i.e., within 1 hour’s drive from the hospital)
were included. Fifty patients who met these criteria
were identified from the ERP database. They were
contacted by post, inviting them to participate and
explaining what the focus groups would involve and
how their participation could help. If willing to par-
ticipate, patients completed a detachable slip at the
bottom of the letter stating their availability for one
of three planned focus group sessions. It was hoped
that 4-6 people would participate in each session.
“A quiet non-clinical room was arranged which had
a circular table allowing participants to maintain eye
contact with each other. Refreshments were offered,
enabling informal interaction before the focus groups
began. The focus groups commenced with an intro-
duction by the facilitators (CT and JB) explaining the
purpose and how the session would run. Written
consent was obtained using a specifically developed
participant information sheet. Initial ground rules
were generated to ensure everyone’s opinions were
respected and heard without interruption, and confi-
dentiality would be maintained.” (Taylor & Burch,
2011, p. 287)
The researchers invited 50 participants, but only 10
patients and 3 relatives participated in the three focus
groups. One group had two patients; the second had
four patients and two relatives; and the third had four
patients and one relative. The low response rate and
the small sample may have limited the data that were
collected. The data reflected those who participated
but may not have accurately reflected other patients in
the enhanced recovery protocol program. The research-
ers could have been clearer about whether the invited
participants were continuing to receive care from the
hospital. If so, some participants may have hesitated
to participate, especially if their perceptions of the
program were less than satisfactory. The data from the
focus groups revealed that environmental details such
as hot meals and available washing facilities for family
members were as important as the clinical aspects of
the care they received.
Electronically Mediated Data
Images created by still and video photography and
Internet communication are newer methods of qualita-
tive data collection that are being used by health-
related researchers. Each is described briefly, and an
 CHAPTER 12  Qualitative Research Methodology 277
example provided. Prior to using one of these forms
of data, the reader is encouraged to study in greater
depth the technology used and the ethical issues due
to potential loss of confidentiality and breach of the
privacy of participants’ protected health information
(see Chapter 9).
Anthropologists and historical researchers have
included photographs as data in their studies for many
years. However, creating photographic images as part
of data collection is emerging as a viable scientific
method in different types of qualitative and quantita-
tive studies the past few years. The ubiquitous nature
of digital photography is likely to speed the accep-
tance of the method. When used as research data,
photographs and videos may be made by participants,
researchers, or a combination of the two. Photovoice
is the idea of “recording and reflecting on the strengths
and concerns” of one’s community and is most often
used in participatory research studies (Findholt,
Michael, & Davis, 2011, p. 186). Caroline Wang and
her colleagues are credited with guiding the first
health-related study during which rural Chinese
women were given cameras to photograph their lives
and especially their health needs (Wang, Burris, &
Ping, 1996). Wang called this practice photo novella,
but others since have used the term photovoice. Nurse
researchers have used photovoice in studies with
rural adolescents to increase awareness of childhood
obesity and the need for prevention (Findholt et al.,
2011) and with siblings in families with children who
have Down syndrome (Rampton et al., 2007). Other
health-related studies conducted by non-nurses have
included studies with older adults who experience
chronic pain (Baker & Wang, 2006) and with African
American men in rural and urban communities to
explore their perceptions of social and cultural factors
that affect individual and community health (Ornelas
et al., 2009).
Photovoice is a creative way to facilitate commu-
nication across cultures. Cooper and Yarborough
(2010) first conducted a focus group with 20 tradi-
tional birth attendants or comodronas in rural Guate-
mala. They defined comodronas as “generally older
women who have become recognized for their knowl-
edge and experience in caring for pregnant women,
delivering babies, and caring for the neonate and the
mother post delivery” (p. 646). After the focus group,
six participants were given disposable 35-mm cameras
and asked to take photographs of positive and negative
aspects of health in their communities. The research-
ers asked the comodronas to obtain written consent
from any persons in the photographs and not to take
any photographs that would put themselves at risk for
harm. These cameras were collected and the photo-
graphs processed by the local research assistant using
research funds. Four months later, the researchers
returned to Guatemala and each of the photographs
was discussed with the photographers in a second
focus group. The researchers noted the differences in
the focus group discussions. “Although the first phase
produced specific and concrete ideas and concerns,
such as specific illnesses and environmental problems,
the second phase elicited, for the most part, a ‘peeling
back’ of health issues” (p. 649). For example, a picture
of a basin of water and a dipper took on new meaning
when the photographer explained that this represented
the entire daily water supply for the family. One of the
strengths of this study was the researchers’ procedures
that protected the rights of the participants and those
photographed. Another strength was the unique com-
bination of data collection methods that produced a
rich picture of health in these rural communities. The
researchers identified limitations of the study to be that
the health ministry nurse was present during the focus
groups and that the photographers were younger
participants.
Photovoice can often generate a deeper understand-
ing of stigmatizing conditions. Fleming, Mahoney,
Carlson, and Engebretson (2009) examined images in
a photovoice exhibit on mental illness. They identified
their data as the “renderings of the artist’s observa-
tions and firsthand narrative accounts of persons living
with mental illness” (p. 18). They selected 15 photo-
graphs with narratives that were available in the
museum exhibit, in the online exhibit, and in the
printed brochure. This choice allowed the researchers
to encounter the data in different formats and strength-
ened the analysis process. The researchers developed
the themes of suffering and stigma. A loss of identity
was an overarching theme linking the other two
themes. The analysis process was rigorous with notes
made on each case that resulted in a three-page inter-
pretive statement for each. The thematic analysis was
validated through repeated meetings of the researchers
to debrief and compare within-case studies with the
original data for logic, coherence, and fittingness”
(p. 19). The researchers provided details about the
analysis process and their conclusions that support
the quality of the study and the credibility of the
findings.
Photovoice may pose unique ethical considerations
because people in photographs can be identified and
may not have consented to participating in the study.
Researchers who are considering photovoice as a
research methodology are urged to read primary
sources and consult with researchers experienced in
278 UNIT TWO  The Research Process
the methodology. The rights of the research partici-
pants must be protected during the conduct and report-
ing of the research.
Internet communication provides a way to collect
data from persons separated by distance. Quantitative
researchers are regularly using World Wide Web–
based surveys and instruments to gather data, but
qualitative researchers are also using Web-based
communities such as online forums and blogs for
research purposes. The number of participants avail-
able for Internet-based research is extensive but does
have the limitation that the sample include only those
who can read and write, are comfortable using a com-
puter, and have access to the Internet (Marshall &
Rossman, 2011). A nurse leader using Internet com-
munication for data collection is Eun-Ok Im. Im has
used mixed-methods studies with quantitative and
qualitative phases. The focus here is on the qualita-
tive phases. Im, Chee, Lim, Liu, and Kim (2008)
used an online forum created for their study to gather
data about physical activity of middle-aged women.
The month-long online forum was completed by 15
of the 30 women who started. The researchers posted
17 topics for discussion with three or four introduced
each week. The topics were about physical activity
and cultural influences on physical activity. The
participants used pseudonyms when posting to the
forum to protect anonymity. The text of the discus-
sions was converted into transcripts for analysis. As
themes were identified, the researchers shared them
with participants and asked for feedback. Im’s
program of research using online forums includes 11
studies. As examples, her team has completed studies
on ethnic differences in cancer pain (Im et al., 2009),
ethnic commonalities and differences in the experi-
ence of menopause (Im, Lee, Chee, Domire, &
Brown, 2010), black women’s experiences with
menopause (Im, Seoung Hee Lee, & Chee, 2010),
Asian Americans’ perspectives on Internet cancer
support groups (Im, Bokim Lee, & Chee, 2010), and
white women’s attitudes toward physical activity (Im,
Lee, Chee, Stuifbergen, and the eMAPA Research
Team, 2011). Im and other colleagues noted in these
studies that one of the limitations was that the data
represent only those who have Internet access and
were comfortable describing personal experiences in
the online forum.
The literature on using Internet communication for
collecting data, or Internet-mediated research, is
growing. Whitehead (2007) produced an integrated
review of the literature on issues of quantitative and
qualitative Internet-mediated research. On the basis of
her review of 46 papers, she concluded that three
major themes affect the credibility of the findings of
Internet-mediated studies. The first is sample bias.
This concern is diminishing as access to the Internet
continues to increase. A researcher could minimize
concerns about sample bias by comparing the demo-
graphic characteristics of the online sample with those
of samples in traditional studies on the same topic.
Whitehead (2007) identified the second concern to be
ethical issues such as seeking consent, assuring ano-
nymity of the participants, and protecting the security
of the site. The third concern was the reliability and
validity of the data collected because the researcher
cannot verify whether participants meet the inclusion
criteria for the study and has no control over distrac-
tions that may occur during data collection. Despite
these issues, studies will continue to be conducted
using the Internet and researchers aware of these
issues can develop studies to minimize the concerns.
Researchers considering this methodology may benefit
from reading the research reports of Im’s teams, which
contain how they addressed issues of confidentiality
and security.
Transcribing Recorded Data
Transcription of verbal data into written data is almost
an assumption in qualitative research. Transcripts
present the data in a form that allows the researcher
to share the data with team members for analysis and
validation. Data collected during a qualitative study
may be narrative descriptions of observations, tran-
scripts from audio recordings of interviews, entries in
the researcher’s diary reflecting on the dynamics of
the setting, or notes taken while reading written docu-
ments. Audio-recorded interviews are generally tran-
scribed verbatim with different punctuation marks
used to indicate laughter, changes in voice tone, or
other nuances. Verbatim transcripts contribute to the
expense and time required for qualitative research and
may not always be needed to answer the research
question (Halcomb & Davidson, 2006). For example,
in qualitative studies used to evaluate a program and
mixed-methods studies, interviews may be used with
structured questions that may not require word-for-
word transcription.
Transcription may require 4 to 6 hours for each 1
hour of interview or focus group time, depending on
the equipment used and the transcriber’s skill. Hiring
a professional transcriptionist may decrease the time
but may be too expensive, depending on the study’s
budget. When hiring a transcriptionist, indicate the
details that you want included and whether you
want the transcript to be verbatim (Davidson, 2009).
 CHAPTER 12  Qualitative Research Methodology 279
Transcribing the recordings yourself has the advan-
tage of immediately immersing you in the data. A
pedal-operated recorder allows you to listen, stop, and
start the recording without removing your hands from
the computer or word processor keyboard. Even when
you hire another person to transcribe the recordings,
you want to check the transcription by listening to the
recording while reviewing the transcript. Voice recog-
nition programs may be of some benefit but require
time for the software to “learn” the voice of the inter-
viewee (Alcock & Iphofen, 2007). For transcription of
focus group recordings, voice recognition software is
not likely to be effective. Other software may allow
conversion of audio recording to digital formats ready
for analysis within computer analysis software. You
also may code the actual recording, negating the need
for a word transcription. Hutchinson (2005) developed
an innovative method of data analysis that uses audio-
editing software to save selected audio bytes from
digital audio recordings. The data are never tran-
scribed but remain in audio form. A database is used
to code and manage the linked audio files and generate
detailed and summary reports. Although the system is
time consuming to set up, it negates the need for
expensive and time-intensive transcription of recorded
data.
Video recordings are maintained in their original
format. However, the researcher may make notes on
sequential segments of the recording, creating a type
of field notes. The researcher may also code the
recordings directly. Video recordings can be used for
quantitative studies, but the recordings are coded and
assigned a numerical value.
Data Management
Because data are collected simultaneously with data
analysis, the study manager, who may be the researcher,
needs to have a plan developed for how to organize
and store data. Label electronic files consistently. For
example, the digital files from recordings can be
labeled with the date and the code number or pseud-
onym of the participant. Make copies of all original
files on a second computer or external storage device.
Similarly, scan or copy all handwritten notes, field
notes, or memos and, if possible, store originals in a
waterproof and fireproof storage box. Any electronic
files containing personally identifiable information
(family member, hospital name, addresses, doctor’s
name) should be encrypted prior to being sent elec-
tronically to a transcriptionist or team member. You
may want to keep a Word document or Excel file
listing all files by date, file name, and type such as
observational memo, transcript, analysis record, or
field note. The study manager may also want to keep
records of who is currently working on that file and
whether it is being transcribed, analyzed, or reviewed
by a team member.
Some researchers may prefer to make notes, mark
text, and label (code) sections of data on a hard copy
of a transcript or field note using colored markers,
pencil, or pen. If hard copies are to be used, ensure
that each page is clearly identified with the file name
in the header or footer of the document. You may want
to format the document with large right-hand margins
to allow more space for coding and notes. It is recom-
mended that you also include line numbers, not for
each page, but for the entire document continuously.
Having line numbers allows the researcher to note the
source of a code by line number within a specific
document.
Other researchers prefer to work on electronic files
within a software program, using tools ranging from
as simple as highlight or comment functions in a docu-
ment within a word processing file to as complex as
analysis of visual images, transcripts, field notes, and
memos within one of several specialized computer
programs. The program does not analyze the data but
records the analysis performed in the mind of the
researcher (Banner & Albarran, 2009; Leech &
Onwuegbuzie, 2011). It can also allow links to be
made between codes, facilitate retrieval of similar
text, or produce diagrams of relationships among
codes.
Computer-assisted qualitative data analysis soft-
ware (CAQDAS) can maintain a file directory, allow
for annotation of coding decisions, and retrieve sec-
tions of text that the researcher has identified with
the same code (Banner & Albarran, 2009; Garcia-
Horta & Guerra-Ramos, 2009; Hoover & Koerber,
2011). Hoover and Koerber (2011) summarize the
advantages of CAQDAS to be efficiency, transpar-
ency, and multiplicity. Table 12-3 provides a more
detailed list of the advantages and disadvantages of
using CAQDAS, and Table 12-4 contains descrip-
tions and suppliers of a selected group of CAQDAS
programs.
Data Analysis
Qualitative data analysis is “a process of examining
and interpreting data in order to elicit meaning, gain
understanding, and develop empirical knowledge”
(Corbin & Strauss, 2008, p. 1). Qualitative data analy-
sis is creative, challenging, time consuming, and, con-
sequently, expensive (Jirwe, 2011). Less experienced
280 UNIT TWO  The Research Process

TABLE 12-3  Advantages and Disadvantages TABLE 12-4  Examples of Computer-Assisted

of Computer-Assisted Qualitative Qualitative Data Analysis Software
Data Analysis Software (CAQDAS) (CAQDAS)
Advantages Store and organize data files Software Description and Website
Provide documentation of coding and ATLAS/ti 6.2 Robust CAQDAS functions; large
analysis searchable data storage including
Click and drag to merge codes media files; multiple users
Search for related codes and quotations allowed; facilitates theory building;
efficiently flexible; supports use of PDF files.
Send coded data files to others http://www.atlasti.com/
Link memos to text Ethnograph v6 Originally developed for use by
Generate a list of all codes ethnographers; import and code
Minimize clerical tasks to allow focus data files; sort and sift codes;
on actual analysis retrieve data and files.
Support and integrate the work of http://www.qualisresearch.com/
multiple team members FrameWork Textual analysis tools plus
Decrease paper usage summarization tool and ability to
Disadvantages Cost of software analyze using a matrix format.
Need to allow time and expend energy http://www.framework-
to learn the software and its functions natcen.co.uk/
Potential that technical/functional HyperRESEARCH Code and retrieval functions; theory
aspects will overwhelm thinking building features added on;
about the analysis handles media files; compatible
Potential for computer problems with Macintosh computers.
interfering with the software and http://www.researchware.com/
causing data and analysis to be lost MAXQDA 10 Robust CAQDAS functions, but less
powerful search tool; allows
integration of quantitative and
qualitative analysis; color-based
researchers may feel uncertain about how to proceed filing; supports different types of
text analysis.
because the process feels ambiguous (Speziale &
http://www.maxqda.com/products
Carpenter, 2007). Data analysis in qualitative research NVivo 9 Robust CAQDAS functions including
occurs concurrently with data collection. Analysis of (combined with several types of queries; familiar
an interview’s data may result in the asking of an NUD*IST) format of file organization system;
additional question in subsequent interviews to handles multimedia files; latest
confirm or not confirm an initial interpretation of the version includes compatibility with
data. The process of interpretation occurs in the mind quantitative analysis and
of the reader. Corbin and Strauss (2008) describe bibliographic software.
interpretation as translating the words and actions of http://www.qsrinternational
participants into meanings that readers and consumers .com/#tab_you/
Weft QDA Free software that provides most
can understand. The virtual text grows in size and
commonly used CAQDAS
complexity as the researcher reads and rereads the functions.
transcripts. Throughout the process of analysis, the http://www.pressure.to/qda/
virtual text develops and evolves. Although multiple
valid interpretations may occur if different researchers Synthesized from Davis and Meyer (2009), Hoover and Koerber (2011),
examine the text, all findings must remain trustworthy Speziale and Carpenter (2007), and websites of suppliers and professional
organizations.
to the data. This trustworthiness applies to the unspo-
ken meanings emerging from the totality of the data,
not just the written words of the text. The first step in and experiences, listening to audio recordings and
data analysis is to be familiar with the data. viewing videotapes until you have become immersed
in the data. Being immersed means that you are fully
Immersion in the Data invested in the data and are spending extensive
Becoming familiar with the data involves reading and amounts of time reading and thinking about the data.
rereading notes and transcripts, recalling observations Recordings contain more than words; they contain
 CHAPTER 12  Qualitative Research Methodology 281
feeling, emphasis, and nonverbal communications.
These aspects are at least as important to the com-
munication as the words are. As you listen to record-
ings, look at photographs, or read transcripts, you
relive the experiences described and become very
familiar with the phrases that different participants
used or the images that were especially poignant. In
phenomenology, this immersion in the data has been
referred to as dwelling with the data (Munhall, 2012).
Byers and France (2008) designed a phenomenologi-
cal study using van Manen’s (1984) approach to phe-
nomenology. They examined the experiences of
registered nurses who cared for dementia patients in
acute care settings. Byers and France described dwell-
ing with the data in the following excerpt:
“Staying true to phenomenology as philosophy and
method, data collection and analysis occurred simul-
taneously and the researchers maintained a stance of
sensitive attunement and openness to the possibilities
of meaning to reach self-reflection, bracketing, and
phenomenological reduction. Following each inter-
view, the researchers transcribed the interview and
dwelt with the data, continually writing and rewriting,
and reading and re-reading the transcript, eventually
formulating judgments about the recurring patterns
and emerging themes of the parts within the whole
staying true to van Manen’s phenomenology.” (Byers
& France, 2008, p. 45)
Coding
Because of the volumes of data acquired in a qualita-
tive study, initial efforts at analysis focus on reducing
the volume of data so that the researcher can more
effectively examine them. The reduction of the data
occurs as you attach meaning to elements in your data
and document that meaning with a word, symbol, or
phrase known as a code. Coding is a means of naming
and labeling. A code is a symbol or abbreviation used
to label words or phrases in the data. Through coding,
the researcher explores the phenomenon of the study.
Therefore, it is important that the codes be consistent
with the philosophical base of the study. Organization
of data, selection of specific elements of the data for
categories, and naming of these categories all reflect
the philosophical basis of the study. Later in the study,
coding may progress to the development of a taxon-
omy or a theoretical framework. For example, you
might develop a taxonomy of types of pain, types of
patients, or types of patient education. Initial catego-
ries should be as broad as possible with minimal
overlap. As data analysis proceeds, the codes may be
merged and relabeled at a higher level of abstraction.
In a study of medication adherence, the initial codes
might be “paying attention to time,” “counting and
recounting,” and “remembering to get prescription.”
These codes might be grouped later into the more
abstract code “attending to logistics.” The first level
of coding is descriptive and uses participant phrases
as the label for the code. The label for the merged
codes is interpretive and might be called a theme if
repeatedly identified in the data.
Trimm and Sanford (2010) studied the experiences
of people waiting while a family member undergoes
surgery. Their grounded theory study resulted in a
middle-range theory that they called “maintaining
balance during the wait” (p. 441). Their analysis
yielded 56 open codes that they transformed into six
axial codes. Further analysis resulted in four domains
that comprise the theory: “focusing on the patient,
passing the time, interplay of thoughts and feelings,
and giving and/or receiving support” (p. 440). Table
12-5 contains excerpts from quotations of participants
with their corresponding codes for one of the substan-
tive domains, focusing on the patient. Trimm and
Sanford (2010) provided extensive detail about their
analysis and validated their findings with some of the
participants. These actions indicate that the study was
rigorously implemented and lend credibility to the
findings.
The type and level of coding vary somewhat
according to the qualitative approach being used.
Table 12-6 displays types of codes described in the
social science literature. The terms can be confusing,
because different writers have given different names
to similar codes.
Content Analysis
Content analysis is designed to classify the words in
a text into categories. The researcher is looking for
repeated ideas or patterns of thought. In exploratory-
descriptive qualitative studies, researchers may ana­
lyze the content of the text using concepts from a
guiding theory, if one was selected during planning for
the study. During historical studies, the researcher
analyzes documents and photographs to describe their
content related to the focus of the study.
Garwick, Seppelt, and Riesgraf (2010) used a par-
ticipatory qualitative study design to explore the chal-
lenges of managing asthma of children in urban Head
Start programs at multiple sites. They conducted focus
groups with teachers and managers of Head Start
Centers and transcribed the data, as described in the
following excerpt:
282 UNIT TWO  The Research Process

TABLE 12-5  Example of Participants’ Quotations and Codes for the Substantive Domain of Focusing
on the Patient
Participants’ Quotations Code
From the wife of a patient having back surgery on why they spent the night before surgery in a hotel: Driving/admitting
“Because we live a long way . . . we had kind of a rough night … because I know he’s nervous.” (p. 443) the patient
“I just wanted to go back there [holding area] and reassure him that I loved him and boost him up a little Reassure/comfort
bit. I definitely wanted to see him before surgery, ’cause it could have been the last time.” (p. 443) the patient
“I was better when I got here last night, just to see dad and that he was really okay … just wondering if he Getting
was okay. Like if he really [italics added] was okay, if he really could talk well and didn’t have any information
physical disabilities or so and he was fine.” (p. 443) first hand
“I love you. I’ll see you in a few minutes.” (p. 444) ‘See you soon’
“I never say goodbye, because it feels so final.” (p. 444)
“We’ll be right here in the waiting room.” (p. 444)
Description of family behavior, not a participant quotation: “When the family entered the waiting area, they Entering the
would look at the room’s arrangement and select a place to sit. Once seated, the conversations and waiting room
actions focused on how the patient looked, when they would hear information, and what activity should
be done first.” (p. 445)
“I was thinking about [spouse’s name] and I know exactly what they are doing in that surgery, because he’s Thinking about
had it done before. So I pictured the saw. I pictured the hammer. I mean I have a wild imagination, so I the surgical
was just crying and thinking of all those things. I can picture the surgery happening and the position that procedure
he’s in and it just made me feel so emotional and that’s why I was crying.” (p. 445)
“This time the surgery will either result in her being able to continue her life or she’s going to have to Worried about
remain an invalid … we’re banking on everything on this move. There’s the potential that she would be the outcome
permanently paralyzed from below the waist. Everything hinges on this operation, because none of her
plans can go forwards unless this works.” (p. 446)

From Trimm, D. R., & Sanford, J. T. (2010). The process of family waiting during surgery. Journal of Family Nursing, 16(4), 435–461.

challenging interactions with parents of children with

“Descriptive content analytic techniques (Neuendorf,
2002) were used to identify and categorize asthma
management issues and action plan strategies. First,
the [principal investigator] PI and research associate
(RA) read and reread each focus group transcript in
its entirety to understand the context of the focus
group discussions. Next, each independently read the
responses in the verbatim transcripts on a line-by-line
basis to identify types of asthma management issues
within and across the focus groups. Then they met to
confirm, through a consensus process, common
themes that participants emphasized in all three of
the teacher focus groups. After the RA systematically
organized the findings by theme and ID number, the
PI validated the categorization of findings. Exemplars
of each of the common themes or challenges are
included in the findings.” (Garwick et al., 2010,
p. 331)
The credibility of the findings was increased
through checking the accuracy of the transcription,
assessing consistency between two data coders, and
validating the findings with participants. The Head
Start teachers and managers noted helpful and
asthma and the complexity of the treatment plans that
must be implemented with no additional staff and
limited equipment.
Narrative Analysis
Narrative inquiry is a qualitative approach that uses
stories as its data (Duffy, 2012). Through a series of
life experiences, people create their identities in the
historical and social context in which they live. As a
philosophical approach to qualitative research, narra-
tive inquiry was not included in this textbook (see
Duffy, 2012, for additional information on the method).
Data analysis, however, may yield stories, and
researchers using other philosophical approaches may
tell a participant’s story in their analysis and presenta-
tion of findings. In addition to being organized chron-
ologically, you might analyze a story as one would a
published book during a literature course, looking for
characters, setting, plot, conflict, and resolution. His-
torical researchers may compare participants’ stories
to present a broader picture of an event. Lesniak
(2010), in her phenomenological study of adolescents
who injure themselves (mentioned earlier in this
chapter), did not indicate that she used narrative analy-
sis. She presented, however, rich descriptions of the
 CHAPTER 12  Qualitative Research Methodology 283

TABLE 12-6  Types of Coding for Qualitative Data Analysis*

Type Description
Axial coding Finding and labeling connections between concepts; assigning codes to categories; also may be called Level II
coding in grounded theory studies
Descriptive coding Classifying elements of data using terms that are close to the participant’s words, usually early in the analysis
Explanatory coding Connecting coded data to an emerging theory; describing coded data as patterns, themes, and links
Interpretive coding Labeling coded data into more abstract terms that represent merged codes
In-vivo coding “Concepts using actual words of research participants rather than being named by analyst” (Corbin & Strauss,
2008, p. 65)
Open coding “Breaking data apart and delineating concepts to stand for blocks of raw data” (Corbin & Strauss, 2008, p.
195); also called Level I coding in grounded theory studies
Process coding Incorporating previously assigned codes into an overall basic social process that is the core of the
phenomenon of interest; used in grounded theory studies (Speziale & Carpenter, 2007)
Selecting coding “Building a ‘story’ that connects the categories” (Creswell, 2007). The researcher may generate propositions
or statements that bridge the categories.
Substantive coding Includes in-vivo coding (using words of participants) and implicit coding, in which the unspoken codes are
constructed by researchers

*These terms are not mutually exclusive, because different writers have used different labels for similar analytical processes.

lived experiences of the adolescents as stories. She
then extracted themes from the stories. This is an
excerpt of one of the stories:
“Emma had just turned 17 years old when she agreed
to be interviewed.… Emma cut as a means to gain a
sense of control over her life. She could not control
her parents’ divorce or the loneliness she experi-
enced during her childhood. She had no friends and
her parents began dating other people, which further
isolated Emma. Cutting was there for her and did not
abandon her. It was a dependable variable in her life.
Emma felt she did not fit in with other girls at school,
becoming a target of their jokes and cruelty. After the
first time she cut, Emma felt immediate relief from
her pent-up emotions, and she was able to sleep. She
stated that before she cut, she would feel numb inside
and after she cut, the emotional pain became tangible
through the physical pain of cutting. Cutting made
Emma not want to die; instead, she cut to feel alive.
The wound became a tangible and visible external
representation of the pain she felt internally. It also
gave her a location upon which to focus as the scars
became a reminder of everything she had gone
through; it was the story of her life. She stated that
cutting was her expression, but on her skin. It became,
in other words, her voice.” (Lesniak, 2010, p. 143)
Lesniak (2010) gave the readers insight into the lives
of the participants by presenting the descriptions as
stories. When one goal of a study is to influence policy
makers or promote behavior change, stories can be a
powerful tool to generate support for the advocated
message. Table 12-7 provides definitions of additional
types of qualitative data analysis.
Memoing
The researcher develops a memo to record insights or
ideas related to notes, transcripts, or codes. Memos
move the researcher toward theorizing and are con-
ceptual rather than factual. They may link pieces of
data together or use a specific piece of data as an
example of a conceptual idea. The memo may be
written to someone else involved in the study or may
be just a note to yourself. The important thing is to
value your ideas and document them quickly. Initially
you might feel that the idea is so clear in your mind
that you can write or record it later. However, you may
soon forget the thought and be unable to retrieve it.
As you become immersed in the data, these ideas will
occur at odd times, such as when you are sleeping,
walking, or driving (Schiellerup, 2008). Whenever an
idea emerges, even if it is vague and not well thought
out, develop the habit of writing it down immediately
or recording it on a hand-held device such as a cell
phone.

Findings and Conclusions

Stories may produce an emotional link between the Qualitative findings reflect the study’s philosophical
research consumer and the participant beyond that roots and the data that were collected. Unlike in quan-
achieved through discussion of codes and themes. titative research, conclusions are formed throughout
284 UNIT TWO  The Research Process

TABLE 12-7  Types of Qualitative Data Analysis

Data Analysis
Chronological analysis
Componential analysis
Constant comparison
Direct interpretation
Domain analysis
Narrative analysis
Taxonomic analysis
Thematic analysis
Theoretical comparison
Three-dimensional analysis
Description
Identifying and organizing major elements in a time-ordered description as events and epiphanies
Identifying units of meaning that are cultural attributes; process allows ethnographer to identify gaps
in observations and selectively collect additional data
Analyzing new data for similarities to and differences from existing data
Identifying a single instance of the phenomenon or topic and drawing out its meaning without
comparing to other instances
Focusing on specific aspects of a social situation such as people involved; used in ethnography
Looking for the story in the data; identifying the characters, setting, plot, conflict, and resolution as an
exemplar of the phenomenon being studied
Identifying categories with a domain (see domain analysis); used in ethnography
Finding within the data three to six overriding abstract ideas that summarize the phenomenon of
interest
Thinking about the properties and characteristics of categories; linking to existing theories and models
Thinking about and identifying continuity, interactions, and situations within a story

Synthesized from Corbin and Strauss (2008) and Creswell (2007).

the data analysis process in qualitative research. Con-
clusions are intertwined with the findings in a qualita-
tive study. For a phenomenological study, the findings
are presented as an exhaustive description (Speziale
& Carpenter, 2007). The findings of a focused ethno-
graphic study would include a description of the
culture that achieved the study objectives or answered
the research questions. In grounded theory studies, the
researcher’s aim is to produce a text or graphic descrip-
tion of social processes. As the description is refined,
a theoretical structure or framework emerges that
might be considered a tentative theory (Fawcett &
Garity, 2009; Marshall & Rossman, 2011; Munhall,
2012).
Reporting Results
In any qualitative study, the first section of a research
report should be a detailed description of the partici-
pants. The ethnography report will also include details
about the setting and the environment in which the
data were gathered. The results of the data analysis
may be displayed in the form of a table with the
themes in the first column of a table and exemplar
quotations in the second column. Using tables in this
way increases the transparency of the analysis and
interpretation. Other writers incorporate supporting or
disconfirming evidence from the literature within the
results section of the report. How the results are pre-
sented depends on the philosophical approach upon
which the study was developed. As previously men-
tioned, phenomenologists provide a thick, rich, and
exhaustive description of the phenomenon that was
studied. Ground theorists present their tentative
theory. Exploratory-descriptive qualitative studies are
reported by addressing each research question and
providing the pertinent findings. The report of a his-
torical study may have limited information about the
methods; rather, the report is the story of the events or
series of events that were studied.
Methods Specific to
Qualitative Approaches
Phenomenological Research Methods
Phenomenological researchers have several choices
about methods that are related to their specific philo-
sophical views on phenomenology. In Chapter 4, dif-
ferences in Husserl’s and Heidegger’s views on
phenomenology were discussed. Researchers sub-
scribing to Husserl’s views would use bracketing,
which is consciously identifying, documenting, and
choosing to set aside one’s own views on the phenom-
enon (Dowling, 2007). Heidegger’s view was that
researchers could not separate their own perspectives
from that of the participants’ during the collection,
analysis, and interpretation of the data. In phenome-
nology, additional philosophical approaches to the
analysis and interpretation of data are available, such
as those advocated by van Kaam (1966), Giorgi
(1970), Colaizzi (1978), and van Manen (1984).
Munhall (2012) calls these men “second-generation
phenomenologists” (p. 126). Prior to selecting an
approach, you are encouraged to read the primary
sources listed in the references. Shorter and Stayt
 CHAPTER 12  Qualitative Research Methodology 285
(2010) conducted a phenomenology study according
to Heidegger’s philosophy. They emphasized the
importance of co-creating the data, as follows:
“A key tenet of Heideggerian phenomenology is
co-construction of knowledge between researcher
and participant, which assumes that both contribute
to understanding the topic. Adequate participant con-
tribution to the construction of knowledge was
ensured in the present study by providing each par-
ticipant with an annotated version of their interview
transcription, detailing subject themes that had been
identified. They were offered the opportunity to
clarify meaning and comment on identified themes.”
(Shorter & Stayt, 2010, p. 161)
Shorter and Stayt (2010) concluded with the fol-
lowing paragraph:
“Confronting death and dying is unavoidable in critical
care settings. End-of-life care is therefore an impor-
tant aspect of critical care nursing. This study has
revealed a complex web of predisposing factors and
occurrences that can shape both the nature of care
for the dying and critical care nurses’ subsequent grief
experiences.” (Shorter & Stayt, 2010, p. 165)
This study was congruent with the Heideggerian
philosophy as evidenced by the validation of the anal-
ysis with the participants. From their findings, the
researchers indicated several areas needing additional
study, such as the informal support structures that
allow critical care nurses to deal with patient deaths.
The inferred clinical implications are that nurses
involved in end-of-life care in acute care settings
experience patients’ deaths in complex ways and use
multiple ways to deal with the grief. Nurses and man-
agers in critical care units need to be aware of the
diversity of grief responses and coping methods.
Grounded Theory Methodology
Philosophical discussions of grounded theory method-
ology center on the nuances of the different approaches
(Cooney, 2010). Sociologists Glaser and Strauss
(1967) worked together during their early years, but
eventually their philosophies resulted in at least two
variations of grounded theory. The original works pro-
vided little detail on data analysis methods, so Corbin
and Strauss (2008) described a structured method of
data analysis (Cooney, 2010). In Table 12-5, substan-
tive and theoretical codes are attributed to Glaser, and
open, axial, and selective are attributed to Strauss
(Cooney, 2010). Researchers considering grounded
theory methodology will want to read the primary
sources on the different methods and choose the one
that is most compatible with the researchers’ beliefs.
During grounded theory studies, data analysis for-
mally begins with the first interview or focus group.
The researchers review the transcript and code each
line, constantly comparing the meaning of one line
with the meanings in the lines that preceded it. Con-
cepts as abstract representations of processes or enti-
ties are named. As the data analysis continues,
relationships between concepts are hypothesized and
then tested for validity (Wuest, 2012). Researchers are
looking for a core category that explains the underly-
ing social process in the experience. Existing theory
and literature are reviewed for contributions to the
researcher’s understanding of the core category.
Banner (2010a) conducted a grounded theory study
of women (N = 30) undergoing coronary-artery bypass
graft (CABG) surgery. She explicitly described the
methodological principles guiding the study, as
follows:
“Key principles of the grounded theory methodology
are the following: (1) recursive data collection and
analysis to the explore merging data, (2) construction
of codes and categories, (3) theoretical sampling
techniques to achieve conceptual density, (4) constant
comparison, (5) generation of substantive theory as
opposed to pure analytical description, (6) use of
theoretical memos to track emerging data and theory
and (7) deferred literature review until after analysis
is complete to encourage analytical sensitivity
(Charmaz 2006).” (Banner, 2010a, p. 3125)
Using Strauss and Corbin’s (1998) approach to data
analysis, Banner (2010a) found six categories in the
pre-operative experiences of women scheduled for
CABG surgery: “help seeking, diagnosis and referral,
conceptualizing surgery, living with coronary heart
disease, and waiting for surgery” (p. 3123). Banner
reported quotations that exemplified each category and
then compared the categories with the findings of other
studies. The underlying social process was described
as “the public-private dialogue around maintaining
and renegotiating normality” (Banner, 2010a, p.
3123). Banner did not develop a graphic representa-
tion of the relationships among the categories and with
286 UNIT TWO  The Research Process
the overall social process, but provided a description
of the theoretical structure. The theoretical structure
could be further refined by Banner or other investiga-
tors on this topic. The clinical implications are that
nurses who read the study report may be sensitized
to the significant life disruption and challenges faced
by women through CABG surgery into recovery.
Because several of the participants did not see them-
selves as being at risk for coronary artery disease,
they may have delayed seeking care. Nurses must
teach women how the symptoms of coronary artery
disease may be different in women and that the women
need to seek care when they experience symptoms.
Ethnographical Methodology
Ethnography is unique among the qualitative
approaches because of its cultural focus. Thus, eth-
nography requires field work, which is spending time
in the selected culture to learn by being present,
observing, and asking questions. Wolf (2012) defines
field work as a “disciplined mode of inquiry that
engages the ethnographer firsthand in data collection
over extended periods of time” (p. 302). Field work
allows the researcher to participate in a wide range of
activities. The observations of the researcher typically
focus on objects, communication patterns, and behav-
iors to understand how values are socially constructed
and transmitted (Wolf, 2012). The researcher looks
below the surface to identify the shared meaning and
values expressed through everyday actions, language,
and rituals (Creswell, 2007). Meanings and values
may reveal power differences, gender issues, opti-
mism, or views of diversity.
One difficulty in planning an ethnographic study is
knowing how much time is needed and actually
what will be observed. Enough time in the field is
needed to achieve some degree of cultural immersion
(Speziale & Carpenter, 2007). The length of an ethno-
graphic study is usually limited by the resources—
money and time—that the researcher has allotted for
the project. When one is studying a different culture,
the time might extend to months or even a year. When
studying the culture of a nursing unit or waiting area,
the researcher will not live on the unit, but would
identify a tentative plan for observing on the unit at
different times during the day and night and on differ-
ent days of the week. The researcher may want to
observe unit meetings, change-of-shift reports, or
other unit rituals, such as holiday meals. Initial accep-
tance into a culture may lead to resistance later if the
researcher’s presence extends beyond the communi-
ty’s expectations or the ethnographer is perceived as
prying or violating the community’s privacy. The
researcher needs to blend into the culture but remain
in an outsider role. A researcher who over-identifies
with the culture being studied and becomes an insider
is said to be going native. In going native, the
researcher becomes a part of the culture and loses all
objectivity—and with it the ability to observe clearly.
Negotiating relationships and roles is a critical skill
for ethnographers, who must possess self-awareness
and social acumen.
Gatekeepers and Informants
Gatekeepers are people who can provide access to the
culture, facilitate the collection of data, and increase
the legitimacy of the researcher (Wolf, 2012). A gate-
keeper may be a formal leader, such as a mayor,
village leader, or nurse manager, or an informal leader,
such as the head of the women’s club, the village
midwife, or the nurse who is considered the unit’s
clinical expert. The support of people who are accepted
in the culture is key to gaining the access needed to
understand the culture. In addition to gatekeepers, you
may seek out other individuals who are willing to
interpret the culture for you. These other individuals
may be informants, insiders in the community who
can provide their perspective on what the researcher
has observed (Wolf, 2012). Not only will the infor-
mants answer questions, but they may also help you
to formulate the questions because they understand the
culture better than you do.
Gathering and Analyzing Data
During field work, the researcher will make extensive
notes about what is observed and thoughts on possible
interpretations. The researcher may seek input on the
possible interpretations with an informant or a person
being interviewed. Data analysis is coding field notes
and interviews for common ideas to allow patterns to
emerge. Data may also be subjected to content analy-
sis. The notes themselves may be superficial. However,
during the process of analysis, you will clarify, extend,
and interpret those notes. Abstract thought processes
such as intuition and reasoning are involved in analy-
sis. The data are then formed into categories and rela-
tionships developed between categories. From these
categories and relationships, the ethnographer
describes patterns of behavior and supports the pat-
terns with specific examples.
The analysis process in ethnography provides
detailed descriptions of cultures. The descriptions
may be presented as cultural themes or a cultural
inventory (Speziale & Carpenter, 2007). These
descriptions may be applied to existing theories of
cultures. In some cases, the findings may lead to the
 CHAPTER 12  Qualitative Research Methodology 287
development of hypotheses, theories, or both. The
results may be useful to nurses when members of the
community that was studied interact with the health
system. The results may be tested by whether another
ethnographer, using the findings of the first ethno-
graphical study, can accurately anticipate human
behavior in the studied culture.
McGibbon, Peter, and Gallop (2010) conducted an
institutional ethnography of nurses’ stress, a specific
type of ethnography, as the following excerpt explains:
“Institutional ethnography is a framework that
describes how people’s activities are socially orga-
nized in a particular way as they go about the routine
activities of their daily lives.… Institutional ethnogra-
phy provides a specific method to link nurses’ every-
day work with institutional structures that shape
practice. Although ethnographic practice in general
provides a framework for rich description of the con-
texts of everyday life, the aim of institutional ethnog-
raphy is not thick description; rather, the aim is to
expose the articulation of the activities of everyday
life with institutional power relations.” (McGibbon et
al., 2010, p. 1356)
The researchers collected and analyzed data from
numerous sources, as described here:
“Data collection included tape-recorded interviews at
a time and place that was convenient for the nurses,
participant observation in the pediatric ICU (PICU)
for a 3-month period, focus groups with participating
nurses, field notes that incorporated a researcher
journal, and the examination of selected nonconfiden-
tial texts related to the nurses’ everyday work.…
The focus of data analysis was explicating the forms
of stress in nurses’ everyday PICU work and their
social organization. Data analysis initially consisted of
coding for the forms of stress in nurses’ work, par-
ticularly given the absence of much of nurses’ experi-
ences in current formulations of nurses’ stress.…
The purpose of attending to thematic codes in the
data was not necessarily to arrive at any unifying
description of the nurses’ experiences across indi-
vidual accounts; rather, thematic coding helped to
identify aspects of the nurses’ experiences related to
the research questions and the social relations
inherent in their work.” (McGibbon et al., 2010,
pp. 1356, 1357)
McGibbon et al. (2010) found six forms of nurses’
stress, “including emotional distress; constancy of
presence; burden of responsibility; negotiating hierar-
chical power; engaging in bodily caring; and being
mothers, daughters, aunts, and sisters” (p. 1357).
“Prolongation of life and active caring for those
declared dead were socially organized and textually
mediated processes that obscured nurses’ emotional
suffering through their scientific and rational charac-
ter.… Nurses were uniquely situated temporally and
spatially in the PICU. Their presence for 12 hours for
almost any given shift and for consecutive shifts meant
that their temporal connection with patients was
markedly different than that of any of the other clini-
cians.” (McGibbon et al., 2010, p. 1359)
The study was rigorous, as evidenced by the use of
multiple sources of data and thorough analysis by the
team. The researchers conclude that nurses’ stress was
increased and perpetrated by the organizational struc-
tures that de-emphasized the centrality of nurses to the
work of hospitals. The study’s findings substantiated
the complexity of nurses’ stress in the workplace and
the need for nurses to develop effective stress manage-
ment strategies.
Exploratory-Descriptive Qualitative
Methodology
Researchers often design exploratory-descriptive qual-
itative studies to address a specific research question.
For example, Mahoney and Ladd (2010) conducted
focus groups at a national meeting of gerontological
nurse practitioners (GNPs) to fill a specific gap in the
literature, the lack of knowledge about the influence
of pharmaceutical marketing on GNP prescribing prac-
tices. The 15 GNPs who participated were not asked
to describe all aspects of their practice: “This study
was undertaken to explore NP participants’ prescrip-
tive decision-making issues related to older adult
patients with specific inquiry about the influence of
pharmaceutical marketing” (Mahoney & Ladd, p. 18).
Kaddoura (2010) designed a study with a slightly
broader focus, as described in this excerpt:
“This study explored new graduate nurses’ percep-
tions of factors that helped to develop their critical
thinking skills throughout their critical care orienta-
tion program. The study attempted to answer the
following research question: How do new graduate
288 UNIT TWO  The Research Process
nurses characterize the role of the ECCO [Essential
of Critical Care Orientation] program in influencing
the critical thinking of new critical care nurses edu-
cated by this program?” (Kaddoura, 2010, p. 426)
Data analysis for exploratory-descriptive qualita-
tive studies is often content analysis, with or without
a guiding theoretical framework. Kaddoura (2010) did
not identify a guiding theory and described her data
analysis in this way.
“Data were analyzed by the qualitative content analy-
sis approach to identify key points that described
graduate nurses’ perceptions of how their critical
thinking skills were developed throughout the critical
care orientation program. Four key codes or themes
emerged as the participants verbalized their percep-
tions and experiences with the ECCO program. The
major themes were related to knowledge and its
application in the development of new graduates’
critical thinking skills as well as the graduates’
perceptions of the merits and concerns related to the
ECCO program presentation. These primary themes
included gathering knowledge, application of knowl-
edge, benefits of the ECCO program, and concerns
about the ECCO program in relation to the new
graduate nurses’ critical thinking skills.” (Kaddoura,
2010, p. 426)
Kaddoura’s (2010) study provided helpful informa-
tion to use in evaluating the orientation program.
Nurse educators designing orientation programs for
new graduates may find the study’s results applicable
to their work. Other than this group, the study has
limited applicability. Kaddoura does provide an
exploratory design that would be helpful in evaluating
programs, especially when combined with quantita-
tive data in a mixed design study.
Exploratory-descriptive qualitative studies may be
designed using a theoretical framework. In their quali-
tative study with low-income young fathers (N =17),
Devault et al. (2008) examined the participants’ life
stories in the context of the Belsky (1984) model of
parental behavior. From Belsky’s model, Devault et
al. (2008) developed a conceptual framework in which
the concepts of individual history, coparental history,
and professional history influenced fathering. Father-
ing influenced meaning and involvement in parenting.
Although the researchers indicated that they were
interested in the life stories, they did not use narrative
data analysis. They selected interview questions that
reflected the different components of Belsky’s model
and then used the components as a structure for content
analysis of the transcripts, as described in the follow-
ing excerpt:
“The life story method is used to understand the
subjective experience of individuals within a given
group (here, vulnerable fathers). Being told by the
person who experienced it, the life story’s purpose
is not only to get factual information but also to
understand the meaning given to events. This enables
the researcher to compare different life stories and
discover common themes that emerge within the
identified group. Thus, data collection was essentially
intended to encourage fathers to tell us about their
lives. The participants’ life stories were examined
from a variety of angles that correspond with those
identified in our conceptual framework.” (Devault,
2008, p. 231)
The researchers presented their conclusions in the
context of the conceptual framework, as follows:
“Returning to our conceptual framework, our study
suggests that the mother of origin plays an important
role in the individual history of vulnerable fathers.…
In regards to the co-parental relationship, we found
that a positive co-parental relationship is manifested
by frequent contacts, shared activities, and mutual
support.… The influence of work on fatherhood was
less clear. In fact, we found that it is the other way
around: becoming a father seems to have an influence
on work. Most fathers try to overcome their profes-
sional instability in order to meet the family needs.
Finally, we found that fatherhood had a very impor-
tant meaning to vulnerable fathers and that it pushes
them to better themselves and become more respon-
sible. Involvement can take different forms but we
found that the most involved fathers were active in
every aspect of their child’s life; they formed a stron-
ger commitment to the father role and status, and
were more centered on their children than on them-
selves.” (Devault et al., 2008, p. 242)
Devault et al. (2008) are not nurses but they con-
ducted a study of interest to nurses working with fami-
lies, especially child-bearing families. The researchers
 CHAPTER 12  Qualitative Research Methodology 289
clearly described their analysis process and recog-
nized the limitations of their study. Their methods lend
credibility to their results.
Historical Research Methodology
The methodology of historical research consists of
(1) identifying a question or study topic; (2) identify-
ing, inventorying, and evaluating sources; and (3)
writing the historical narrative. Whether motivated
by curiosity, personal factors, or professional reasons,
the researcher’s interest in a specific topic needs to
be explainable to others (Lundy, 2012). One way to
do this is for the researcher to develop a clear,
concise statement of the topic. The topic may be nar-
rowed to be manageable with available resources.
Although the historical researcher may be interested
in the effect of World War II on nursing science, the
researcher may need to narrow the study to one or a
few nurse theorists who were nurses during the war
or the nurse scientists educated at one university. The
statement of the topic may evolve into a title for the
study, which includes the period being addressed.
Prior to determining the years to be studied, you
must have knowledge of the broader social, political,
and economic factors that would have an impact on
the topic. Using this knowledge, you can identify the
questions you will examine during the research
process.
Sources
Sources in a historical study may be documents such
as books, letters, newspaper clippings, professional
journals, and diaries. Sources may also be people who
were alive during the time being studied or who heard
stories from older relatives. Review the literature that
is available on the topic you have selected, and start a
bibliography or inventory of materials you want to
review. Library searches identify published materials
and may have some archives, unpublished materials
purchased or donated for their historical value (Spe-
ziale & Carpenter, 2007). Pay attention to the organi-
zations and institutions with which the person was
affiliated. These organizations and affiliations provide
clues to the location of primary sources (Lundy,
2012). Primary sources are “firsthand accounts of
the person’s experience, an institution, or an event and
may lack critical analysis” (Speziale & Carpenter,
2007, p. 263). For example, historical researchers
interested in Martha Rogers and the effect of World
War II on her theory would note that Rogers was the
Dean of New York University, increasing the likeli-
hood that the university has documents written by her.
In the case of Rogers, however, an Internet search
reveals that many of her materials are housed
in Boston University’s Howard Gotlieb Archival
Research Center. Accessing these documents would
include obtaining permission to review the docu-
ments, traveling to Boston, and making notes about or
images of the documents.
Secondary sources are those written about the
time or the people involved, but not by the person of
interest. Secondary sources are also examined because
they may validate or corroborate primary sources or
present additional information or opinions (Lundy,
2012; Speziale & Carpenter, 2007). In fact, validation
and corroboration are important for determining
whether sources are genuine and authentic. External
criticism determines the “genuineness of primary
sources” (Lundy, 2012, p. 265). The researcher needs
to know where, when, why, and by whom a document
was written, which may involve verifying the hand-
writing or determining the age of the paper on which
it was written. Internal criticism involves establish-
ing the authenticity of the document. The researcher
may ask whether the document’s content is consistent
with what was known at the time the document was
written. Are dates, locations, and other details consis-
tent across sources? The researcher remains open to
the views presented in the documents or other sources
but remains somewhat skeptical until sources are
verified.
Historical Data Analysis
Data gathering and analysis occur simultaneously as
the researcher samples documents, seeking descrip-
tions, conflicting records, or contextual details. As
with other qualitative approaches, historical research-
ers become immersed in the data. Content analysis
yields data that the researcher will use to develop a
description of the topic. The connections made among
documents, opinions, and stories constitute the inter-
pretation of the data that are essential to an unfolding,
deep understanding of the topic. Figuring out when to
stop examining sources may be one of the major
challenges faced by historical researchers. Like phe-
nomenological researchers, who stop interviewing
participants when redundancy in the data is confirmed,
historical researchers decide to stop gathering data
when new data are no longer being found. The
researcher may return to data gathering if gaps or
questions emerge as the findings are being written.
Writing the Historical Narrative
The historical researcher keeps extensive records of
the source of each fact, event, and story that is
extracted. The extracted data may be organized as a
290 UNIT TWO  The Research Process
chronology or attached to an outline. The chronology
or outline will become the skeleton of the narrative
that will be written. The historical narrative may take
the form of a case study, a rich narrative, or a biogra-
phy. The links made by the historical researcher from
the past to the present give historical research its sig-
nificance to nursing (Lundy, 2012).
Irwin (2011) conducted a historical study but gave
few details of the methods, a situation that is common
with this qualitative research approach. Historical
researchers focus on the final product, which may be
a book, a documentary, or an essay. Early in her essay
of findings, Irwin indicates her purpose in conducting
the study:
“The archives, sources, and figures that are central to
the field [nursing] provide a ready means to trace the
spread of U.S. global influence in the early twentieth
century. The discipline should not be relegated to the
peripheries of U.S. international history, therefore,
but must be made central in any historical consider-
ation of the United States in the world.… To demon-
strate how the history of nursing can inform the study
of both U.S. international history and the broader
history of medicine, this essay traces the careers of
four nurses during World War I and the following
decade.” (Irwin, 2011, p. 80)
The four nurses who Irwin highlighted were
selected because of their service with the American
Red Cross, as described in this excerpt:
“Pansy V. Besom, Helen L. Bridge, Kathleen D’Olier,
and Alice L. Fitzgerald were among nearly 20,000
nurses from the United States who volunteered for
overseas service with the American Red Cross (ARC)
in World War I and the 1920s. These women gradu-
ated from U.S. hospital training schools during the
first decade of the twentieth century and went to
work as nurses in U.S. urban centers.” (Irwin, 2011,
p. 80)
A review of Irwin’s reference list makes it clear
that she used personal letters, press releases, and
organizational documents from the National Archives,
College Park, Maryland. Each citation includes the
box number and the file from which data were
extracted. In addition, Irwin listed more than 100
books as references.
Irwin (2011) intertwines the career of each nurse
with the concurrent events of that era. She concludes
as follows, with support for a fresh view of the United
States’ role in international health:
“Taking the histories of U.S. nurses into account and
highlighting their centrality to the study of U.S. inter-
national history invites a reconsideration of the history
of U.S. foreign relations in new and vital ways. These
nurses—Besom, Bridge, D’Olier, Fitzgerald, and
many others—are our U.S. international history, for
they were the actual, physical embodiments of the
United States in the world. Their letters, writings, and
biographies are an invaluable archive for scholars
interested in examining the on-the-ground workings
of U.S. influence and soft power.” (Irwin, 2011, p. 96)
This is a rigorous study with extensive archive and
literature support. The study could have been strength-
ened, however, if Irwin had provided more informa-
tion about how she integrated the facts from various
documents into a coherent whole.
KEY POINTS
• Qualitative methods are more flexible than quanti-
tative methods to ensure that the participant’s voice
is heard.
• Qualitative data collection and data analysis occur
simultaneously.
• Researchers and participants in qualitative studies
co-create the data that will be analyzed.
• Qualitative methods of data collection are observa-
tion, interviews, focus groups, images, and elec-
tronically mediated communication.
• Recordings and notes are transcribed into data files
prior to analysis.
• Qualitative researchers select coding and analysis
strategies consistent with the philosophical
approach of their studies.
• Phenomenological methods may include bracket-
ing and interviewing to elicit rich descriptions of
lived experiences.
• Methods specific to grounded theory studies are
coding, describing concepts, and identifying links
between the concepts for the purpose of developing
a theory.
• Ethnographic methods are characterized by exten-
sive field work that includes observations and inter-
views for the purpose of describing aspects of the
culture being studied.
 CHAPTER 12  Qualitative Research Methodology 291
• Exploratory qualitative studies may use a theoreti-
cal perspective on the research topic as the basis
for data analysis.
• Historical researchers extract the meaning from
primary and secondary source documents to
describe and analyze the context and chronology of
past events.
• Rigorous qualitative researchers are reflexive, a
characteristic that requires the ability to be aware
of nuances of the research situation and one’s own
biases.
REFERENCES
Alcock, J. & Iphofen, R. (2007). Computer-assisted software tran-
scription of qualitative interviews. Nurse Researcher, 15(1),
16–26.
Anatchkova, M. D. & Bjorner, J. B. (2010). Health and role func-
tioning: The use of focus groups in the development of an item
bank. Quality of Life Research, 19(1), 111–123.
Baker, T. A. & Wang, C. C. (2006). Photovoice: Use of a participa-
tory action research method to explore the chronic pain experi-
ence of older adults. Qualitative Health Research, 16(10),
1405–1413.
Banner, D. (2010a). Becoming a coronary artery bypass graft
surgery patient: A grounded theory study of women’s experiences.
Journal of Clinical Nursing, 19(21/22), 3123–3133.
Banner, D. (2010b). Qualitative interviewing: Preparation for
practice. Canadian Journal of Cardiovascular Nursing, 20(2),
27–30.
Banner, D. & Albarran, J. W. (2009). Computer-assisted qualitative
data analysis software: A review. Canadian Journal of Cardiovas-
cular Nursing, 19(3), 24–27.
Belsky, J. (1984). The determinants of parenting: A process model.
Child Development, 55(1), 83–96.
Byers, D. C. & France, N. E. M. (2008). The lived experience
of registered nurses providing care for patients with dementia in
the acute care setting: A phenomenological study. International
Journal for Human Caring, 12(4), 44–49.
Charmaz, K. (2006) Constructing grounded theory: A practical
guide through qualitative analysis. London, England: Sage
Publications.
Chen, H.-Y. (2010). The lived experience of moving forward for
clients with spinal cord injury: A Parse research method study.
Journal of Advanced Nursing, 66(5), 1132–1141.
Cheng, H. C., Qin, L. X., & Tee, H. K. (2009). An exploratory study
on the isolation experiences of patients with haematological dis-
orders. Singapore Nursing Journal, 35(1), 15–23.
Coffman, M. J., Shobe, M. A., O’Connell, B. (2008). Self-
prescription practices in recent Latino immigrants. Public Health
Nursing, 25(3), 203–211.
Colaizzi, P. (1978). Psychological research as the phenomenologist
views it. In R.S. Valle, M. King (Eds.), Existential phenomeno-
logical alternatives for psychology (pp. 48–71). New York, NY:
Oxford University Press.
Cooney, A. (2010). Choosing between Glaser and Strauss: An
example. Nurse Researcher, 17(4), 18–28.
Cooper, C. M. & Yarborough, S. P. (2010). Tell me-show me: Using
combined focus group and photovoice methods to gain under-
standing of health issues in rural Guatemala. Qualitative Health
Research, 20(5), 644–653.
Corbin, J. & Strauss, A. (2008). Basics of qualitative research:
Techniques and procedures for developing grounded theory (3rd
ed.). Thousand Oaks, CA: Sage Publications.
Creswell, J. W. (2007). Qualitative inquiry and research design:
Choosing among five approaches (2nd ed.). Thousand Oaks, CA:
Sage Publications.
Davidson, C. (2009). Transcription: Imperative for qualitative
research. International Journal of Qualitative Methods, 8(1),
36–52.
Devault, A., Milcent, M.-P., Ouellet, F., Laurin, I., Jauron, M., &
Lacharite, C. (2008). Life stories of young fathers in the contexts
of vulnerability. Fathering, 6(3), 226–248.
Dickinson, A., Machan, I., Horton, K., Jain, D., & Maddex, J. (2011).
Fall prevention in the community: What older people say they
need. British Journal of Community Nursing, 16(4), 174–180.
Dodd, V., Glassman, T., Arthur, A., Webb, M., & Miller, M. (2010).
Why underage college students drink in excess: Qualitative
research findings. American Journal of Health Education, 41(2),
93–101.
Dowling, M. (2007). From Husserl to van Manen: A review of dif-
ferent phenomenological approaches. International Journal of
Nursing Studies, 44(1), 131–142.
Duffy, M. (2012). Narrative inquiry: The method. In P. L. Munhall
(Ed.), Nursing research: A qualitative perspective (5th ed.) (pp.
421–440). Sudbury, MA: Jones & Bartlett.
Fawcett, J. & Garity, J. (2009). Evaluating research for evidence-
based nursing practice. Philadelphia, PA: F. A. Davis.
Findholt, N. E., Michael, Y. L., & Davis, M. M. (2011). Photovoice
engages rural youth in childhood obesity prevention. Public
Health Nursing, 28(2), 186–192.
Finn, N. A. & Stube, J. E. (2010). Post stroke fatigue: Qualitative
study of three focus groups. Occupational Therapy International,
17(2), 81–91.
Fleming, J., Mahoney, J., Carlson, E., & Engebretson, J. (2009). An
ethnographic approach to interpreting a mental illness photovoice
exhibit. Archives of Psychiatric Nursing, 23(1), 16–24.
Garcia-Horton, J. B. & Guerra-Ramos, M. T. (2009). The use of
CADQDAS in educational research: Some advantages, limita-
tions, and potential risks. International Journal of Research &
Method in Education, 32(2), 151–165.
Garwick, A. W., Seppelt, A., & Riesgraf, M. (2010). Addressing
asthma management challenges in multisite, urban Head Start
Program. Public Health Nursing, 27(4), 329–336.
Gharaibeh, M. & Owels, A. (2009). Why do Jordanian women stay
in an abusive relationship: Implications for health and social well-
being. Journal of Nursing Scholarship, 41(4), 376–384.
Giorgi, A. (1970). Psychology as a human science: A phenomeno-
logically based approach. New York, NY: Harper & Row.
Glaser, B. G. & Strauss, A. (1967). The discovery of grounded
theory: Strategies for qualitative research. Chicago, IL: Aldine.
Gray, J. (2009). Rooms, recording, and responsibilities: The logis-
tics of focus groups. Southern Online Journal of Nursing Research,
9(1), Article 5. Available from http://snrs.org/publications/
SOJNR_articles2/Vol09Num01Art05.html.
292 UNIT TWO  The Research Process
Greco, K. E., Nail, L. M., Kendall, J., Cartwright, J., & Messacar,
D. C. (2010). Mammography decision making in older women
with a breast cancer family history. Journal of Nursing Scholar-
ship, 42(3), 348–356.
Haahr, A., Kirkevold, M., Hall, E. O. C., & Ostergaard, K. (2010).
Living with advanced Parkinson’s disease: A constant struggle
with unpredictability. Journal of Advanced Nursing, 67(2),
408–417.
Halcomb, E. J. & Davidson, P. M. (2006). Is verbatim transcription
of interview data always necessary? Applied Nursing Research,
19(1), 38–42.
Hall, L. M., Pedersen, C., & Fairley, L. (2010). Losing the moment:
Understanding interruptions to nurses’ work. Journal of Nursing
Administration, 40(4), 169–176.
Hamilton, J. B., Moore, C. E., Powe, B. D., Agarwai, M., & Martin,
P. (2010). Perceptions of support among older African American
cancer survivors. Oncology Nurse Forum, 37(4), 484–493.
Henneman, E. A., Gawlinski, A., Blank, F. S., Henneman, P. L.,
Jordan, D., & McKenzie, J. B. (2010). Strategies used by critical
care nurses to identify, interrupt, and correct medical errors.
American Journal of Critical Care, 19(6), 500–509.
Hoover, R. S. & Koerber, A. L. (2011). Using NVivo to answer the
challenges of qualitative research in professional communication:
Benefits and best practices. IEEE Transations on Professional
Communication, 54(1), 68–82.
Hutchinson, A. (2005). Analysing audio-recorded data: Using com-
puter software applications. Nurse Researcher, 12(3), 20–31.
Im, E.-O., Chee, W., Lim, H.-J., Liu, Y., & Kim, H. K. (2008).
Midlife women’s attitudes toward physical activity. Journal of
Obstetric, Gynecologic and Neonatal Nursing, 37(2), 203–213.
Im, E.-O., Lee, B., & Chee, W. (2010). Shielded from the real world:
Perspectives on Internet cancer support groups by Asian Ameri-
cans. Cancer Nursing, 33(3), E10–E20.
Im, E.-O., Lee, B., Chee, W., Domire, S., & Brown, A. (2010). A
national multiethnic online forum study on menopausal symptom
experience. Nursing Research, 59(1), 26–33.
Im, E.-O., Lee, B., Chee, W., Stuifbergen, A. and the eMAPA
Research Team. (2011). Attitudes toward physical activity of
white midlife women. Journal of Obstetric, Gynecologic and
Neonatal Nursing, 40(3), 312–321.
Im, E.-O., Lee, S. H., Liu, Y., Lim, H.-J., Guevara, E., & Chee, W.
(2009). A national online forum on ethnic differences in cancer
pain experiences. Nursing Research, 58(2), 86–91.
Im, E.-O., Lee, S. H., & Chee, W. (2010). Black women in meno-
pausal transition. Journal of Obstetric, Gynecologic and Neonatal
Nursing, 39(4), 435–443.
Irwin, J. F. (2011). Nurses without borders: The history of nursing
as U.S. international history. Nursing Historical Review, 19(1),
78–102.
Jirwe, M. (2011). Analysing qualitative data. Nurse Researcher,
18(3), 4–5.
Kaddoura, M. A. (2010). New graduate nurses’ perceptions of the
effects of clinical simulation on their critical thinking, learning,
and Confidence. The Journal of Continuing Education in Nursing,
41(11), 506–516.
Kostovich, C., Saban, K., & Collins, E. (2010). Becoming a nurse
researcher: The importance of mentorship. Nursing Science Quar-
terly, 23(4), 281–286.
Lange, B. (2007). The prescriptive power of caring for self: Women
in recovery from substance use disorders. International Journal
of Human Caring, 11(2), 74–80.
Lauzon Clabo, L. M. (2008). An ethnography of pain assessment
and the role of social context on two postoperative units. Journal
of Advanced Nursing, 61(5), 531–539.
Leech, N. L. & Onwuegbuzie, A. J. (2011). Beyond constant com-
parison qualitative data analysis: Using NVivo. School Psychol-
ogy Quarterly, 26(1), 70–84.
Leininger, M. M. (1985). Transcultural care diversity and universal-
ity: A theory of nursing. Nursing and Health Care, 6(4),
208–212.
Lesniak., R. G. (2008). The lived experience of adolescent females
who self-injure by cutting (Doctoral dissertation). Dissertation
Abstracts International, B60 (05). (UMI No. 9932285).
Lesniak, R. G. (2010). The lived experience of adolescent females
who self-injure by cutting. Advanced Emergency Nursing Journal,
32(2), 137–147.
Liamputtong, P. & Ezzy, D. (2005). Qualitative research methods
(2nd ed.). Victoria, Australia: Oxford University Press.
Lopez, R. P. (2009). Decision-making for acutely ill nursing home
residents: Nurses in the middle. Journal of Advanced Nursing,
65(5), 1001–1009.
Lundy, K. S. (2012). Historical research. In P. L. Munhall (Ed.),
Nursing research: A qualitative perspective (5th ed.) (pp. 381–
397). Sudbury, MA: Jones & Bartlett.
Mahoney, D. F. & Ladd, E. (2010). More than a prescriber: Geron-
tological nurse practitioners’ perspectives on prescribing and
pharmaceutical marketing. Geriatric Nursing, 31(1), 17–27.
Maslow., A. H. (1987).Motivation and personality (3rd ed.). New
York City, NY: Harper & Row Publishers.
Marshall, C. & Rossman, G. B. (2011). Designing qualitative
research (5th ed.). Los Angeles, CA: Sage.
McCloskey, S. & Taggart, L. (2010). How much compassion have
I left? An exploration of occupational stress among children’s
palliative care nurses. International Journal of Palliative Care,
16(5), 233–240.
McGibbon, E., Peter, E. & Gallop, R., (2010) An Institutional Eth-
nography of Nurses’ Stress. Qualitative Health Research, DOI:
10.1177/1049732310375435.
Moyle, W. (2002). Unstructured interviews: Challenges when par-
ticipants have major depressive illness. Journal of Advanced
Nursing, 39(3), 266–273.
Munhall, P. L. (2012). Nursing research: A qualitative perspective
(5th ed.). Sudbury, MA: Jones & Bartlett.
Neuendorf, K. A. (2002). The content analytic guidebook. Newbury
Park, CA: Sage.
Nikander, P. (2008). Working with transcripts and translated data.
Qualitative Research in Psychology, 5(3), 225–231.
Ollson, L., Ostlund, G., Strang, P., Grassman, E. J., & Friedrichsen,
M. (2010). Maintaining hope when close to death: Insight from
cancer patients in palliative home care. International Journal of
Palliative Nursing, 16(12), 607–612.
Ornelas, I. J., Arnell, J., Tran, A. N., Royster, M., Armstrong-
Brown, J., & Eng, E. (2009). Understanding African American
men’s perceptions of racism, male gender socialization, and social
capital through photovoice. Qualitative Health Research, 19(4),
552–565.
 CHAPTER 12  Qualitative Research Methodology 293
Owen-Smith, A., Sterk, C., McCarty, F., Hankerson-Dyson, D., &
DiClemente, R. (2010). Development and evaluation of a comple-
mentary and alternative medicine use survey in African-Americans
with acquired immune deficiency syndrome. Journal of Alterna-
tive & Complementary Medicine, 16(5), 569–577.
Parse, R. R. (2001). Qualitative inquiry: The path of sciencing.
Boston, MA: Jones and Bartlett Publishers.
Rampton, T. B., Rosemann, J. L., Latta, A. L., Mandleco, B. L.,
Roper, S. O., & Dyches, T. T. (2007). Images of life: Siblings of
children with Downsyndrome. Journal of Family Nursing, 13(4),
420–442.
Richards, K. M. (2008). RAP Project: An instrument development
study to determine common attributes for pain assessment among
men and women who represent multiple pain-related diagnoses.
Pain Management Nursing, 9(1), 33–43.
Roy, C. (1976). Introduction to nursing: An adaptation model.
Englewood Cliffs, NJ: Prentice-Hall.
Schiellerup, P. (2008). Stop making sense: The trials and tribula-
tions of qualitative data analysis. Area, 40(2), 161–171.
Schumacher, G. (2010). Culture care meanings, beliefs, and prac-
tices in rural Dominican Republic. Journal of Transcultural
Nursing, 21(2), 93–103.
Scott, A. J. & Wilson, R. F. (2011). Social determinants of health
among African Americans in a rural community in the Deep
South: An ecological study. Rural and Remote Health, 11, 1634
(Online). Available from http://www.rrh.org.au.
Scott, D., Sharpe, H., O’Leary, K., Dehaeck, U., Hindmarsh, K.,
Moore, J. G., & Osmond, M. H. (2009). Court reporters: A viable
solution for the challenges of focus group data collection? Quali-
tative Health Research, 19(1), 140–146.
Seidman, I. (2006). Interviewing as qualitative research: A guide
for researchers in education and the social sciences. New York
City, NY: Teachers College Press.
Shorter, M. & Stayt, L. C. (2010) Critical care nurses’ experiences
of grief in an adult intensive care unit. Journal of Advanced
Nursing, 66(1), 159–167.
Simmons, D. & Sherwood, G. (2010). Self-prescription practices in
recent Latino immigrants. Critical Care Nursing Clinics of North
America, 22(2), 253–260.
Speziale, H. J. S. & Carpenter, D. R. (2007). Qualitative research
in nursing: Advancing the humanistic perspective (4th ed.). Phila-
delphia. PA: Lippincott Williams & Wilkins.
Steele, R. G., Wu, Y. P., Jensen, C. D., Pankey, S., Davis, A. M., &
Aylward, B. S. (2011). School nurses’ perceived barriers to dis-
cussing weight with children and their families: A qualitative
approach. Journal of School Health, 81(3), 128–137.
Strauss, A. & Corbin, J. (1998). Basics of qualitative research:
Techniques and procedures for developing grounded theory (2nd
ed.). Thousand Oaks, CA: Sage Publications.
Swanlund, S. L. (2010). Successful medication management pro-
cesses as perceived by community-dwelling adults over age 74.
Applied Nursing Research, 23(1), 22–29.
Taylor, C. & Burch, J. (2011). Feedback on an enhanced recovery
programme for colorectal surgery. British Journal of Nursing,
20(5), 286–290.
Trimm, D. R. & Sanford, J. T. (2010). The process of family waiting
during surgery. Journal of Family Nursing, 16(4), 435–461.
van Kaam, A. (1966). Existential foundations of psychology. Pitts-
burgh, PA: Duquesne University Press.
van Manen, M. (1984). “Doing” phenomenological research and
writing. Alberta, Canada: University of Alberta Press.
Walls, P., Parahoo, K., & Fleming, P. (2010). The role and place of
knowledge and literature in grounded theory. Nurse Researcher,
17(4), 8–17.
Wang, C., Burris, M. A., & Ping, X. Y. (1996). Chinese village
women as visual anthropologists: A participatory approach to
reaching policymakers. Social Science & Medicine, 42(10),
1391–1400.
Whitehead, L. C. (2007). Methodological and ethical issues in
Internet-mediated research in the field of health: An integrated
review of the literature. Social Science & Medicine, 65(4),
782–791.
Wolf, Z. E. (2012). Ethnography: The method. In P. L. Munhall
(Ed.), Nursing research: A qualitative perspective (5th ed.) (pp.
285–338). Sudbury, MA: Jones & Bartlett.
Wuest, J. (2012). Grounded theory: The method. In P. L. Munhall
(Ed.), Nursing research: A qualitative perspective (5th ed.) (pp.
225–256). Sudbury, MA: Jones & Bartlett.
Zengerle-Levy, K. (2006). Nursing the child who is alone in the
hospital. Pediatric Nursing, 32(3), 226–231, 237.

13
CHAPTER
Outcomes Research
O
utcomes research, now an established field
of health research, focuses on the end results
of patient care. Numerous studies have been
conducted by nursing and medicine over the last three
decades in the United States (Aiken, Clarke, Sloane,
Lake, & Cheney, 2008; Bakker et al., 2011; Stone
et al., 2007), Canada (Tourangeau, 2003; Tourangeau
et al., 2007), and internationally (Van den Heede et al.,
2009) that explore the relationships among nursing
interventions and patient outcomes. Nursing-related
interventions and variables that have been studied
include skill mix and configuration of nursing person-
nel; staffing levels; assignment patterns (primary,
functional, or team); shift patterns; levels of nursing
education, experience, and expertise; ratios of full-
time to part-time nurses; level and type of nursing
leadership available centrally and on units; cohesion
and communication among the nursing staff and
between nurses and physicians; the implementation of
clinical care maps for patients with selected diagnoses;
and the interrelationships of these factors.
The momentum propelling outcomes research
comes primarily from policy makers, insurers, and the
public. There is a growing demand for data from
providers that justify interventions and costs and for
systems of care that demonstrate improved patient
outcomes. By linking the care people receive to the
outcomes they experience, outcomes research has
become the key to developing better ways to monitor
and improve the quality of care. There has been a
major shift in published nursing studies as the number
of studies using traditional quantitative or qualitative
methods is dwarfed by the number of outcomes
studies.
This chapter describes the theoretical basis of out-
comes research, provides a brief history of the emerg-
ing endeavors to examine outcomes, explains the
importance of outcomes research designed to examine
nursing practice, and highlights methodologies used
294
  http://evolve.elsevier.com/Grove/practice/
in outcomes research. A broad base of literature from
a variety of disciplines was used to develop the content
for this chapter, in keeping with the multidisciplinary
perspective of outcomes research.
Theoretical Basis of
Outcomes Research
The theory on which outcomes research is based
emerged from evaluation research (Structure-
Process-Outcomes Framework). The theorist Avedis
Donabedian (1976, 1978, 1980, 1982, 2005) proposed
a theory of quality health care and the process of
evaluating it. Quality is the overriding construct of the
theory; however, Donabedian never defined this
concept himself (Mark, 1995). He noted that quality
of care is a “remarkably difficult notion to define and
is a reflection of the values and goals current in the
medical care system and in the larger society of which
it is a part” (Donabedian, 2005, p. 692). The World
Health Organization (2009, p. 13) defined quality of
care as the “degree to which health services for indi-
viduals and populations increase the likelihood of
desired health outcomes and are consistent with
current professional knowledge.”
The cube shown in Figure 13-1 explains the ele-
ments of quality health care. The three dimensions of
the cube are health, subjects of care, and providers of
care. The concept health has many aspects; three
are shown on the cube: physical-physiological func-
tion, psychological function, and social function.
Donabedian (1987, p. 4) proposed that “the manner in
which we conceive of health and of our responsibility
for it, makes a fundamental difference to the concept
of quality and, as a result, to the methods that we use
to assess and assure the quality of care.”
The concept subjects of care has two primary
aspects: patient and person. A patient is defined as

Plan, institution, system
Organized team
Several practitioners*
Individual practitioner
Physical-psychological function
Psychological function
Social function
Individual
Aggregate:
case load
Patient
*Of the same profession or of different professions
Figure 13-1  Level and scope of concern as factors in the definition of quality.
someone who has already gained access to some care,
and a person as someone who may or may not have
gained access to care. Each of these concepts is further
categorized by the concepts individual and aggregate.
Within patient, the aggregate is a caseload; within
person, the aggregate is a target population or a
community.
The concept providers of care shows levels of
aggregation and organization of providers. The first
level is the individual practitioner. At this level, con-
sideration is given to the individual provider rather
than others who might be involved in the subject’s
care, whether individual or aggregate. As the levels
progress, providers of care include several practitio-
ners, who might be of the same profession or different
professions and “who may be providing care con­
currently, as individuals, or jointly, as a team”
(Donabedian, 1987, p. 5). At higher levels of aggrega-
tion, the provider of care is institutions, programs, or
the healthcare system as a whole.
Donabedian theorized that the dimensions of health
are defined by the subjects of care, not by the provid-
ers of care, and are based on “what consumers expect,
want, or are willing to accept” (Donabedian, 1987, p.
5). Thus, practitioners cannot unilaterally enlarge the
definition of health to include other aspects; this action
CHAPTER 13  Outcomes Research 295
Individual
Aggregate:
target
population,
community
Person
requires social consensus that “the scope of profes-
sional competence and responsibility embraces these
areas of function” (Donabedian, 1987, p. 5). Donabe-
dian indicated, however, that providers of care may
make efforts to persuade subjects of care to expand
their definition of the dimensions of health.
The essence of Donabedian’s framework is the
physical-physiological function of the individual
patient being cared for by the individual practitioner.
Examining quality at this level is relatively simple. As
one moves outward to include more of the cubical
structure, the notions of quality and its assessment
become increasingly difficult (see Figure 13-1). When
more than one practitioner is involved, both individual
and joint contributions to quality must be evaluated.
Concepts such as coordination and teamwork must be
conceptually and operationally defined. When a person
is the subject of care, an important attribute is access.
When an aggregate is the subject of care, an important
attribute is resource allocation. Access and resource
allocation are interrelated, because they each define
who gets care, the kind of care received, and how
much care is received.
As more elements of the cube are included, con-
flicts among competing objectives emerge. The chief
conflict is between the practitioner’s responsibilities
296 UNIT TWO  The Research Process
to the individual and to the aggregate. The practitioner
is expected to have an exclusive commitment to each
patient, yet the aggregate demands a commitment to
the well-being of society, a situation that may lead to
ethical dilemmas for the practitioner. Spending more
time with an individual patient decreases access for
other patients. Society’s demand to reduce costs for an
overall financing program may require raising costs to
the individual. From an examination of the cube, logic
would suggest that one could build up quality begin-
ning with the primordial, or elemental, cell and
increase by increments with the assumption that each
increment would contribute positively to a greater
total quality (see Figure 13-1). However, the conflicts
among competing objectives may preclude this pos-
sibility and lead instead to moral dilemmas.
Donabedian (1987) identified three objects to eval-
uate when appraising quality: structure, process, and
outcome. A complete quality assessment program
requires the simultaneous use of all three constructs
and an examination of the relationships among the
three. However, researchers have had little success in
accomplishing this theoretical goal. Studies designed
to examine all three constructs would require suffi-
ciently large samples of various structures, each with
the various processes being compared and large
samples of subjects who have experienced the out-
comes of those processes. The funding and the coop-
eration necessary to accomplish this goal have not yet
been realized; however, examples of nursing research
in which two or more aspects have been evaluated are
provided in this chapter.
Evaluating Outcomes
The goal of outcomes research is the evaluation of
outcomes as defined by Donabedian. However, this
goal is not as simple as it might immediately appear.
Donabedian’s theory requires that identified outcomes
be clearly linked with the process that caused the
outcome. Researchers need to define the process and
justify the causal links with the selected outcomes.
The identification of desirable outcomes of care
requires dialogue between the subjects of care and the
providers of care. Although the providers of care may
delineate what is achievable, the subjects of care must
clarify what is desirable. The outcomes must also be
relevant to the goals of the health professionals, the
healthcare system of which the professionals are a
part, and society.
Outcomes are time dependent. Some outcomes
may not be apparent for a long period after the process
that is purported to cause them, whereas others may
be apparent immediately. Some outcomes are
temporary, and others are permanent. Thus, an appro-
priate time frame for determining the selected out-
comes must be established.
A final obstacle to outcomes evaluation is deter-
mining attribution. This requires assigning the place
and degree of responsibility for the outcomes observed.
Many factors other than health care may influence
outcomes, and precautions must be taken to hold all
significant factors other than health care constant or
account for their effect if valid conclusions can be
drawn from outcomes research. A particular outcome
is often influenced by a multiplicity of factors. Patient
factors, such as compliance, predisposition to disease,
age, propensity to use resources, high-risk behaviors
(e.g., smoking, poor dietary habits, and drug abuse),
and lifestyle, must be taken into account. Environmen-
tal factors such as air quality, public policies related
to smoking, and occupational hazards must be
included. The responsibility for outcomes may be dis-
tributed among providers, patients, employers, insur-
ers, the community, and the government.
There is as yet little scientific basis for judging the
precise relationship between each of these factors and
the selected outcome. Many of the influencing factors
may be outside the jurisdiction or influence of the
healthcare system or of the providers within it. One
solution to this problem of identifying relevant out-
comes is to define a set of proximate outcomes spe-
cific to the condition for which care is being provided.
Critical pathways and care maps may help the
researcher to define at least proximate outcomes.
However, proximate outcomes do not provide the
level of evidence that examining the desired outcomes
does. In addition, researchers must be ever vigilant not
to confound causality and correlation, because an
association or relationships between or among factors
cannot imply that certain factors caused particular
outcomes.
Evaluating Process
Clinical management has been, for most health profes-
sionals, an art rather than a science. Understanding the
process sufficiently to study it must begin with much
careful reflection, dialogue, and observation. There are
multiple components of clinical management, many of
which have not yet been clearly defined or tested.
Three components of process that are of particular
interest to Donabedian are standards of care, practice
styles, and costs of care.
Standards of Care
A standard of care is a norm on which quality of care
is judged. Clinical guidelines, critical paths, and

care maps define standards of care for particular situ-
ations. According to Donabedian (1987), a practi­
tioner has legitimate responsibility to apply available
knowledge when managing a dysfunction or disease
state. This management consists of (1) identifying or
diagnosing the dysfunction; (2) deciding whether or
not to intervene; (3) choosing intervention objectives;
(4) selecting methods and techniques to achieve the
objectives; and (5) skillfully executing the selected
techniques or interventions.
Donabedian (1987) recommended the development
of criteria to be used as a basis for judging the quality
of care. These criteria may take the form of clinical
guidelines or care maps based on prior validation that
the care contributed to the desired outcomes. The clini-
cal guidelines published by the Agency for Healthcare
Research and Quality (AHRQ, 2011) establish norms
or standards against which the validity of clinical man-
agement can be judged. These norms are now estab-
lished through clinical practice guidelines available
through the National Guideline Clearinghouse within
the AHRQ (see http://www.guideline.gov/). However,
the core of the problem, from Donabedian’s perspec-
tive, is clinical judgment. Analysis of the process of
making diagnoses and therapeutic decisions is critical
to the evaluation of the quality of care. The emergence
of decision trees and algorithms is a response to
Donabedian’s concerns and provides a means of evalu-
ating the adequacy of clinical judgments.
Practice Styles
The style of practitioners’ practice is another dimen-
sion of the process of care that influences quality;
however, it is problematic to judge what constitutes
“goodness” in style and to justify the decisions regard-
ing it. Donabedian (1987) identified the following
problem-solving styles: (1) routine approaches to care
versus flexibility; (2) parsimony versus redundancy;
oasis-ebl|Rsalles|1475959208
(3) variations in degree of tolerance of uncertainty; (4)
propensity to take risks; and (5) preference for Type I
errors versus Type II errors. There are also diverse
styles of interpersonal relationships. Westert and
Groenewegen (1999, p. 174) suggest that “differences
in practice styles are a result of differences in oppor-
tunities, incentives, and influences.” They suggest that
“there is an (often implicit) idea of what should be
done and how, and this shared (local) standard influ-
ences the choices made by individual practitioners.”
This alternative originates in the borders between eco-
nomics and sociology and it can be characterized as
the social production function approach. The Medical
Outcomes Study, described later in this chapter, was
designed to determine whether variations in patient
CHAPTER 13  Outcomes Research 297
outcomes are explained by differences in system of
care, clinician specialty, and clinicians’ technical and
interpersonal styles (Tarlov et al., 1989).
Practice pattern is a concept closely related to
practice style. Although practice style represents
variation in how care is provided, practice pattern rep-
resents variation in what care is provided. Small area
analysis is an example of research that attempts to
describe variation in practice patterns. Researchers of
variations in practice patterns have found that such
variation is not totally explained by patients’ clinical
conditions. For example, previous researchers have
found that prescribing practices differ by region of the
country and are influenced in part by drug company
resources and marketing practices. The research meth-
odology involved in small area analysis is described
later in this chapter, under the term geographical
analysis.
Evidence-based practice (EBP) is another dimen-
sion of the process of care that is considered a critical
aspect of professional practice (Stetler & Caramanica,
2007). The ultimate goals of EBP are improved patient
health status and quality of care (Graham, Bick,
Tetroe, Strause, & Harrison, 2011). Thus, the impact
of EBP should be assessed through measurement of
patient outcomes. Graham et al. (2011) suggest that
when planning evaluation of EBP, evaluators should
begin by defining a clear question to guide their evalu-
ation. They use the PICO question as a framework for
structuring their evaluation question (Straus, Tetroe,
Graham, Zwarenstein, & Bhattacharyya, 2009). P
refers to the population of interest, which could be the
public or a specific patient population. I refers to the
implementation of a knowledge translation (KT) inter-
vention. C refers to a comparison or control group who
did not receive the intervention, and O refers to the
outcome of interest. For instance, the outcome could
be improvement in health status. Very few empirical
studies have assessed the impact of evidence-based
nursing practice on patient outcomes. Davies, Edwards,
Ploeg, and Virani (2008) found implementation of best
practice guidelines in nursing resulted in improved
outcomes in diverse settings, but there was consider-
able variability in the indicators evaluated, suggesting
the need for more research in this area. Chapter 19
provides more details on the synthesis of research
evidence and the implementation of EBP in nursing.
Costs of Care
A third dimension of the examination of the process
of care is cost. There are cost consequences to
maintaining a specified level of quality of care. Pro-
viding more and better care is likely to raise costs but
298 UNIT TWO  The Research Process

is also likely to produce savings. Economic benefits
(savings) result from preventing illness, preventing
complications, maintaining a higher quality of life,
and prolonging productive life. The Institute for
Healthcare Improvement has identified the following
simple framework for identifying the return on invest-
ment for quality initiatives (Sadler, Joseph, Keller, &
Rostenberg, 2009):
Step 1: Identify your improvement goal.
Step 2: Estimate improvement costs.
Step 3: Calculate revenue improvement through cost
avoidance.
Step 4: Calculate the return on investment.
Schifalacqua, Mamula, and Mason (2011) used this
framework in their development of a cost of care cal-
culator for evaluating an EBP program. The goal was
to compare the costs of care before and after imple-
mentation of an EBP program. The cost of care analy-
sis focused on potential cost savings realized through
the prevention of serious adverse events, such as
healthcare-associated infections, pressure ulcers, and
falls. Many challenges were revealed in identifying
costs that were comparable among healthcare facilities
participating in the implementation of the EBP
program for preventing serious adverse events. Base-
line costs associated with each adverse event were
estimated on the basis of U.S. published studies and
refined using organizational event data.
A related issue is who bears the costs of care. Some
measures purported to reduce costs have instead
simply shifted costs to another party. For example, a
hospital might reduce its costs by discharging a par-
ticular type of patient early, but total costs could
increase if the necessary community-based health care
raised costs above those incurred by keeping the
patient hospitalized longer or if complications resulted
in rehospitalization. In this case, the third-party pro-
vider could experience higher costs. In many cases,
the costs are shifted from the healthcare system to the
family as out-of-pocket costs or costs that are not
covered by healthcare reimbursement systems. Studies
examining changes in costs of care should consider
total costs, which include out-of-pocket costs. Table
13-1 provides a few examples of studies that examine
the direct and indirect costs of care.
Evaluating Structure
Structures of care are the elements of organization
and administration, as well as provider and patient
characteristics, that guide the processes of care. The
first step in evaluating structure is to identify and
describe the elements of the structure. Various admin-
istration and management theories could be used to
identify the elements of the structure. These elements
might be leadership, tolerance of innovativeness,

TABLE 13-1  Studies Examining Costs of Care

Year Source
2011 Madsen, L. B., Christiansen, T., Kirkegaard, P., & Pedersen, E. B. (2011). Economic evaluation of home blood
pressure telemonitoring: A randomized controlled trial. Blood Pressure, 20(2), 117–125.
2010 Spekle, E. M., Heinrich, J., Hoozemans, M. J., Blatter, B. M., van der Beek, A. J., van Dieen et al. (2010). The
cost-effectiveness of the RSI QuickScan program for computer workers: Results of an economic evaluation
alongside a randomized controlled trial. BMC Musculoskeletal Disorders, 11, 259–271.
2010 Van Oostrom, S. H., Heymans, M. W., de Vet, H. C., van Tulder, M. W., van Mechelen, W., & Anema, J. R.
(2010). Economic evaluation of a workplace intervention for sick-listed employees with distress. Occupational
& Environmental Medicine, 67(9), 603–610.
2009 Marchetti, A., & Rossiter, R. (2009). Managing acute acetaminophen poisoning with oral versus intravenous
N-acetylcysteine: A provider-perspective cost analysis. Journal of Medical Economics, 12(4), 384–391.
2007 Griffiths, P. D., Edwards, M. H., Forbes, A., Harris, R. L., & Ritchie, G. (2007). Effectiveness of intermediate
care in nursing-led inpatient units. Cochrane Database of Systematic Reviews, (2), CD002214.
2006 Phibbs, C. S., Holty, J. C., Goldstein, M. K., Garber, A. M., Wang, Y., Feussner, J.R., et al. (2006). The effect of
geriatrics evaluation and management on nursing home use and health care costs: Results from a randomized
trial. Medical Care, 44(1), 91–95.
2005 Harris, R., Richardson, G., Griffiths, P., Hallett, N., & Wilson-Barnett, J. (2005). Economic evaluation of a
nursing-led inpatient unit: The impact of findings on management decisions of service utility and sustainability.
Journal of Nursing Management, 13(5), 428–438.
2004 Altimier, L. B., Eichel, M., Warner, B., Tedeschi, L., & Brown, B. (2004). Developmental care: Changing the
NICU physically and behaviorally to promote patient outcomes and contain costs. Neonatal Intensive Care,
17(2), 35–39.
 CHAPTER 13  Outcomes Research 299

TABLE 13-2  Studies Investigating the Relationship between Structural Variables and Outcomes
Year Source
2011 McHugh, M. D., Shang, J., Sloane, D. M., & Aiken, L. H. (2011). Risk factors for hospital-acquired ‘poor glycemic
control’: A case control study. International Journal for Quality in Health Care, 23(1), 44–51.
2011 Trinkoff, A. M., Johantgen, M., Storr, C. L., Gurses, A. P., Liang, Y., & Han, K. (2011). Nurses’ work schedule
characteristics, nurse staffing, and patient mortality. Nursing Research, 60(1), 1–8.
2010 Flynn, L., Liang, Y., Dickson, G. L., & Aiken, L. H. (2010). Effects of nursing practice environments on quality
outcomes in nursing homes. Journal of the American Geriatrics Society, 58(12), 2401–2406.
2010 Fries, C. R., Earle, C. C., & Silber, J. H. (2010). Hospital characteristics, clinical severity, and outcomes for surgical
oncology patients. Surgery, 147(5), 602–609.
2010 Cummings, G. G., Midodzi, W. K., Wong, C. A., & Estabrooks, C. A. (2010). The contribution of hospital nursing
leadership styles to 30-day patient mortality. Nursing Research, 59(5), 331–339.
2009 Silber, J. H., Rosenbaum, P. R., Romano, P. S., Rosen, A. K., Wang, Y., Teng, Y., et al. (2009). Hospital teaching
intensity, patient race, and surgical outcomes. Archives of Surgery, 144(2), 113–120.
2008 Kutney-Lee, A., & Aiken, L. H. (2008). Effect of nurse staffing and education on the outcomes of surgical patients with
comorbid serious mental illness. Psychiatric Services, 59(12), 1466–1469.
2007 Castle, N. G., & Engberg, J. (2007). The influence of staffing characteristics on quality of care in nursing homes. Health
Services Research, 42(5), 1822–1847.
2007 Goldman, L. E., Vittinghoff, E., & Dudley, R. A. (2007). Quality of care in hospitals with a high percent of Medicaid
patients. Medical Care, 45(6), 579–583.
2007 Standing, M. (2007). Clinical decision-making skills on the developmental journal from student to registered nurse: A
longitudinal inquiry. Journal of Advanced Nursing, 60(3), 257–269.
2006 Mor, V. (2006). Defining and measuring quality outcomes in long term care. Journal of the American Medical Directors
Association, 7(8), 532–538.
2006 Rubin, F. H., Williams, J. T., Lescisin, D. A., Mook, W. J., Hassan, S., & Innouye, S. K. (2006). Replicating the Hospital
Elder Life Program in a community hospital and demonstrating effectiveness using quality improvement methodology.
Journal of the American Geriatrics Society, 54(6), 969–974.
2005 Donaldson, N., Brown, D. S., Aydin, C. E., Bolton, M. L., & Rutledge, D. N. (2005). Leveraging nurse-related
dashboard benchmarks to expedite performance improvement and document excellence. Journal of Nursing
Administration, 35(4), 163–172.
2005 Kleinpell, R., & Gawlinski, A. (2005). Assessing outcomes in advanced practice nursing practice: The use of quality
indicators and evidence-based practice. AACN Clinical Issues, 16(1), 43–57.
2005 Wilson, I. B., Landon, B. E., Hirschhorn, L. R., McInnes, K., Ding, L., Marsden, P. V., et al. (2005). Quality of HIV care
provided by nurse practitioners, physician assistants, and physicians. Annals of Internal Medicine, 143(10), 729–736.

organizational hierarchy, decision-making processes, sample of another structure providing the same pro-
oasis-ebl|Rsalles|1476144234

distribution of power, financial management, and
administrative decision-making processes. Nurse
researchers investigating the influence of structural
variables on quality of care and outcomes have studied
factors such as nurse staffing, nursing education,
nursing work environment, hospital characteristics,
and organization of care delivery (Table 13-2).
The second step is to evaluate the impact of various
structure elements on the process of care and on out-
comes. This evaluation requires comparing different
structures that provide the same processes of care. In
evaluating structures, the unit of measure is the struc-
ture. The evaluation requires access to a sufficiently
large sample of “like” structures with similar processes
and outcomes, which can then be compared with a
cesses and examining the same outcomes. For example,
in your research you might want to compare various
structures providing primary health care, such as the
private physician office, the health maintenance orga-
nization (HMO), the rural health clinic, the community-
oriented primary care clinic, and the nurse-managed
center. You might examine surgical care provided
within the structures of a private outpatient surgical
clinic, a private hospital, a county hospital, and a teach-
ing hospital associated with a health science center.
Within each of these examples, the focus of your study
would be the impact of structure on processes and
outcomes of care. Table 13-2 lists some current out-
comes studies that examined the impact of structure
of care and process of care on patient outcomes.
300 UNIT TWO  The Research Process
The federal government requires nursing homes,
home healthcare agencies, and hospitals to collect and
report specifically measured quality variables to the
government. The mandate was established because of
considerable variation in the quality of care in these
structures. Various government agencies analyze the
quality of these structures so that they can adequately
oversee the quality of care provided to the American
public. These data are made available to the general
public so that individuals can make their own deter-
mination of the quality of care provided by various
nursing homes, home healthcare agencies, or hospi-
tals. Researchers can also access these data for studies
of the quality of various structures. To access these
data on the Internet, you can search using the phrases
nursing home compare, home health compare, and
hospital compare. In addition to being able to select a
specific hospital, nursing home, or home healthcare
agency, you can access considerable general informa-
tion about quality related to these structures of health
care. You can also do a search for Magnet hospitals
on the American Nurses Credentialing Center website
(http://www.nursecredentialing.org/Magnet/Finda
MagnetFacility.aspx/) to check the status of a particu-
lar hospital regarding its recognition for excellence in
nursing care.
Federal Government
Involvement in
Outcomes Research
Agency for Healthcare Research and Quality
Nurses participated in the initial federal involvement
in studying the quality of health care. In 1959, two
National Institutes of Health (NIH) study sections,
the Hospital and Medical Facilities Study Section and
the Nursing Study Section, met to discuss concerns
about the adequacy and appropriateness of medical
care, patient care, and hospital and medical facilities.
As a result of their dialogue, a Health Services
Research Study Section was initiated. This study
section eventually became the Agency for Health Ser-
vices Research (AHSR), and subsequently the Agency
for Health Care Policy and Research (AHCPR). With
a growing budget and strong political support, propo-
nents of the AHCPR were becoming a powerful
force. They insisted on a change in health care
because of the demand for healthcare reform that
existed throughout the government and among the
public. A reauthorization act changed the name of the
AHCPR to the Agency for Healthcare Research and
Quality (AHRQ). The AHRQ is designated as a sci-
entific research agency. The term policy was removed
from the agency name to avoid the perception that
the agency determined federal healthcare policies
and regulations. The word quality was added to the
agency’s name, establishing the AHRQ as the lead
federal agency on quality of care research, with a
new responsibility to coordinate all federal quality
improvement efforts and health services research.
The new legislation eliminated the requirement that
the AHRQ develop clinical practice guidelines.
However, the AHRQ (2011) still supports these
efforts through EBP centers and the dissemination of
evidence-based guidelines through its National
Guideline Clearinghouse (see Chapter 19 for a more
detailed discussion of EBP guidelines).
The AHRQ, as a part of the U.S. Department of
Health and Human Services (DHHS), supports
research designed to improve the outcomes and quality
of health care, reduce its costs, address patient safety
and medical errors, and broaden access to effective
services. The AHRQ website (at http://www.ahrq.gov/)
is a valuable source of information about outcomes
research, funding opportunities, and results of recently
completed research, including nursing research. In
2010, AHRQ awarded $25 million in funding to
support efforts by states and health systems to imple-
ment and evaluate patient safety approaches and
medical liability reform models. In addition, AHRQ
invested $17 million to expand projects to help prevent
healthcare-associated infections (HAIs), the most
common complication of hospital care. The AHRQ
initiated several major research efforts to examine
medical outcomes and improve quality of care. One
of the latest, which is described in the next section, is
their comparative effectiveness research.
American Recovery and Reinvestment Act
Funding from the American Recovery and Reinvest-
ment Act (Recovery Act), signed into law in February
2009, allowed AHRQ to expand its work in support
of comparative effectiveness research, including
enhancing the Effective Health Care Program. A total
of $473 million was designated for funding patient-
centered outcomes research (AHRQ, 2010). This
AHRQ program provides patients, clinicians, and
others with evidence-based information to make
informed decisions about health care, through activi-
ties such as comparative effectiveness reviews con-
ducted through AHRQ’s Evidence-based Practice
Center (EPC) (see Chapter 19). The AHRQ has a
broad research portfolio that touches on nearly every
aspect of health care, including:
 CHAPTER 13  Outcomes Research 301

• Clinical practice
• Outcomes and effectiveness of care
• Evidence-based practice
• Primary care and care for priority populations
• Healthcare quality
• Patient safety/medical errors
• Organization and delivery of care and use of health-
care resources
• Healthcare costs and financing
• Health information technology
• Knowledge transfer
The United States is not the only country demanding
improvements in quality of care and reductions in
healthcare costs. Many countries are experiencing
similar concerns and addressing them in relation to
their particular government structures. Thus, the move-
ment into outcomes research and the approaches
described in this chapter are a worldwide phenomenon.
Medical Outcomes Study
The Medical Outcomes Study (MOS) was the first
large-scale study in the United States to examine
factors influencing patient outcomes. The study was
designed to identify elements of physician care associ-
ated with favorable patient outcomes. Figure 13-2
shows the conceptual framework for the MOS. The
following describes the MOS:
“The Medical Outcomes Study was designed to (1)
determine whether variations in patient outcomes
are explained by differences in system of care, clini-
cian specialty, and clinicians’ technical and interper-
sonal styles and (2) develop more practical tools for
the routine monitoring of patient outcomes in medical
practice. Outcomes included clinical end points; phys-
ical, social, and role functioning in everyday living;
patients’ perceptions of their general health and well-
being; and satisfaction with treatment. Populations of
clinicians (n = 523) were randomly sampled from dif-
ferent healthcare settings in Boston, MA; Chicago, IL;
and Los Angeles, CA. In the cross-sectional study,
adult patients (n = 22,462) evaluated their health
status and treatment. A sample of these patients (n =
2349) with diabetes, hypertension, coronary heart
disease, and/or depression were selected for the lon-
gitudinal study. Their hospitalizations and other treat-
ments were monitored and they periodically reported
outcomes of care. At the beginning and end of the
longitudinal study, Medical Outcomes Study staff per-
formed physical examinations and laboratory tests.
Results [were] reported serially, primarily in the
[Journal of the American Medical Association].” (Tarlov
et al., 1989, p. 925)

STRUCTURE OF CARE PROCESS OF CARE OUTCOMES

System characteristics Technical style Clinical end points
• Organization • Visits • Symptoms and signs
• Specialty mix • Medications • Laboratory values
• Financial incentives • Referrals • Death
• Workload • Test ordering
• Access/convenience • Hospitalizations
• Expenditures Functional status
• Continuity of care • Physical
Provider characteristics • Coordination • Mental
• Age • Social
• Gender • Role
• Specialty training Interpersonal style
• Economic incentives • Interpersonal manner Figure 13-2  Conceptual framework of the
• Patient participation General well-being
• Beliefs/attitudes • Health perceptions Medical Outcomes Study.
• Preferences • Counseling
• Communication level • Energy/fatigue
• Job satisfaction • Pain
• Life satisfaction
Patient characteristics
• Age Satisfaction with care
• Gender • Access
• Diagnosis/condition • Convenience
• Severity • Financial coverage
• Comorbid conditions • Quality
• Health habits • General
• Beliefs/attitudes
• Preferences
302 UNIT TWO  The Research Process
MOS failed to control for the effects of nursing
interventions, staffing patterns, and nursing practice
delivery models on medical outcomes. Coordination
of care, counseling, and referral activities, which are
more commonly performed by nurses than physicians,
were inappropriately considered in the MOS to be
components of medical practice. Kelly, Huber,
Johnson, McCloskey, and Maas (1994) suggested
modifications to the MOS framework that would rep-
resent the collaboration among physicians, nurses, and
allied health practitioners and the influence of their
interactions on patient outcomes. These researchers
also suggested adding the domain of societal out-
comes to include such outcome variables as cost. They
noted that “the MOS outcomes framework incorpo-
rated areas in which nursing science contributed to
health and medical care effectiveness. It also includes
structure, process, and outcome variables in which
nursing practice overlaps with that of other health
professionals” (p. 213). Kelly et al. (1994) further
observed that “client outcome categories of the MOS
framework that go beyond the scope of physician
treatment and intervention alone include functional
status, general well-being, and satisfaction with care”
(p. 213). A review of the state of the science on
nursing-sensitive outcomes published in 2011 con-
firmed the relevance of these outcomes to nursing
practice and suggested several more, including self-
care; therapeutic self-care, defined as patients’ ability
to manage their disease and its treatment; symptom
control; psychosocial functioning; healthcare utiliza-
tion; and mortality (Doran, 2011).
Origins of Outcomes/
Performance Monitoring
Florence Nightingale has been credited as being the
first nurse to collect data in order to identify nursing’s
contribution to quality care and to conduct research
into patient outcomes (Magnello, 2010; Montalvo,
2007). However, efforts to systematically collect data
to assess outcomes in more modern times did not gain
widespread attention in the United States until the late
1970s. At that time, concerns about quality of care
prompted the development of the “Universal Minimum
Health Data Set,” which was followed shortly there-
after by the Uniform Hospital Discharge Data Set
(Kleib, Sales, Doran, Mallette, & White, 2011). These
data sets facilitated consistency in data collection
among healthcare organizations by prescribing the
data elements to be gathered. The aggregated data
were then used to perform the assessment of quality
of care in hospitals and provide information on patients
discharged from hospitals.
Over time other countries developed similar data
sets. In Canada, “Standards for Management Informa-
tion Systems” (MIS) were developed in the 1980s.
Upon the establishment of the Canadian Institute for
Health Information (CIHI) in 1994, the MIS became
a set of national standards used to collect and report
financial and statistical data from health service orga-
nizations’ daily operations (Canadian Institute for
Health Information, 2012). Simultaneously, CIHI
implemented a national Discharge Abstract Database
(DAD), which has become a key resource. However,
as was the case with the Medical Outcomes Study
discussed earlier, these data sets did not include infor-
mation about nursing care delivered to patients in the
hospital (Kleib et al., 2011). Without this information,
the contribution of nursing care to patient, organiza-
tional, and system outcomes was rendered invisible.
This major gap in information was addressed by the
development of nursing minimum data sets in the
Unites States, Canada, and other countries around
the world.
Outcomes Research and
Nursing Practice
Outcome studies provide rich opportunities to build a
stronger scientific underpinning for nursing practice.
Nurse researchers have been actively involved in the
effort to examine the outcomes of patient care. Ideally,
we would like to understand the outcomes of nursing
practice within a one-to-one nurse/patient relation-
ship. However, in most cases, more than one nurse
cares for a patient. Therefore, the nursing effect is
shared. In addition, nurse managers and nurse admin-
istrators have control over the nursing staff and the
environment of nursing practice, and this control
affects the autonomy of the nurse to implement prac-
tice. Therefore, outcomes research must first focus on
how nursing care is organized rather than what nurses
do. Then, perhaps, we can begin to determine how
what nurses do influences patient outcomes (Lake,
2006). We know that nurses do have an effect on
patient outcomes. Kramer, Maguire, and Schmalen-
berg (2006) indicated that a growing body of evi-
dence supports a relationship between empowered
shared leadership/governance structure and the imple-
mentation of nursing practice. The importance of
autonomy in clinical nursing practice is being recog-
nized as critically important to positive patient out-
comes. It is important to identify autonomy-enabling
structures in the organizational structures of nursing
practice. One such structure revealed in a number of
nursing studies is the Magnet hospital designation,

which represents excellence in nursing in the agency
with this designation.
Nursing-Sensitive Patient Outcomes
A nursing-sensitive patient outcome (NSPO) is
“sensitive” because it is influenced by nursing care
decisions and actions. It may not be caused by nursing
but is associated with nursing. In various situations,
“nursing” might be the individual nurse, nurses as a
working group, the approach to nursing practice, the
nursing unit, or the institution that determines numbers
of nurses, salaries, educational levels of nurses, assign-
ments of nurses, workload of nurses, management of
nurses, and policies related to nurses and nursing prac-
tice. It might even include the architecture of the
nursing unit. In whatever form, nursing actions have
a role in the outcome, even though acts of other pro-
fessionals, organizational acts, and patient character-
istics and behaviors often are involved in the outcome.
What patient outcomes can you think of that might be
nursing-sensitive?
Nursing-sensitive outcomes have become an issue
because of national concerns related to the quality of
care. The demand for professional accountability
regarding patient outcomes dictates that nurses be able
to identify and document outcomes influenced by
nursing care. Efforts to study nursing-sensitive out-
comes were initiated by the American Nurses Associa-
tion (ANA). In 1994, the ANA, in collaboration with
the American Academy of Nursing Expert Panel on
Quality Health Care, launched a plan to identify indi-
cators of quality nursing practice and to collect and
analyze data using these indicators throughout the
United States (Mitchell, Ferketich, & Jennings, 1998).
The goal was to identify and/or develop nursing-
sensitive quality measures. Donabedian’s theory was
used as the framework for the project. Together, these
indicators were referred to as the ANA Nursing Care
Report Card, which could facilitate benchmarking,
or setting a desired standard that would allow com-
parisons of hospitals in terms of their nursing care
quality.
No one knew empirically what indicators were sen-
sitive to nursing care provided to patients or what the
relationships were between nursing inputs and patient
outcomes. Every hospital had a different way of mea-
suring the indicators that the ANA had selected. Per-
suading them to change to a standardized measure of
the indicators for consistency among hospitals was a
major endeavor (Jennings, Loan, DePaul, Brosch, &
Hildreth, 2001; Rowell, 2001). Multiple pilot studies
were conducted as nurse researchers and cooperating
hospitals put in place the mechanisms required for
data collection. These pilot studies identified multiple
CHAPTER 13  Outcomes Research 303
problems that had to be resolved before the project
could go forward. Researchers learned that not only
must the indicators be measured consistently but data
collection must also be standardized. As studies con-
tinued, indicators were amplified and continue to be
tested.
The ANA proposed that all hospitals collect and
report on the nursing-sensitive quality indicators. To
encourage researchers to collect these indicators, the
ANA accredited organizations and the federal gov-
ernment helped by sharing the data with key groups.
The ANA also encouraged state nurses’ associations
to lobby state legislatures to include the nursing-
sensitive quality indicators in regulations or state law.
In 1998, the ANA provided funding to develop a
national database to house data collected using
nursing-sensitive quality indicators. This became the
National Database of Nursing Quality Indicators
(NDNQI). Currently, NDNQI has more than 1500
participating organizations. Participation in NDNQI
meets requirements for the Magnet Recognition
Program®, and 20% of database members participate
for that reason. The remaining 80% of the members
participate voluntarily to support their evaluation and
improvement of nursing care quality and outcomes
(Montalvo, 2007).
Detailed guidelines for data collection, including
definitions and decision guides, are provided by
NDNQI (2010). Healthcare organizations submit data
electronically via the internet. Statistical methods such
as hierarchical mixed models are used to examine the
correlation between the nursing workforce character-
istics and outcomes (Montalvo, 2007). Quarterly and
annual reports of structure, process, and outcome indi-
cators are available 6 weeks after the close of each
reporting period. The database is housed at the
Midwest Research Institute (MRI), Kansas City, Mis-
souri, and is managed by MRI in partnership with the
University of Kansas School of Nursing (Alexander,
2007). The NDNQI nursing sensitive indicators are as
follows:
1. Patient falls/injury falls
2. Pressure ulcers (hospital acquired, unit acquired)
3. Physical/sexual assault
4. Pain assessment/intervention/reassessment cycle
5. Peripheral intravenous (IV) infiltration
6. Physical restraints
7. Registered Nurse (RN) survey: Job satisfaction;
Practice Environment Scale
8. Healthcare associated infections:
(a) Catheter-associated urinary tract infections
(UTI)
(b) Central line–associated bloodstream infection
(c) Ventilator-associated pneumonia
304 UNIT TWO  The Research Process
9. Staff mix (RN; Licensed Practical Nurse [LPN]/
Licensed Vocational Nurse [LVN]/Unlicensed
Assistive Personnel [UAP])
10. Nursing care hours provided per patient day
11. Nurse turnover (total, adapted National Quality
Forum voluntary, Magnet controllable)
12. RN education/certification
13. RN survey:
(a) Practice Environment Scales
(b) Job Satisfaction Scales (option)
(c) Job Satisfaction Scales–Short Form (option)
Other organizations currently involved in efforts to
study nursing-sensitive outcomes include the National
Quality Forum (NQF), Collaborative Alliance for
Nursing Outcomes California Database, Veterans
Affairs Nursing Outcomes Database, the Center for
Medicare and Medicaid Services’ (CMS) Hospital
Quality Initiative, the American Hospital Association,
the Federation of American Hospitals, The Joint Com-
mission, and the AHRQ. A description of the Collab-
orative Alliance for Nursing Outcomes California
Database project follows.
The Collaborative Alliance for Nursing Outcomes
California Database Project
California Nursing Outcomes Coalition (CalNOC)
was a statewide nursing quality report card pilot
project launched in 1996. CalNOC was a joint venture
of the ANA/California and the Association of Califor-
nia Nurse Leaders that was funded by ANA. Member-
ship is voluntary and is composed of approximately
300 hospitals from the United States, with pilot work
in Sweden, England, and Australia. As its membership
grew nationally, CalNOC was renamed the Collabora-
tive Alliance for Nursing Outcomes (CALNOC,
2010). It is a not-for-profit corporation, and member
hospitals pay a size-based annual data management
fee to participate and access CALNOC industry Web-
based benchmarking reporting system.
Hospital-generated unit-level acute nurse staffing
and workforce characteristics and processes of care
data as well as key endorsed nursing-sensitive out-
comes measures are submitted electronically via the
Web. In addition, the CALNOC database includes
unique measures such as its Medication Administra-
tion Accuracy metric, which helps in tracking medica-
tion errors. CALNOC data are stratified by unit type
and hospital characteristics, and reports can be aggre-
gated to division, hospital, and system/group/geo-
graphical levels. The following list of nursing-sensitive
indicators has been extracted from Nurse Minimum
Data Sets and CALNOC:
Structure indicators:
1. Hours per patient-day (RN, LPN, UAP)
2. Skill mix
3. Nurse/patient ratios
4. Percentage of contracted staff utilization (hours)
5. Staff voluntary turnover rate
6. Workload intensity (admissions, discharges,
transfers)
7. Sitter hours as percentage of total care hours
8. RN characteristics (education, certification, years
of experience)
Process indicators:
1. Risk assessment for pressure ulcers (Braden
Scale)
2. Time since last risk assessment
3. Risk score (pressure ulcers)
4. Risk status (pressure ulcers, falls)
5. Prevention protocols in place (pressure ulcers,
falls)
6. Medication administration accuracy: observed
prevalence of 6 key safe practices
7. Peripherally inserted central catheter (PICC):
line insertion practices (who inserted, where,
presence of a dedicated team)
8. Restraint use: type and clinical justification
Outcome measures:
1. Community-acquired pressure ulcer prevalence
2. Hospital-acquired pressure ulcer prevalence
stages 1-4
3. Patient fall rate per 1000 patient-days and con-
sequences (injury fall rate)
4. Restraint prevalence
5. Central line–associated bloodstream infections
(CABSIs) in PICC lines
6. Medication administration error rates
For further information on these outcome initia-
tives, you can review Doran, Mildon, and Clarke’s
(2011) knowledge synthesis of the state of science on
nursing outcomes measurement and international
nursing report card initiatives. The knowledge synthe-
sis was a review of nursing-sensitive outcome and
report card initiatives in the United States, Canada, the
United Kingdom, and Belgium.
National Quality Forum
The National Quality Forum was created in 1999 as a
national standard-setting organization for healthcare
performance measures (NQF, 2011). The NQF portfo-
lio of voluntary consensus standards includes perfor-
mance measures, serious reportable events, and
preferred practices (i.e., safe practices). A complete
list of measures included in the NQF portfolio can
be found online (www.qualityforum.org/Measures_
List.aspx/). Approximately one third of the measures

in NQF’s portfolio are measures of patient outcomes.
Examples are mortality, readmissions, health function-
ing, depression, and experience of care (NQF, 2011).
The NQF includes in their performance measure-
ment portfolio several nursing-sensitive measures.
Those that were submitted by the ANA under the
NDNQI include the following:
• Nursing staff skill mix
• Nursing hours per patient-day
• Catheter-associated urinary tract infection (UTI)
rate
• Central line–associated bloodstream infection rate
• Fall/injury rates
• Hospital/unit-acquired pressure ulcer rates
• Nurse turnover rate
• RN practice environment scale
• Ventilator-associated pneumonia rate
These indicators are the first nationally standard-
ized performance measures of nursing-sensitive out-
comes in acute care hospitals, and they are designed
to assess healthcare quality, patient safety, and a pro-
fessional and safe work environment. Although most
measures currently used focus on the failure to meet
the expected standards, the NQF believes that quality
is as much about influencing positive outcomes as
about avoiding negative outcomes. Thus, the NQF is
interested in developing measures that reflect the posi-
tive effects of nursing care. Priority areas for indica-
tors include assessment, patient education, and care
coordination (Naylor, 2007).
Oncology Nursing Society
The Oncology Nursing Society (ONS) is a profes-
sional organization of more than 35,000 RNs and
other healthcare providers dedicated to excellence in
patient care, education, research, and administration
in oncology nursing (ONS, 2012). The ONS has
taken a leadership role among specialty nursing orga-
nizations in developing an EBP resource area on its
website (http://www.ons.org/ClinicalResources/). The
site provides nurses with a guide to identify, critically
appraise, and use evidence to solve clinical problems.
The ONS website also assists nurses—especially
advanced practice nurses—who are helping others
develop EBP protocols. The outcomes resource area
helps nurses to achieve desired outcomes for people
with cancer by providing outcome measures, resource
cards, and evidence tables.
Advanced Practice Nursing
Outcomes Research
Studies of outcomes of advanced practice nurses
(APNs) are now appearing in the literature. APNs are
CHAPTER 13  Outcomes Research 305
RNs educationally prepared at the master’s or doctoral
level. These practitioners have expertise in a particular
area of clinical practice and provide direct patient care.
The ANA recognizes four types of APNs: certified
registered nurse anesthetists (CRNAs), certified nurse
midwives (CNMs), clinical nurse specialists (CNSs),
and nurse practitioners (NPs). Studying APNs requires
determining what happens during the process of APN
care. This care involves a set of activities within,
among, and between practitioners and patients and
includes both technical and interpersonal elements.
This process of care is complex and somewhat mysteri-
ous. However, clearly describing what occurs during
this process of care is essential to developing a com-
prehensive understanding of how APNs affect out-
comes.Although researchers have provided descriptions
of APN care, considerable detailed work must still be
done to more thoroughly describe the activities and
interactions that occur between APNs and patients
during the process of care (Cunningham, 2004).
The next step is to establish the relationship
between APN interventions and outcomes. The out-
comes must be clearly defined and measurable or
observable. Outcomes may require risk adjustments
for factors that may confound the results, such as
comorbidity, stage of illness, severity of illness, and
demographic characteristics. Failure to use rigor in
measurement will limit your ability to interpret study
findings meaningfully. It is important for variables to
be measured using the same measurement methods
across studies so that results are more readily com-
pared. Understanding which outcomes are sensitive to
APN interventions is critical to building knowledge in
this area. We need a classification of outcomes of APN
practice. Ingersoll, McIntosh, and Williams (2000)
generated a beginning list of 27 relevant outcome indi-
cators of APN practice. The nine highest outcomes are
as follows:
• Satisfaction with care delivered
• Symptom resolution or reduction
• Perception of being well cared for
• Compliance with or adherence to treatment plan
• Knowledge of patients and families
• Trust of care provider
• Collaboration among care providers
• Frequency and type of procedures ordered
• Quality of life
Doran, Sidani, and DiPietro (2010) conducted a
systematic review of the empirical evidence on out-
comes of CNS practice. CNSs are licensed registered
professional nurses with graduate preparation demon-
strated in an earned master’s or doctorate degree
(National Association of Clinical Nurse Specialists
306 UNIT TWO  The Research Process
[NACNS], 2004). CNS-sensitive outcomes are those
that can be theoretically linked to the activities of
CNSs and observed in the three domains of CNS
practice: patient and patient care, nurse and nursing
practice, and organization and system (NACNS,
2004). Examples of CNS-sensitive outcomes are
disease-specific patient outcomes, physical and psy-
chosocial patient outcomes, nurse outcomes (job sat-
isfaction), and system outcomes (costs of care) (Doran
et al., 2010).
Methodologies for
Outcomes Studies
Outcomes research methodologies have been devel-
oped to link the care people receive to the results they
experience, thereby providing better ways to monitor
and improve the quality of care (Clancy & Eisenberg.
1998). This section describes some of the current
methodologies used in conducting outcomes research,
including sampling methods, research strategies or
designs, measurement processes, and statistical
approaches. These descriptions are not sufficient to
guide you in using the approaches described; rather
they provide a broad overview of the variety of meth-
odologies being used. This knowledge will help you
understand and critically appraise the methodologies
used in published outcomes studies. For additional
information, you can refer to the citations in each
section. Outcomes studies cross a variety of disci-
plines; thus, the emerging methodologies are being
enriched by a cross-pollination of ideas, some of
which are new to nursing research.
Samples and Sampling
The preferred methods of obtaining samples are
different in outcomes studies; random sampling is
Figure 13-3  Types of databases emanating from
patient care encounters.
Large clinical databases
Computerized medical records
Disease or organ-specific databases
not considered desirable and is seldom used. Hetero-
geneous, rather than homogeneous, samples are
obtained. Rather than using sampling criteria—which
restrict subjects included in the study to decrease pos-
sible biases and also reduce the variance and increase
the possibility of identifying a statistically significant
difference—outcomes researchers seek large hetero-
geneous samples that reflect, as much as possible, all
patients who would be receiving care in the real world.
For example, samples need to include patients with
various comorbidities and patients with varying levels
of health status. In addition, persons should be identi-
fied who do not receive treatment for their condition.
Devising ways to evaluate the representativeness of
such samples is problematic. As noted in Chapter 15
of this text, for a sample to be representative it must
be as much like the target population as possible,
particularly in relation to the variables being studied.
Because the target population in outcomes research is
often heterogeneous, there are a large number of vari-
ables for which sample representativeness needs to be
determined. Another challenge in outcomes research
is developing strategies to locate untreated individuals
and include them in follow-up studies. To address
some of these challenges, outcomes researchers have
used large databases as sample sources in observa-
tional research designs.
Large Databases as Sample Sources
One source of samples for outcomes studies is large
databases. Two broad categories of databases emerge
from patient care encounters: clinical databases and
administrative databases, as illustrated by Figure 13-3
(Waltz, Strickland, & Lenz, 2010). Clinical databases
are created by providers such as hospitals, HMOs, and
healthcare professionals. The clinical data are gener-
ated either as a result of routine documentation of care
or in relation to a research protocol. Some databases
Patient care
encounter
Clinical data Administrative data
Large administrative databases
Insurance claims databases
Tumor or disease registries
Vital statistics databases
 CHAPTER 13  Outcomes Research 307

are data registries that have been developed to gather
data related to a particular disease, such as heart
disease or cancer (Lee & Goldman, 1989). With a
clinical database, you can link observations made by
many practitioners over long periods. Links can be
made between the process of care and outcomes
(Mitchell et al., 1994; Moses, 1995).
Administrative databases are created by insur-
ance companies, government agencies, and others not
directly involved in providing patient care. Adminis-
trative databases have standardized sets of data for
enormous numbers of patients and providers (Deyo
et al., 1994; McDonald & Hui, 1991). An example is
the Medicare database managed by the Centers for
Medicare & Medicaid Services. These databases can
be used to determine the incidence or prevalence of
disease, geographical variations in medical care utili-
zation, characteristics of medical care, outcomes of
care, and complementarity with clinical trials. Wray
et al. (1995) cautioned, however, that analyses should
be restricted to outcomes specific to a particular sub-
group of patients rather than to one adverse outcome
of all disease states. Examples of large database indi-
cators used to assess the quality of care are provided
in Table 13-3.
Large databases are used in outcomes studies to
examine patient care outcomes. The outcomes that can
be examined are limited to those recorded in the data-
base and thus tend to be general. Existing databases
can be used for analyses such as (1) assessing nursing
care delivery models; (2) varying nursing practices; or
(3) evaluating patients’ risk of hospital-acquired

TABLE 13-3  Examples of Large Database Indicators Used to Monitor Nursing Structural, Process, and
Outcome Indicators
Type of Indicator Indicator Source
Structural Nursing (e.g., RN, LPN, UAP) National Database Nursing Quality Indicators (NDNQI, 2010)
hours per patient day Collaborative Alliance for Nursing Outcomes (CALNOC, 2010)
National Quality Forum (NQF, 2011)
Staff mix (RN, LPN, LVN, UAP) NDNQI (2010)
CALNOC (2010)
NQF (2011)
Nurse turnover NDNQI (2010)
CALNOC (2010)
NQF (2011)
RN Practice Environment NDNQI (2010)
NQF (2011)
Process Risk assessment for pressure ulcers CALNOC (2010)
Physical restraints NDNQI (2010)
CALNOC (2010)
Prevention protocols in place CALNOC (2010)
B-NMDS (Belgian nursing minimum data set) (Van den Heede, et al.,
2009; Sermeus, Delesie, Van den Heede, Diya, & Lesaffre, 2008)
Medication administration accuracy CALNOC
Outcome Patient falls/injury falls NDNQI (2010)
CALNOC (2010)
NQF (2011)
Catheter-associated urinary tract NDNQI (2010)
infection rate NQF (2011)
Hospital-acquired pressure ulcer NDNQI (2010)
CALNOC (2010)
NQF (2011)
Central line–associated blood- NDNQI (2010)
stream infection rate CALNOC (2010)
NQF (2011)

LPN, licensed practical nurse; LVN, licensed vocational nurse; RN, registered nurse; UAP, unlicensed assistive personnel.
308 UNIT TWO  The Research Process
infection, hospital-acquired pressure ulcer, or falls. To
examine these questions, nurses must develop the sta-
tistical and methodological skills needed for working
with large databases. Large databases contain patient
and institutional information from huge numbers of
patients. They exist in computer-readable form, require
special statistical methods and computer techniques,
and can be used by researchers who were not involved
in the creation of the database.
Initiatives such as CALNOC (2010), NDNQI
(2010; Montalvo, 2007), and the Veterans Affairs
Nursing Outcome Database (Alexander, 2007) are
making nursing data more accessible for large data-
base research. The following nursing classification
schemes have been used in national databases:
• The North American Nursing Diagnosis Associa-
tion (NANDA) Classification
• The Omaha System: Applications for Community
Health Nursing Classification
• The Home Health Care Classification
• The Nursing Interventions Classification (NIC)
• The Nursing Outcomes Classification (NOC)
Research Strategies for Outcomes Studies
Outcomes research programs usually consist of studies
with a mix of strategies carried out sequentially.
Although these strategies could be referred to as
designs, for some the term design as used in Chapters
10 and 11 is inconsistent with the strategies used in
outcomes studies. Research strategies for outcomes
studies have emerged from a variety of disciplines,
and innovative new strategies continue to appear in
the literature. Strategies for outcomes studies tend to
employ fewer controls than traditional research
designs and cannot be as easily categorized. The
research strategies described in this section are only a
sampling from the outcomes research literature and
include consensus knowledge building, practice
pattern profiling, prospective cohort studies, retro-
spective cohort studies, population-based studies, geo-
graphical analyses, economic studies, and ethical
studies.
Consensus Knowledge Building
Consensus knowledge building is usually performed
by a multidisciplinary group representing a variety of
constituencies. Initially, the group conducts an exten-
sive international search of the literature on the topic
of concern, including unpublished studies, studies in
progress, dissertations, and theses. Several separate
reviews may be performed, focusing on specific ques-
tions about the outcomes of care, diagnosis, preven-
tion, or prognosis. The results are dispersed to
researchers and clinical experts in the field, who are
asked to carefully examine the material and then par-
ticipate in a consensus conference. The consensus
conference yields clinical guidelines, which are pub-
lished and widely distributed to clinicians. The clinical
guidelines are also used as practice norms to study
process and outcomes in that field. Gaps in the knowl-
edge base are identified and research priorities deter-
mined by the consensus group.
Preliminary steps in this process might include con-
ducting extensive integrative reviews and seeking
consensus from a multidisciplinary research team and
locally available clinicians. A review could be accom-
plished by establishing a website and conducting dia-
logue with experts via the internet. The review could
be published in Sigma Theta Tau’s online journal,
Knowledge Synthesis in Nursing, and then dialogue
related to the review could be conducted over the
internet. The Delphi method has also been used to seek
consensus (Tork, Dassen, & Lohrmann, 2008).
Described as a group process to obtain judgments
from a panel of experts, the Delphi method involves
a series of questionnaires or rounds until a predeter-
mined level of group consensus is reached (see Chapter
17). The experts are questioned individually, and a
summary of the individual judgments is distributed to
panel members, with subsequent questionnaires to
influence the panel members through group feedback
(Keeney, Hasson, & McKenna, 2011).
Practice Pattern Profiling
Practice pattern profiling is an epidemiological tech-
nique that focuses on patterns of care rather than indi-
vidual occurrences of care. Researchers use large
database analysis to identify a provider’s pattern of
practice and compare it with that of similar providers
or with an accepted standard of practice. The tech-
nique has been used to determine overutilization and
underutilization of services, to examine costs associ-
ated with a particular provider’s care, to uncover prob-
lems related to efficiency and quality of care, and to
assess provider performance. The provider being pro-
filed could be an individual practitioner, a group of
practitioners, or a healthcare organization such as a
hospital or an HMO. The provider’s pattern is
expressed as a rate aggregated over time for a defined
population of patients under the provider’s care. For
example, using data on the universe of deliveries in
Florida and New York over a 15-year period, Epstein
and Nicholson (2009) examined why treatment styles
differ among obstetricians at a point in time and why
styles change over time. They found that variation in
cesarean section rates among physicians within a

market is about twice as large as variation between
markets.
Profiling can be used when the data contain hierar-
chical groupings: Patients could be grouped by nurse,
nurses by unit, and units by larger organizations. The
analysis uses regression equations to examine the rela-
tionship of an outcome to the characteristics of the
various groupings. To be effective, the analysis must
include data on the different sources of variability that
might contribute to a given outcome.
The structure of the analysis reflects the structure
of the data. For example, patient characteristics could
be data on disease severity, comorbidity, emergent or
developing status, behavioral characteristics, socio-
economic status, and demographics. Nurse character-
istics might consist of level of education, specialty
status, years of practice, age, gender, and certifica-
tions. Unit characteristics could comprise number of
beds, nursing management style used on the unit, ratio
of patients to nurses, and the proportion of staff who
are RNs (Doran et al., 2006). Profiles are designed to
generate some type of action, such as to inform the
provider that his or her rates of patient improvement
are high or too low in comparison with the norm. By
examining aggregate patterns of practice, profiling can
be used to compare the care provided by different
organizations or received by different populations of
patients. Critical pathways or care maps can then be
used to determine the proportion of patients who
diverged from the pathway for a particular nurse,
group of nurses, or group of nursing units. Profiling
can be used to improve quality, assess provider per-
formance, and review utilization patterns.
Profiling does not address methods of improving
outcomes, although this process can identify problem
areas. It can be used to determine how performance
should be changed to improve outcomes and who
should make those changes. Profiling can also identify
outliers, allowing more detailed examination of these
individuals and their practice. To date, most of the
profiling research has been about medical practice.
The development of nursing-sensitive structure,
process, and outcome indicators that are assessed and
benchmarked over time, as is being done with NDNQI
(Montalvo, 2007; NDNQI, 2010) will enable profiling
of nursing practice.
Prospective Cohort Studies
A prospective cohort study is an epidemiological
study in which the researcher identifies a group of
people who are at risk for experiencing a particular
event. Sample sizes for these studies often must be
very large, particularly if only a small portion of the
CHAPTER 13  Outcomes Research 309
at-risk group will experience the event. The entire
group is followed over time to determine the point at
which the event occurs, the variables associated with
the event, and the outcomes for those who experienced
the event in comparison with those who did not.
The Harvard Nurses’ Health Study is an example
of a prospective cohort study. This study recruited
100,000 nurses to determine the long-term conse-
quences of the use of birth control pills. Every 2 years
or more often, nurses complete a questionnaire about
their health and health behaviors. The study has now
been in progress for more than 20 years. Multiple
studies reported in the literature have used the large
data set yielded by the study. The following summary
describes a prospective cohort study on smoking and
risk of psoriasis in women, using the Nurses’ Health
Study II, a second study using a younger population
than the Harvard study (Setty, Curhan, & Choi, 2007).
These researchers were able to obtain an extremely
large heterogeneous sample for their study by using
data from the Nurses’ Health Study.
Background: Psoriasis is a common, chronic, inflam-
matory skin disorder. Smoking may increase the risk
of psoriasis.
Methods: Over a 14-year time period from 1991 to
2005, the relation between smoking status, duration,
intensity, cessation, exposure to second-hand smoke,
and incident of psoriasis was prospectively examined
in 78,532 women from the Nurses Health Study II.
The primary outcome was incident, self-reported,
physician-diagnosed psoriasis.
Results: Eight hundred eighty-seven incident cases
of psoriasis were documented. The multivariate rela-
tive risk (RR) of psoriasis was 1.78 (95% confidence
interval [CI], 1.46 to 2.16) for current smokers and
1.37 (95% CI, 1.17 to 1.59) for past smokers in
comparison with persons who had never smoked.
The multivariate RR of psoriasis was 1.60 (95% CI,
1.31 to 1.97) for those who had smoked 11 to 20
pack-years and 2.05 (95% CI, 1.66 to 2.53) for those
who had smoked 21 or more pack-years in compari-
son with nonsmokers. The multivariate RR of psoria-
sis was 1.61 (95% CI, 1.30 to 2.00) for those who
quit smoking less than 10 years ago, 1.31 (95% CI,
1.05 to 1.64) for those who had quit 10 to 19 years
ago, and 1.15 (95% CI, 0.88 to 1.51) for those who
had quit 20 or more years ago in comparison with
persons who had never smoked. An increased risk of
psoriasis was associated with prenatal and childhood
exposure to passive smoke.
 310 UNIT TWO  The Research Process
Conclusions: The prospective analysis suggests that
current and past smoking, and cumulative measures
of smoking, were associated with the incidence of
psoriasis. After 20 years of smoking cessation, the risk
of the incident psoriasis among ex-smokers decreases
nearly to that of persons who have never smoked.
(Setty et al., 2007, p. 953)
Retrospective Cohort Studies
A retrospective cohort study is an epidemiological
study in which the researcher identifies a group of
people who have experienced a particular event. This
is a common research technique used in the field of
epidemiology to study occupational exposure to
chemicals. Events of interest to nursing that could be
studied in this manner include a procedure, an episode
of care, a nursing intervention, and a diagnosis. Nurses
might use a retrospective cohort study to follow a
cohort of women who had received a mastectomy for
breast cancer or of patients in whom a urinary bladder
catheter was placed during and after surgery. The
cohort is evaluated after the event to determine the
occurrence of changes in health status, usually the
development of a particular disease or death. Nurses
might be interested in the pattern of recovery after an
event or, in the case of catheterization, the incidence
of bladder infections in the months after surgery.
On the basis of the study findings, epidemiologists
calculate the relative risk of the identified change in
health for the group. For example, if death were the
occurrence of interest, the expected number of deaths
would be determined. The observed number of deaths
divided by the expected number of deaths and multi-
plied by 100 yields a standardized mortality ratio
(SMR), which is regarded as a measure of the relative
risk of the studied group to die of a particular condi-
oasis-ebl|Rsalles|1476144236
tion. In nursing studies, patients might be followed
over time after discharge from a healthcare facility to
determine complication rates and the SMR (Swaen &
Meijers, 1988).
In retrospective studies, researchers commonly ask
patients to recall information relevant to their previous
health status. This information is often used to deter-
mine the amount of change occurring before and after
an intervention. Recall can easily be distorted, thereby
misleading researchers in determining outcomes.
Thus, recall should be used with caution. Herrmann
(1995) identified three sources of distortion in recall:
(1) the question posed to the subject may be conceived
or expressed incorrectly; (2) the recall process may be
in error; and (3) the research design used to measure
recall can result in the recall’s appearing to be differ-
ent from what actually occurred. Herrmann (1995, p.
AS90) also identified four bases of recall:
Direct recall: The subject “accesses the memory
without having to think or search memory,” result-
ing in correct information.
Indirect recall: The subject “accesses the memory
after thinking or searching memory,” resulting in
correct information.
Limited recall: “Access to the memory does not occur
but information that suggests the contents of the
memory is accessed,” resulting in an educated
guess.
No recall: “Neither the memory nor information rel-
evant to the memory may be accessed, resulting in
a wild guess.”
The following abstract developed by Rozen, Ugoni,
and Sheehan (2011), on their study of vaginal birth
after cesarean section, is presented as an example of
a retrospective cohort study:
“Background: Previous studies assessing the safety of
vaginal birth after caesarean section (VBAC) have
compared VBAC to elective repeat caesarean section
(ERCS), despite the fact that the risks posed by each
are considerably different. Explaining the complica-
tions of VBAC in a way that is meaningful to women
can be challenging, and thus a comparison to a similar
group of women who have also not undergone
previous vaginal delivery may be a more relevant
comparison.
Research Question: When counseling women
undergoing planned VBAC, should a comparison of
outcomes be made to women undergoing ERCS, or
is a comparison to other nulliparous women undergo-
ing vaginal birth a more valid comparison in terms of
risk outcomes?
Participants and Methods: A retrospective cohort
study was undertaken comprising a consecutive
cohort of 21,389 women who delivered, stratified by
Robson’s criteria into Robson groups 1-5. Those in
Robson groups 6-10 were not included. Demographic
data and maternal/neonatal outcomes were reviewed,
with main outcome measures comprising uterine
rupture, post-partum hemorrhage (PPH), 3rd/4th
degree tears, and neonatal morbidity.
Results: There was no increase in PPH, vaginal
tears, or neonatal complications in the VBAC group
when compared to Robson groups 1 and 2 (nullipa-
rous women in spontaneous or induced labour,
respectively). Uterine rupture rates were low in all
groups, with no correlation identified.

Discussion: The maternal and neonatal morbidity
associated with VBAC is comparable to primiparous
women undergoing a vaginal birth.
Conclusion: In demonstrating the low relative mor-
bidity in this comparison, these outcomes may aid in
counseling women faced with the choice of VBAC
versus ERCS.” (Rozen et al., 2011, p. 3)
Population-Based Studies
Population-based studies are also important in out-
comes research. Conditions must be studied in the
context of the patient’s community rather than of the
medical system. With this method, all cases of a condi-
tion occurring in the defined population are included,
rather than only patients treated at a particular health-
care facility, because the latter could introduce a selec-
tion bias. The researcher might make efforts to include
individuals with the condition who had not received
treatment.
Community-based norms of tests and survey instru-
ments obtained in this manner provide a clearer picture
of the range of values than the limited spectrum of
patients seen in specialty clinics. Estimates of instru-
ment sensitivity and specificity are more accurate.
This method enables researchers to understand the
natural history of a condition or of the long-term risks
and benefits of a particular intervention (Guess et al.,
1995). Bakker et al. (2011) conducted a study examin-
ing the differences in birth outcomes related to mater-
nal age. The following is an abstract of their study:
“Background: Previous studies have shown that birth
weight and preterm birth are strong predictors of
neonatal morbidity and mortality. Maternal age might
be a modifiable determinant of weight and gestational
age at birth. In most Western countries the age of
mothers having their first child is increasing due to
prolonged education, professional commitment,
delayed marriage, and other personal reasons. It has
been suggested that older maternal age is associated
with increased risks of pregnancy complications, such
as gestational hypertension or diabetes, preterm
delivery, fetal malformations, and fetal death.
Methods: This is a population-based prospective
cohort study with 8,568 mothers and their children
based in Rotterdam, Netherlands. Maternal age,
sociodemographic, lifestyle-related determinants,
and birth outcomes were obtained from question-
naires and hospital records. The main outcome
CHAPTER 13  Outcomes Research 311
measures were birth weight, preterm delivery,
small-for-gestational-age, and large-for gestational-
age babies. Multivariate linear and logistic regression
analyses were used to analyze study data.
Results: In this study, mothers aged 30-34.9 years
had no differences in risk of preterm delivery. Mothers
<20 years had the highest risk of delivering small-
for-gestational-age babies (OR 1.6, 95% CI: 1.1-2.5);
however, after adjustment for sociodemographic and
lifestyle-related determinants this increased risk was
not present. Mothers >40 years had the highest risk
of delivering large-for-gestational-age babies (OR 1.3,
95% CI: 0.8-2.4); however no associations of
maternal age with the risks of delivering large-
for-gestational-age babies could be explained by
sociodemographic and lifestyle-related determinants.
Conclusions: Younger mothers have an increased
risk of small-for-gestational age babies, whereas older
mothers have an increased risk of large-for-
gestational-age babies when compared with mothers
aged 30-34.9 years. Sociodemographic and lifestyle-
related determinants cannot entirely explain these
differences.” (Bakker et al., 2011, p. 500)
Geographical Analyses
Geographical analyses examine variations in health
status, health services, patterns of care, or patterns of
use by geographical area and are sometimes referred
to as small area analyses. Variations may be associ-
ated with sociodemographic, economic, medical, cul-
tural, or behavioral characteristics. Locality-specific
factors of a healthcare system, such as capacity, access,
and convenience, may play a role in explaining varia-
tions. The social setting, environment, living condi-
tions, and community may also be important factors.
The interactions between the characteristics of a
locality and of its inhabitants are complex. The char-
acteristics of the total community may transcend the
characteristics of individuals within the community
and may influence subgroup behavior. High educa-
tional levels in the community are commonly associ-
ated with greater access to information and
receptiveness to ideas from outside the community.
Regression analyses are commonly used to develop
models using all the risk factors and the characteristics
of the community. Results are often displayed through
the use of maps (Kieffer, Alexander, & Mor, 1992).
After the analysis, the researcher must determine
whether differences in rates are due to chance alone
and whether high rates are too high. From a more
312 UNIT TWO  The Research Process
theoretical perspective, the researcher must then
explain the geographical variation uncovered by the
analysis (Volinn, Diehr, Ciol, & Loeser, 1994).
Geographical information systems (GISs) can
provide an important tool for performing geographical
analyses. A GIS uses relational databases to facilitate
processing of spatial information. The software tools
in a GIS can be used for mapping, data summaries,
and analysis of spatial relationships. GISs have the
capability of modeling data flows so that the effect of
proposed changes in interventions applied to individu-
als or communities on outcomes can be modeled.
Dunn, Anderson, and Bierman (2009) conducted a
study of temporal and regional trends in intrauterine
device (IUD) insertion. Their study abstract is reprinted
here:
“Background: In Canada, intrauterine device (IUD) use
is low and declined between 1985 and 1995. This
study examines temporal and regional trends in IUD
insertion in Ontario, Canada, from 1996 to 2006.
Study Design: Physician billing data was used to
determine annual age-adjusted IUD insertion rates
for women aged 15-55 years and proportions inserted
by gynecologists and family physicians (FPs). Small
area variation statistics were used to analyze variation
in rates across the province.
Results: Annual insertion rates followed a U-shaped
distribution and were lowest in 2001 and highest in
2006. From 1996 to 2006, the proportion inserted by
FPs fell from 38.2% to 31.6% (p<0.001). In 2006,
women in regions with the highest rates were twice
as likely to have an IUD inserted as those in the
lowest-rate regions.
Conclusions: IUD insertion rate began to increase
in 2001, the year of introduction of levonorgestrel-
releasing intrauterine system. Regional variation in
rates suggests that access is not equal across the
province and that strategies to support FPs to insert
IUDs may be important to ensure sufficient access.”
(Dunn et al., 2009, p. 469)
Economic Studies
Many of the problems studied in health services
research address concerns related to the efficient use
of scarce resources and, thus, to economics. Health
economists are concerned with the costs and benefits
of alternative treatments or ways of identifying the
most efficient means of care. The economist’s defini-
tion of efficiency is the least expensive method of
achieving a desired end while obtaining the maximum
benefit from available resources. If available resources
must be shared with other programs or other types of
patients, an economic study can determine whether
changing the distribution of resources will increase
total benefit or welfare.
Economic evaluation is a “set of formal, quantita-
tive methods used to compare two or more treatments,
programs, or strategies with respect to their resource
use and their expected outcomes” (Guyatt, Rennie,
Meade, & Cook, 2008, p. 781). In 1993, the U.S.
Public Health Service established the Panel on Cost-
Effectiveness in Health and Medicine to address the
Public Health System’s growing need for evaluating
techniques that could guide decision-making pro-
cesses in an era of rising cost restrictions (Gold, 1996).
The Panel was charged with the development of rec-
ommendations for addressing the (1) poor quality of
many economic analyses; (2) lack of comparability
across cost-effectiveness analyses (CEAs) of different
interventions or types of illness; and (3) lack of data
on the costs and effects of interventions (Gold, 1996).
Cost-effectiveness analysis, rather than cost-benefit
analysis, was identified as the appropriate tool for
economic analysis for the purpose of public health
uses because of the former’s focus on health rather
than economic outcomes. The recommendations of
the Panel are summarized here.
CEAs compare different ways of accomplishing a
clinical goal, such as diagnosing a condition, treating
an illness, or providing a service. The alternative
approaches are compared in terms of costs and bene-
fits. The purpose is to identify the strategy that pro-
vides the most value for the money. There are always
tradeoffs between costs and benefits. When making
decisions for patient groups, clinicians need to con-
sider whether the benefits of providing treatment are
worth the healthcare costs. It is also important to indi-
cate whose perspective the analysis will consider, for
example, that of patients, providers, insurers, or the
broader society. The Panel on Cost-Effectiveness in
Health and Medicine assumed a societal perspective.
The Panel’s recommendations fell into the following
eight categories (Weinstein, Siegel, Gold, Kamlet, &
Russell, 1996):
1. The nature and limits of CEA and of the reference
case for the analysis.
2. Components (variables) belonging in the numera-
tor and denominator of a cost/effectiveness ratio.
3. Measuring terms in the numerator (i.e., costs).
4. Valuing the health consequences in the denomi­
nator.
5. Estimating effectiveness of the intervention.
6. Time preference and discounting.

7. Handling uncertainty in CEA.
8. Reporting guidelines.
With regard to category #1, the Panel chose the
societal perspective for the reference case they con-
sidered in their analysis, because it does not represent
the viewpoint of any specific group and provides a
benchmark against which to assess results from other
perspectives. “The reference case is defined by a stan-
dard set of methods and assumptions. It includes a set
of standard results: the reference case results” (Russell,
Gold, Siegel, Daniels, & Weinstein, 1996, p. 1173).
As to category #2, the “results of CEA are summarized
in cost-effectiveness ratios that demonstrate the cost
of achieving a unit of health effect (e.g., cost per year
of life gained) for diverse types of patients and for
variations of the intervention” (Russell et al., 1996, p.
1173). Changes in health due to the intervention are
captured in the denominator, and changes in resource
use due to the intervention are captured in the numera-
tor. The societal perspective dictates that all important
effects on human health (e.g., social and physical
function, quality of life, years of life) and on resources
must be included in the analysis. Therefore the panel
recommended that the denominator of the C/E ratio
be reserved for the improvement in health associated
with an intervention (Weinstein et al., 1996). With
regard to category #3, there were several recommen-
dations related to what to include in the numeration of
the C/E ratio, which reflects the change in costs or use
of resources caused by a health intervention. For
example, the Panel recommended that costs be mea-
sured in constant dollars, an approach that requires
adjusting dollars for different years because of infla-
tion. Valuing the health consequences in the denomi-
nator (category #4) requires a measurement of adjusted
life-years expressed in an interval scale so that the
ratio of differences between values is meaningful.
Acceptable data for estimating effectiveness of the
intervention, which is category #5, may come from a
variety of sources such as randomized controlled
trials, observational studies, uncontrolled experi-
ments, or descriptive series (Weinstein et al., 1996).
As to category #6, on time preference and discounting,
the Panel recommended that costs and health out-
comes occurring during different time periods should
be discounted to their present value and that they
should be discounted at the same rate. The Panel rec-
ommended a discount rate of 3%. It is important to
address uncertainty in conclusions of CEA (category
#7) because CEA conclusions might change with
changes in assumptions or values (e.g., discount rate).
Uncertainty is addressed by varying the values/
assumptions in the analysis and assessing whether the
CHAPTER 13  Outcomes Research 313
conclusions still hold. For the last category, reporting
guidelines, the Panel developed a set of recommenda-
tions for reporting CEA that addressed how to report
the background to the problem, data and methods,
results, and discussion (Siegel, Weinstein, Russell, &
Gold, 1996). Building on these recommendations,
Stone (1998) described the methodology for conduct-
ing and reporting cost-effectiveness analyses in
nursing.
Costs can vary hugely across jurisdictions, because
hospitals may have different success in negotiating
with suppliers for drugs and equipment (Drummond,
Groeree, Moayyedi, & Levine, 2008). Costs also
depend on how care is organized. “The same service
may be delivered by a physician or a nurse practitio-
ner, in the outpatient setting or in the hospital, and
with or without administrative costs related to the
adjudication of patient eligibility to receive the
service” (Drummond et al., 2008, p. 623). It is time
for nurses to take a more active role in conducting
cost-effectiveness research. Nurses are well positioned
to evaluate healthcare practices and have the incentive
to conduct the studies. However, nursing practice is
seldom a subject of cost-effectiveness analyses. The
knowledge gained from this effort could enable nurses
to refine their practice by substituting interventions
that maximize nurses’ time (a cost variable) to the best
advantage of patient outcomes.
Fraher, Collins, Bourke, Phelan, and Lynch (2009)
conducted an economic analysis of employing a total
parenteral nutrition surveillance nurse for the preven-
tion of catheter-related bloodstream infections. The
following is an abstract of that study.
Background: “The cost of catheter-related blood-
stream infection (CRBSI) is substantial in terms of
morbidity, mortality, and financial resources. Total
parenteral nutrition (TPN) is a recognized risk factor
for CRBSI. In 1997, an intravenous nutrition nurse
was promoted to TPN surveillance clinical nurse
manager (CNM) and quarterly infection audit meet-
ings were introduced to monitor trends in CRBSI.”
Methods: “Data were prospectively collected over
a 15-year period using specific TPN records in a
535-bed tertiary acute university hospital. A total of
20,439 CVC-days and 307 CRBSIs were recorded.
Mean number of infections before, and after, the
introduction of a dedicated TPN surveillance CNM
were compared.”
Results: “Mean CRBSI per 1000 catheter-days ± SD
was 20.5 ± 6.34 prior to 1997 and 14.64 ± 7.81 after
314 UNIT TWO  The Research Process
1997, representing a mean reduction of 5.84 CRBSIs
per 1000 catheter-days (95% CI: –4.92 to 16.60;
p=0.05). Mean number of CRBSIs per year ± SD was
28.3 ± 4.93 prior to 1997 and 18.5 ± 7.37 after 1997,
representing a mean decrease of 9.8 infections per
year (95% CI: 0.01 to 19.66; p<0.05). The savings
made by preventing 9.8 infections per year were cal-
culated from data on bed-days obtained from the
hospital finance office. The cost in hospital days saved
per annum was c 135,000.”
Conclusions: “Introduction of the TPN surveillance
CNM saved the hospital at least c 78,300 per annum
and led to a significant decrease in CRBSIs in TPN
patients.” (Fraher et al., 2009, p. 129)
Ethical Studies
Outcomes studies often lead to policies for allocating
scarce resources. Ethicists take the position that moral
principles, such as justice, constrain the use of costs
and benefits to choose treatments that might maximize
the benefit per unit cost. Value commitments are inher-
ent in choices about research methods and about the
selection and interpretation of outcome variables, and
researchers should acknowledge these commitments.
“The choices researchers make should be documented
and the reasons for those choices should be given
explicitly in publications and presentations so that
readers and other users of the information are enabled
and expected to bear more responsibility for interpret-
ing and applying the findings appropriately” (Lynn &
Virnig, 1995, p. AS292). Veatch (1993) proposed that
by analyzing the implications of rationing decisions in
terms of the principles of justice and autonomy, we
would establish more acceptable criteria than we
would by using outcomes predictors alone. As an
example, Veatch performed an ethical analysis of the
use of outcome predictors in decisions related to the
early withdrawal of life support. Ethical studies need
to play an important role in outcomes programs of
research.
Sperling and Simon (2010) investigated physician
attitudes and policies regarding access to fertility care
and assisted reproductive technologies in Israel. The
abstract of that study follows:
“Despite the high profile of fertility care and assisted
reproductive technologies, their social and regulatory
contexts remain largely unexplored. Yet, studies
reveal a practice of candidate screening on a some-
what arbitrary basis. Examining the above issues is of
special importance to Israel, given its high fertility
rates. To this end, this study conducted a survey of
physicians’ attitudes regarding access to fertility care
and treatment. An anonymous questionnaire was dis-
tributed among IVF [in vitro fertilization] providers in
all fertility clinics in Israel during 2008-2009. A total
of 46 physicians (> 40%) responded. Although all
agree that every person has a right to procreate,
15.25% believe it is important to screen candidates
and 55.6% believe they should consider non-medical
criteria when providing care. Only 47.8% of physi-
cians acknowledge the existence of guidelines in their
units, but where they exist, 22.5% state they do not
follow them. Furthermore, between 24.4-63.0% of
physicians are willing to perform controversial proce-
dures if backed by official guidelines. In conclusion,
existing guidelines are often vague or ignored. Con-
trary to the USA, IVF providers in Israel are shaped
by the pro-natalist [an attitude or policy that encour-
ages childbearing] approach highly encouraged by the
state and they act less as trustees and gatekeepers to
the future child.” (Sperling & Simon, 2010, p. 854)
Measurement Methods
The selection of appropriate outcome variables is criti-
cal to the success of a study (Bernstein & Hilborne,
1993). As in any study, the researcher must evaluate
the evidence of validity and the reliability of the mea-
surement methods. Outcomes selected for nursing
studies should be those most consistent with nursing
practice and theory (Harris & Warren, 1995). In some
studies, rather than selecting the final outcome of care,
which may not occur for months or years, researchers
use measures of intermediate end points. Intermedi-
ate end points are events or markers that act as pre-
cursors to the final outcome. It is important, however,
to document the validity of the intermediate end point
in predicting the outcome (Freedman & Schatzkin,
1992). In early outcomes studies, researchers selected
outcome measures that they could easily obtain rather
than those most desirable for outcomes studies. Later
outcome studies have selected outcome measures
from secondary data sources (e.g., Aiken et al., 2008;
Cummings, Midodzi, Wong, & Estabrooks, 2010).
Secondary analysis is “any reanalysis of data or infor-
mation collected by another researcher or organiza-
tion, including analysis of data sets collected from a
variety of sources to create time-series or area-based
data sets” (Shi, 2008, p. 129). Outcomes researchers
have used secondary data from sources such as hospi-
tal discharge data (Aiken et al., 2008; Cummings

et al., 2010). Data collected through NDNQI or
CALNOC can also be utilized in nursing outcomes
research.
Table 13-4 identifies characteristics important to
evaluate in selecting methods of measuring outcomes.
In evaluating a particular outcome measure, the
researcher should consult the literature for previous
studies that have used that particular method of mea-
surement, including the publication describing devel-
opment of the method of measurement. This approach
was used by Doran et al. (2011) when they reviewed
the state of the science on nursing-sensitive outcomes.
For each outcome concept (e.g., functional status,
pain, pressure ulcer, self-care), the empirical literature
investigating the concept in nursing research was
reviewed. The approach to measurement of the
outcome concept was identified, and the reliability,
sensitivity, and validity of the measurement tools were
appraised and summarized in tabular form. Sensitivity
to change is an important measurement property to
consider in outcomes research because researchers are
often interested in evaluating how outcomes change
in response to healthcare interventions. As the sensi-
tivity of a measure increases, statistical power
increases, allowing smaller sample sizes to detect sig-
nificant differences. For a full discussion of reliability
and validity of scales and questionnaires, precision
and accuracy of physiological measures, and sensitiv-
ity and specificity of diagnostic tools, refer to Chapter
16 of this text.
Statistical Methods for Outcomes Studies
Although outcomes researchers test for the statistical
significance of their findings, this evaluation is not
considered sufficient to judge the findings as impor-
tant. Their focus is the clinical importance of study
findings (see Chapter 21 for more information on
clinical importance). In analyzing data, outcomes
researchers have moved away from statistical analyses
that use the mean to test for group differences. They
place greater importance on analyzing change scores
and use exploratory methods for examining the data
to identify outliers.
Analysis of Change
With the focus on outcomes studies has come a
renewed interest in methods of analyzing change.
Gottman and Rushe (1993) reported that the first book
addressing change in research, Problems in Measuring
Change, edited by Harris (1967), is the basis for most
of the current approaches to analyzing change.
However, some new ideas have emerged regarding the
analysis of change, such as the studies by Tracy et al.
CHAPTER 13  Outcomes Research 315
(2006) and Bettger, Coster, Latham, and Keysor
(2008).
For some outcomes, the changes may be nonlinear
or may go up and down rather than always increasing.
Thus, it is as important to uncover patterns of change
as it is to test for statistically significant differences at
various time points. Some changes may occur in rela-
tion to stages of recovery or improvement. These
changes may occur over weeks, months, or even years.
A more complete picture of the process of recovery
can be obtained by examining the process in greater
detail and over a broader range. With this approach,
the examiner can develop a recovery curve, which
provides a model of the recovery process that can then
be tested (Boz et al., 2004; Hernandez, Fernandez,
Luzon, Cuena, & Montejo, 2007; McCauley, Hannay,
& Swank, 2001).
Analysis of Improvement
In addition to reporting the mean improvement
score for all patients treated, it is important to report
what percentage of patients improved. Do all patients
improve slightly, or is there a divergence among
patients, with some improving greatly and others not
improving at all? This divergence may best be illus-
trated by plotting the data. Researchers studying a
particular treatment or approach to care might develop
a standard or index of varying degrees of improvement
that might occur. The index would allow better com-
parisons of the effectiveness of various treatments.
Characteristics of patients who experience varying
degrees of improvement should be described, and out-
liers should be carefully examined. This step requires
that the study design include baseline measures of
patient status, such as demographic characteristics,
functional status, and disease severity measures. An
analysis of improvement allows better judgments of
the appropriate use of various treatments (Fasting &
Gisvold, 2003).
Variance Analysis
Variance analysis is used to track individual and
group variance from a specific critical pathway. The
goal is to decrease preventable variance in process,
thus helping patients and their families achieve optimal
outcomes. Some of the variance is due to comorbidi-
ties. You may find that keeping a patient with comor-
bidities on the desired pathway may require you to
utilize more resources early in the patient’s care. Thus,
it is important to track both variance and comorbidi-
ties. Studies examining variations from pathways may
make it easier for healthcare providers to tailor exist-
ing critical pathways for specific comorbidities.
316 UNIT TWO  The Research Process

TABLE 13-4  Characteristics of Outcomes Assessment Instruments*

Characteristic Considerations in Patient Outcomes Evaluation References
Applicability Consider purpose of instruments Deyo & Carter, 1992
Discriminate between subjects at a point in time Stewart et al., 1989
Predict future outcomes Guyatt, Walter, & Norman, 1987
Evaluate changes within subjects over time Feinstein, Josephy, & Wells, 1986
Screen for problems Deyo, 1984
Provide case-mix adjustment
Assess quality of care
Consider whether norms are established for clinical population of interest
Instrument format is compatible with assessment approach (e.g.,
observer-rated versus self-administered)
Setting in which instrument was developed
Practicality The Instrument: Leidy, 1991
(clinical utility) Includes outcomes important to the patient Nelson, Landgraf, Hays, Wasson,
Is short and easy to administer (low respondent burden) includes & Kirk, 1990
questions that are easy to understand and acceptable to patients and Stewart et al., 1989
interviewers Lohr, 1988
Has scores that reflect condition severity and condition-specific features, Bombardier & Tugwell, 1987
and discriminate those with conditions from those without Feinstein et al., 1986
Is easily scored and has readily understandable scores Kirshner & Guyatt, 1985
Uses a level of measurement that allows a change score to be determined Deyo, 1984
Provides information that is clinically useful
Is performance or capacity based
Includes patient rating of magnitude of effort and support needed for
performance of physical tasks
Comprehensiveness Generic measures are designed to summarize a spectrum of concepts Nelson et al., 1990
applied to different impairments, illnesses, patients, and populations Patrick & Deyo, 1989
Disease-specific measures are designed to assess specific patients with Deyo, 1984
specific conditions or diagnoses
Dimensions of the instrument; a core set of physical, mental, and role
functions-desirable
Reliability Can be influenced by day-to-day variations in patients, differences Nelson et al., 1990
between observers, items in the scale, mode of administration Spitzer, 1987
Is the critical determinant of usefulness of an instrument Guyatt et al., 1987
Is designed for discriminative purpose Deyo, 1984
Validity No consensus of what are scientifically admissible criteria for many Spitzer, 1987
indices Deyo, 1984
No “gold standard” exists for establishing criterion validity for many
indices
Responsiveness Not yet indexed for virtually any evaluative measures Stewart & Archbold, 1992
Coarse scale rating may not detect changes Leidy, 1991
Aggregated scores may obscure changes in subscales Jaeschke, Singer, & Guyatt, 1989
Useful for determining sample size and statistical power Bombardier & Tugwell, 1987
Reliable instruments are likely to be responsive, but reliability not Guyatt et al., 1987
adequate as sole index of consistent results over time Deyo & Centor, 1986
Detail in scaling should be considered Deyo, 1984
As baseline variability of score changes within stable subjects, larger
treatment effects may be needed to demonstrate efficacy
Temporal relationship between intervention and outcome should be
considered

*Examples are illustrative; for complete list of structural, process, and outcome indicators, refer to original sources cited in the table.
Modified from Harris M. R., & Warren, J. J. (1995). Patient outcomes: Assessment issues for the CNS. Clinical Nurse Specialist, 9(2), 82.

Variance analysis can also be used to identify
at-risk patients who might benefit from the services of
a case manager. Variance analysis tracking is expressed
through the use of graphics, with the expected pathway
plotted on a graph. The care providers plot deviations
(negative variance) on the graph, allowing immediate
comparison with the expected pathway. Deviations
may be related to the patient, system, or provider
(Okita et al., 2009; Olive & Solomonides, 2009;
Tidwell, 1993).
Latent Transition Analysis
Latent transition analysis (LTA) is used in situations
in which stages or categories of recovery have been
defined and transitions between stages can be identi-
fied. To use this analysis method, the researchers
assign each member of the population to a single cat-
egory or stage for a given point of time. However,
stage membership changes over time. The analysis
tests stage membership to provide a realistic picture
of development.
Roberts and Ward (2011) described an example of
latent transition analysis (LTA) for nursing research
by presenting a case example, using secondary analy-
sis. The secondary analysis involved a psychoeduca-
tional intervention for cancer pain management in
which LTA was used to describe for whom and in what
direction the tailored intervention resulted in change
with respect to attitudinal barriers and pain symptoms.
The study sample consisted of 791 participants from
the original study. Participants’ response patterns to
the Barriers Questionnaire II (Ward, Wang, Serlin,
Peterson, & Marray, 2009) were used to determine
latent classes for the LTA, as described here:
“The LTA model was specified using a four-step
process. … Specifically, these steps were (a) defining
latent class structure using latent class analysis, (b)
predicting class membership by adding demographic
predictors to the latent class model at Time 1, (c)
testing distal pain outcomes across latent classes at
Time 2, and (d) modeling and testing transition prob-
abilities for change in class membership over time.”
(Roberts & Ward, 2011, p. 75)
Roberts and Ward (2011) found that the older, less
educated individuals in the study were more likely to
be in the High Barriers class at Time 1. For those
individuals who transitioned across classes, those who
received the intervention were more likely to move in
a favorable direction (to the Low Barriers class). The
CHAPTER 13  Outcomes Research 317
results from LTA provided useful information about
for whom and in what direction the intervention
resulted in change. In addition, LTA is a useful proce-
dure for nurse researchers because it collapses large
arrays of categorical data into meaningful patterns
(Roberts & Ward, 2011). For example, Cain, Epler,
Steinley, and Sher (2010) used LTA to examine change
in patterns of concerns related to eating, weight, and
shape in young adult women.
Multilevel Analysis
Multilevel analysis is a statistical technique that
is useful when data exist at several different levels,
for example, if data are nested or if a study has
several different units of analysis. It has been used in
epidemiology to study how environmental factors
(aggregate-level characteristics) and individual attri-
butes and behaviors (individual-level characteristics)
interact to influence individual-level health behaviors
and disease risks. Growing numbers of nursing studies
have used multilevel analysis as a statistical technique
to study phenomena measured at different units of
analysis. One such example is the study by Simon,
Muller, and Hasselhorn (2010), who used multilevel
analysis to determine which variables were associated
with nurses’ intention to leave the profession and
which were associated with their intention to leave the
organization. As described in the following excerpt
from the abstract, data existed at individual, depart-
mental, and hospital levels:
“A secondary data analysis of the German sample
of the European Nurses’ Early Exit-Study was per-
formed, using a generalized linear mixed-model
approach… Data from 2119 Registered Nurses in 71
departments of 16 hospitals from 2003 were ana-
lyzed. Models for intentions to leave the profession
explained more variance (R2 = 0.46) than models for
intentions to leave the organization (R2= 0.28). Both
leaving intentions were associated with age, profes-
sional commitment, and job satisfaction. Intentions to
leave the profession were strongly associated with
variables related to the personal background and the
work/home interface whereas intentions to leave the
organization were related to organizational leader-
ship and the local context.” (Simon et al., 2010,
p. 616)
In this example, the multilevel analysis was a
useful technique for differentiating the determinants
of nurses’ intention to leave, demonstrating that
318 UNIT TWO  The Research Process
departmental level variables (e.g., leadership) had
an influence on intention to leave the organization
whereas individual level variables (e.g., personal
background) had an influence on intention to leave the
profession. In another nursing study, Tak, Sweeney,
Alterman, Baron, and Calvert (2010) examined deter-
minants of workplace assaults on nursing assistants in
U.S. nursing homes. Data from the National Nursing
Assistant Survey (individual level) were linked to
facility information (organizational level). The inves-
tigators found that mandatory overtime and not having
enough time to assist residents with their activities of
daily living (i.e., workload) were associated with
experiencing injuries from assault as well as working
in a facility with Alzheimer care units.
Disseminating Outcomes
Research Findings
Once you have completed fieldwork and data analysis
and have documented your findings, you need to
ensure that your research findings are disseminated.
Often funding agencies have particular requirements
about the dissemination of research findings, which
involves the communication of studies through pre-
sentations and publications. This means that the first
step in deciding the format of your reporting is to
understand the responsibilities that you have to funders
and others. Including plans for the dissemination of
findings is an essential component of a program of
research. Strategies for the dissemination of research
findings need to be carefully planned and need to
include at least the following:
• Decide on the objectives of dissemination.
• Identify the audience for your material.
• Ascertain whether you will be preparing a written
or oral report (usually both).
• Identify other methods for disseminating results.
• Identify timelines for each section of the reporting
strategy.
The audiences for your research could include (1)
the clinicians, who will apply the knowledge to prac-
tice; (2) the public, who may make healthcare deci-
sions on the basis of the information; (3) healthcare
institutions, which must evaluate care in their facilities
on the basis of the information; (4) health policy
makers, who may set standards on the basis of the
information; and (5) researchers, who may use the
information in designing new studies. Disseminating
information to these various constituencies through
presentations at meetings and publications in a variety
of journals, as well as releasing the information to the
news media, requires careful planning (see Chapter
27, Disseminating Research Findings).
KEY POINTS
• Outcomes research examines the end results of
patient care.
• The scientific approaches used in outcomes studies
differ in some important ways from those used in
traditional research.
• Donabedian (1987, 2005) developed the theory on
which outcomes research is based.
• Quality is the overriding construct of the theory,
although Donabedian never defined this term.
• The three major concepts of the theory are health,
subjects of care, and providers of care.
• Donabedian identified three objects of evaluation
in appraising quality: structure, process, and out­
come.
• The goal of outcomes research is to evaluate out-
comes as defined by Donabedian, whose theory
requires that identified outcomes be clearly linked
with the process that caused the outcome.
• Clinical guideline panels are developed to incorpo-
rate available evidence on health outcomes.
• Outcome studies provide rich opportunities to
build a stronger scientific underpinning for nursing
practice.
• A nursing-sensitive patient outcome is “sensitive”
because it is influenced by nursing.
• Organizations currently involved in efforts to study
nursing-sensitive outcomes include the American
Nurses Association, the National Quality Forum,
the Collaborative Alliance for Nursing Outcomes,
the Veterans Affairs Nursing Outcomes Database,
the Center for Medicare and Medicaid Services’
Hospital Quality Initiative, the American Hospital
Association, the Federation of American Hospitals,
The Joint Commission, and the Agency of Health-
care Research and Quality.
• Another area of interest in terms of outcomes
research is the process of care delivered by
advanced practice nurses (nurse practitioners, nurse
midwives, nurse anesthetists, and clinical nurse
specialists).
• Outcome design strategies tend to have less control
than traditional research designs and, except for the
clinical trial, seldom use random samples; rather,
they use large representative samples.
• Statistical approaches used in outcomes studies
include new approaches to examining measure-
ment reliability, strategies to analyze change, the

analysis of improvement, variance analysis, latent
transition analysis, and multilevel analysis.
• Dissemination is an important part of the outcomes
research process to ensure the study results have an
impact on patients, providers, and healthcare
organizations.
REFERENCES
Agency for Healthcare Research and Quality (AHRQ, 2010). HHS
awards $437 million in patient-centered outcomes research
funding. Retrieved from http://www.ahrq.gov/news/press/pr2010/
recoveryawpr.htm/.
Agency for Healthcare Research and Quality (AHRQ, 2011).
National Guideline Clearinghouse. Retrieved from http://www
.guideline.gov/.
Aiken, L. H., Clarke, S. P., Sloane, D. M., Lake, E. T., & Cheney,
T. (2008). Effects of hospital care environment on patient mortal-
ity and nurse outcomes. Journal of Nursing Administration, 38(5),
223–229.
Alexander, G. R. (2007). Nursing sensitive databases: Their exis-
tence, challenges and importance. Medical Care Research and
Review, 64(2), 44S–63S.
Bakker, R., Steegers, E., Biharie, A., Mackenbach, J., Hofman, A.,
& Jaddoe, V. (2011). Explaining differences in birth outcomes
in relation to maternal age: The Generation R Study. BJOG: An
International Journal of Obstetrics and Gynaecology, 118(4),
500–509.
Bernstein, S. J., & Hilborne, L. H. (1993). Clinical indicators: The
road to quality care? Joint Commission Journal on Quality
Improvement, 19(11), 501–509.
Bettger, J. A., Coster, W. J., Latham, N. K., & Keysor, J. J. (2008).
Analyzing change in recovery patterns in the year after acute
hospitalization. Achieves of Physical Medicine & Rehabilitation,
89(7), 1267–1275.
Bombardier, C., & Tugwell, P. (1987). Methodological consider-
ations in functional assessment. Journal of Rheumatology,
14(Suppl. 15), 7–10.
Boz, C., Ozmenoglu, M., Alioglu, Z., Velioglu, S., Altunayoglu, V.,
& Gazioglu, S. (2004). Local cold effect on the excitability recov-
ery curve of the sympathetic skin response. Electromyography &
Clinical Neurophysiology, 44(8), 497–501.
Cain, A. S., Epler, A. J., Steinley, D., & Sher, K. J. (2010). Stability
and change in patterns of concerns related to eating, weight, and
shape in young adult women: A latent transition analysis. Journal
of Abnormal Psychology, 119(2), 255–267.
Canadian Institute for Health Information. (2012). Frequently-asked
questions about the MIS Standards. Ottawa, Canada: Author.
Retrieved from http://www.cihi.ca/cihi-ext-portal/internet/en/
document/standards+and+data+submission/standards/
mis+standards/mis_faq/.
Clancy, C. M., & Eisenberg, J. M. (1998). Outcomes research:
Measuring the end results of health care. Science, 282(5387),
245–246.
Collaborative Alliance for Nursing Outcomes (CALNOC, 2010).
CALNOC: About us. Retrieved from https://www.calnoc.org/
globalPages/mainpage.aspx/.
CHAPTER 13  Outcomes Research 319
Cummings, G. G., Midodzi, W. K., Wong, C. A., & Estabrooks,
C. A. (2010). The contribution of hospital nursing leadership
styles to 30-day patient mortality. Nursing Research, 59(5),
331–339.
Cunningham, R. S. (2004). Advanced practice nursing outcomes: A
review of selected empirical literature. Oncology Nursing Forum,
31(2), 219–230.
Davies, B., Edwards, N., Ploeg, J., & Virani, T. (2008). Insights
about the process and impact of implementing nursing guidelines
on delivery of care in hospitals and community settings. BMC
Health Services Research, 8, 29.
Deyo, R. A. (1984). Measuring functional outcomes in therapeutic
trials for chronic disease. Controlled Clinical Trials, 5(3),
223–240.
Deyo, R. A., & Carter, W. B. (1992). Strategies for improving and
expanding the application of health status measures in clinical
settings. Medical Care, 30(5 suppl), MS176–MS186.
Deyo, R. A., & Centor, R. M. (1986). Assessing the responsiveness
of functional scales to clinical change: An analogy to diagnostic
test performance. Journal of Chronic Disease, 39(11), 897–906.
Deyo, R. A., Taylor, V. M., Diehr, P., Conrad, D., Cherkin, D. C.,
Ciol, M., et al. (1994). Analysis of automated administrative and
survey databases to study patterns and outcomes of care. Spine,
19(18), 2083S–2091S.
Donabedian, A. (1976). Benefits in medical care programs. Cam-
bridge, MA: Harvard University Press.
Donabedian, A. (1978). Needed research in quality assessment and
monitoring. Hyattsville, MD: U.S. Department of Health, Educa-
tion, and Welfare, Public Health Service: National Center for
Health Services Research.
Donabedian, A. (1980). Explorations in quality assessment and
monitoring. Ann Arbor, MI: Health Administration Press.
Donabedian, A. (1982). The criteria and standards of quality. Ann
Arbor, MI: Health Administration Press.
Donabedian, A. (1987). Some basic issues in evaluating the quality
of health care. In: L. T. Rinke (Ed.), Outcome measures in home
care (p. 338) (Vol. I). New York, NY: National League for
Nursing. (Original work published 1976.)
Donabedian, A. (2005). Evaluating the quality of medical care. The
Milbank Quarterly, 83(4), 691–729.
Doran, D. M. (Ed.) (2011). Nursing outcomes: The state of the
science (2nd ed.). Sudbury, MA: Jones & Bartlett.
Doran, D. M., Mildon, B. & Clarke, S. (2011). Toward a national
report card in nursing: A knowledge synthesis. Canadian Journal
of Nursing Leadership, 24(2), 38–57.
Doran, D. M., Sidani, S., & Di Pietro, T. (2010). Nursing sensitive
outcomes. In J. S. Fulton, B. Lyon, & K. Goudreau (Eds.), Foun-
dations of Clinical Nurse Specialist Practices (pp. 35–37), New
York, NY: Springer.
Doran, D., Harrison, M. B., Laschinger, H., Hirdes, J., Rukholm,
E., Sidani, S., et al. (2006). Relationship between nursing inter-
ventions and outcome achievement in acute care settings. Research
in Nursing & Health, 29(1), 61–70.
Drummond, M., Goeree, R., Moayyedi, P., & Levine, M. (2008).
Economic analysis. In G. Guyatt, D. Rennie, M. O. Meade, &
D. J. Cook (Eds.), Users’ guides to the medical literature: A
manual for evidence-based practice (2nd ed.) (pp. 619–641). New
York, NY: McGraw-Hill Medical.
320 UNIT TWO  The Research Process
Dunn, S., Anderson, G. M., & Bierman, A. S. (2009). Temporal and
regional trends in IUD insertion: A population-based study in
Ontario, Canada. Contraception, 80(5), 469–473.
Epstein, A. J., & Nicholson, S. (2009). The formation and evolution
of physician treatment styles: An application to cesarean sections.
Journal of Health Economics, 28(6), 1126–1140.
Fasting, S., & Gisvold, S. E. (2003). Statistical process control
methods allow the analysis and improvement of anesthesia care.
Canadian Journal of Anesthesia, 50(8), 767–774.
Feinstein, A. R., Josephy, B. R., & Wells, C. K. (1986). Scientific
and clinical problems in indexes of functional disability. Annals
of Internal Medicine, 105(3), 413–420.
Fraher, M. H., Collins, C. J., Bourke, J., Phelan, D., & Lynch, M.
(2009). Cost-effectiveness of employing a total parenteral nutri-
tion surveillance nurse for the prevention of catheter-related
bloodstream infections. Journal of Hospital Infection, 73(2),
129–134.
Freedman, L. S., & Schatzkin, A. (1992). Sample size for study-
ing intermediate endpoints within intervention trials or observa-
tional studies. American Journal of Epidemiology, 136(9),
1148–1159.
Gold, M. (1996). Panel on cost-effectiveness in health and medi-
cine. Medical Care, 34(12), DS197–DS199.
Gottman, J. M., & Rushe, R. H. (1993). The analysis of change:
Issues, fallacies, and new ideas. Journal of Consulting and Clini-
cal Psychology, 61(6), 907–910.
Graham, I. D., Bick, D., Tetroe, J., Straus, S. E., & Harrison, M. B.
(2011). Measuring outcomes of evidence-based practice: Distin-
guishing between knowledge use and its impact. In D. Bick &
I. D. Graham (Eds.), Evaluating the impact of implementing
evidence-based practice (pp. 18–37). Oxford, United Kingdom:
Wiley-Blackwell.
Guess, H. A., Jacobsen, S. J., Girman, C. J., Oesterling, J. E., Chute,
C. G., & Panser, L. A., et al. (1995). The role of community-based
longitudinal studies in evaluating treatment effects. Example:
Benign prostatic hyperplasia. Medical Care, 33(Suppl. 4),
AS26–AS35.
Guyatt, G., Rennie, D., Meade, M. O., & Cook, D. J. (2008). Users’
guides to the medical literature: A manual for evidence-based
practice (2nd ed.). New York, NY: McGraw-Hill Medical.
Guyatt, G., Walter, S., & Norman, G. (1987). Measuring change
over time: Assessing the usefulness of evaluative instruments.
Journal of Chronic Disease, 40(2), 171–178.
Harris, C. W. (1967). Problems in measuring change. Madison, WI:
University of Wisconsin Press.
Harris, M. R., & Warren, J. J. (1995). Patient outcomes: Assessment
issues for the CNS. Clinical Nurse Specialist, 9(2), 82–86.
Hernandez, G., Fernandez, R., Luzon, E., Cuena, R., & Montejo,
J. C. (2007). The early phase of the minute ventilation recovery
curve predicts extubation failure better than the minute ventilation
recovery time. Chest, 131(5), 1315–1322.
Herrmann, D. (1995). Reporting current, past, and changed health
status: What we know about distortion. Medical Care, 33(Suppl.
4), AS89–AS94.
Ingersoll, G. L., McIntosh, E., & Williams, M. (2000). Nurse-
sensitive outcomes of advanced practice. Journal of Advanced
Nursing, 32(5), 1272–1282.
Jaeschke, R., Singer, J., & Guyatt, G. H. (1989). Measurement of
health status: Ascertaining the minimal clinically important dif-
ference. Controlled Clinical Trials, 10(4), 407–415.
Jennings, B. M., Loan, L. A., DePaul, D., Brosch, L. R., & Hildreth,
P. (2001). Lessons learned while collecting ANA indicator data.
Journal of Nursing Administration, 31(3), 121–129.
Keeney, S., Hasson, F., & McKenna, H. P. (2011). The Delphi
technique in nursing and health research. Oxford, United
Kingdom: Wiley-Blackwell.
Kelly, K. C., Huber, D. G., Johnson, M., McCloskey, J. C., & Maas,
M. (1994). The Medical Outcomes Study: A nursing perspective.
Journal of Professional Nursing, 10(4), 209–216.
Kieffer, E., Alexander, G. R., & Mor, J. (1992). Area-level predic-
tors of use of prenatal care in diverse populations. Public Health
Reports, 107(6), 653–658.
Kirshner, B., & Guyatt, G. (1985). A methodological framework for
assessing health indices. Journal of Chronic Diseases, 38(1),
27–36.
Kleib, M., Sales, A., Doran, D. M., Malette, C. & White, D. (2011).
Nursing minimum data sets. In D. M. Doran (Ed.), Nursing out-
comes: The state of the science (2nd ed.) (pp. 487–512). Sudbury,
MA: Jones & Bartlett.
Kramer, M., Maguire, P., & Schmalenberg, C. (2006). Excellence
through evidence: The what, when, and where of clinical auton-
omy. Journal of Nursing Administration, 36(10), 479–491.
Lake, E. T. (2006). Multilevel models in health outcomes research
Part I: Theory, design, and measurement. Applied Nursing
Research, 19(1), 51–53.
Lee, T. H., & Goldman, L. (1989). Development and analysis of
observational data bases. Journal of the American College of
Cardiology, 14(Suppl. 3A), 44A–47A.
Leidy, N. K. (1991). Survey measures of functional ability and dis-
ability of pulmonary patients. In: B. L. Metzger (Ed.), Synthesis
conference on altered functioning: Impairment and disability (pp.
52–79). Indianapolis, IN: Nursing Center Press of Sigma Theta
Tau International.
Lohr, K. N. (1988). Outcome measurement: Concepts and ques-
tions. Inquiry, 25(1), 37–50.
Lynn, J., & Virnig, B. A. (1995). Assessing the significance of treat-
ment effects: Comments from the perspective of ethics. Medical
Care, 33(4), AS292–AS298.
Magnello, M. E. (2010). The passionate statistician. In S. Nelson &
A. M. Rafferty (Eds.), Notes on Nightingale: The influence and
legacy of a nursing icon (pp. 115–129). Ithaca, NY: Cornell
University.
Mark, B. A. (1995). The black box of patient outcomes research.
Image: Journal of Nursing Scholarship, 27(1), 42.
McCauley, S. R., Hannay, H. J., & Swank, P. R. (2001). Use of the
disability rating scale recovery curve as a predictor of psychoso-
cial outcome following closed-headed injury. Journal of the Inter-
national Neuropsychology Society, 7(4), 457–467.
McDonald, C. J., & Hui, S. L. (1991). The analysis of humongous
databases: Problems and promises. Statistics in Medicine, 10(4),
511–518.
Mitchell, J. B., Bubolz, T., Pail, J. E., Pashos, C. L., Escarce, J. J.,
Muhlbaier, L. H., et al. (1994). Using Medicare claims for out-
comes research. Medical Care, 32(Suppl. 7), JS38–JS51.

Mitchell, P. H., Ferketich, S., & Jennings, B. M. (1998). American
Academy of Nursing Expert Panel on Quality Health Care 1998
Quality Health Outcomes Model. Image—Journal of Nursing
Scholarship, 30(1), 43–46.
Montalvo, I. (2007). The national database of nursing quality
indicators (NDNQI). The Online Journal of Issues in Nursing,
12(3). Retrieved from http://www.nursingworld.org/ojin.
Moses, L. E. (1995). Measuring effects without randomized trials?
Options, problems, challenges. Medical Care, 33(4), AS8–AS14.
National Association of Clinical Nurse Specialists. (NACNS, 2004).
Statement on clinical nurse specialist practice and education (2nd
ed.). Harrisburg, PA: Author.
National Database of Nursing Quality Indicators. (NDNQI, 2010).
ANA’s NQF-endorsed measure specifications: Guidelines for data
collection on the American Nurses Association’s national quality
forum endorsed measures: Nursing care hours per patient day;
skill mix; falls and falls with injury. Kansas City, KS: Author.
Retrieved from home page link https://www.nursingquality.org/
Default.aspx/.
National Quality Forum. (NQF, 2011). NQF Advancing perfor-
mance measurement. Report to the Congress and the Secretary of
the U.S. Department of Health and Human Services, covering the
period of January 14, 2010–January 13, 2011. Retrieved from
http://www.qualityforum.org/About_NQF/HHS_Performance_
Measurement.aspx/.
Naylor, M. D. (2007). Advancing the science in the measurement
of health care quality influenced by nurses. Medical Care Research
& Review, 64(Suppl. 4), 144S–169S.
Nelson, E. C., Landgraf, J. M., Hays, R. D., Wasson, J. H., & Kirk,
J. W. (1990). The functional status of patients: How can it be mea-
sured in physicians’ offices? Medical Care, 28(12), 1111–1126.
Okita, A., Yamashita, M., Abe, K., Nagel, C., Matsumoto, A.,
Akehi, M., et al. (2009). Variance analysis of a clinical pathway
of video-assisted single lobectomy for lung cancer. Surgery Today,
39(2), 104–109.
Olive, M., & Solomonides, T. (2009). Variance analysis as practice-
based evidence. Studies in Health Technology & Informatics, 147,
190–198.
Oncology Nursing Society (ONS, 2012). About the ONS. Retrieved
from http://www.ons.org/about/.
Patrick, D. L., & Deyo, R. A. (1989). Generic and disease-specific
measures in assessing health status and quality of life. Medical
Care, 27(Suppl. 3), S217–S232.
Roberts, T. J., & Ward, S. E. (2011). Using latent transition analysis
in nursing research to explore change over time. Nursing Research,
60(1), 73–79.
Rowell, P. (2001). Lessons learned while collecting ANA indicator
data: The American Nurses Association responds. Journal of
Nursing Administration, 31(3), 130–131.
Rozen, G., Ugoni, A. M. & Sheehan, P. M. (2011). A new perspec-
tive on VBAC: A retrospective cohort study. Women and Birth,
24(1), 3–9.
Russell, L. B., Gold, M. R., Siegel, J. E., Daniels, N., & Weinstein,
M. C. (1996). The role of cost-effectiveness analysis in health and
medicine. Panel on Cost-Effectiveness in Health and Medicine.
Journal of the American Medical Association, 276(14), 1172–
1177.
CHAPTER 13  Outcomes Research 321
Sadler, B. L., Joseph, A., Keller, A., & Rostenberg, B. (2009). Using
evidence-based environmental design to enhance safety and
quality. Institute for Healthcare Improvement (IHI): Innovation
series white paper. Cambridge, MA: IHI. Retrieved from http://
www.IHI.org/.
Schifalacqua, M. M., Mamula, J., & Mason, A. R. (2011). Return
on investment imperative: The cost of care calculator for an
evidence-based practice program. Nursing Administration Quar-
terly, 35(1), 15–20.
Sermeus, W., Delesie, L., Van den Heede, K., Diya, L., & Lesaffre,
E. (2008). Measuring the intensity of nursing care: Making use of
the Belgian nursing minimum data set. International Journal of
Nursing Studies, 45(7), 1011–1021.
Setty, A. R., Curhan, G. & Choi, H. K. (2007). Smoking and the
risk of psoriasis in women: Nurses’ Health Study II. American
Journal of Medicine, 120(11), 953–959.
Shi, L. (2008). Health services research methods (2nd ed.). Clifton
Park, NY: Delmar Cengage Learning.
Siegel, J. E., Weinstein, M. C., Russell, L. B., & Gold, M. R. (1996).
Recommendations for reporting cost-effectiveness analyses.
Panel on Cost-Effectiveness in Health and Medicine. Journal of
the American Medical Association, 276(16), 1339–1341.
Simon, M., Muller, B. H., & Hasselhorn, H. M. (2010). Leaving the
organization or the profession—a multilevel analysis of nurses’
intentions. Journal of Advanced Nursing, 66(3), 616–626.
Sperling, D., & Simon, Y. (2010). Attitudes and policies regarding
access to fertility care and assisted reproductive technologies in
Israel. Reproductive Biomedicine Online, 21(7), 854–861.
Spitzer, W. O. (1987). State of science 1986: Quality of life and
functional status as target variables for research. Journal of
Chronic Disease, 40(6), 465–471.
Stetler, C. B., & Caramanica, B. (2007). Evaluation of an evidence-
based practice initiative: Outcomes, strengths, and limitations of
a retrospective conceptually based approach. Worldviews on
Evidence-Based Nursing, 4(4), 187–199.
Stewart, A. L., Greenfield, S., Hays, R. D., Wells, K., Rogers,
W. H., Berry, S. D., et al. (1989). Functional status and well-being
of patients with chronic conditions. JAMA, 262(7), 907–913.
Stewart, B. J., & Archbold, P. G. (1992). Nursing intervention
studies require outcome measures that are sensitive to change:
Part 2. Research in Nursing & Health, 16(1), 77–81.
Stone, D. W. (1998). Methods for conducting and reporting cost-
effectiveness analysis in nursing. Image—Journal of Nursing
Scholarship, 30(3), 229–234.
Stone, P. W., Mooney-Kane, C., Larson, E. L., Horan, T., Glance,
L. G., Zwanziger, J., & Dick, A. W. (2007). Nurse working
conditions and patient safety outcomes. Medical Care, 45(6),
571–578.
Straus, S., Tetroe, J., Graham, I. D., Zwarenstein, M., & Bhattacha-
ryya, O. (2009). Monitoring and evaluating knowledge. In: S.
Straus, J. Tetroe, & I. D. Graham (Eds.), Knowledge translation
in health care (pp. 151–159). Oxford, UK: Wiley-Blackwell.
Swaen, G. M., & Meijers, J. M. (1988). Influence of design char-
acteristics on the outcomes of retrospective cohort studies. British
Journal of Industrial Medicine, 45(9), 624–629.
Tak, S., Sweeney, M. H., Alterman, T., Baron, S., & Calvert, G. M.
(2010). Workplace assaults on nursing assistants in U.S. nursing
322 UNIT TWO  The Research Process
homes: A multilevel analysis. American Journal of Public Health,
100(10), 1938–1945.
Tarlov, A. R., Ware, J. E., Jr., Greenfield, S., Nelson, E. C., Perrin,
E., & Zubkoff, M. (1989). The Medical Outcomes Study: An
application of methods for monitoring the results of medical care.
JAMA, 262(7), 925–930.
Tidwell, S. L. (1993). A graphic tool for tracking variance and
comorbidities in cardiac surgery case management. Progress in
Cardiovascular Nursing, 8(2), 6–19.
Tork, H. K., Dassen, T., & Lohrmann, C. (2008). Care dependency of
children in Egypt. Journal of Clinical Nursing, 17(3), 287–295.
Tourangeau, A. E. (2003). Modeling the determinants of mortality
for hospitalized patients. International Nursing Perspectives,
3(1), 37–48.
Tourangeau, A. E., Doran, D. M., Hall, L. M., O’Brien Pallas, L.,
Pringle, D., Tu, J. V., et al (2007). Impact of hospital nursing care
on 30-day mortality for acute medical patients. Journal of
Advanced Nursing 57(1), 32–44.
Tracy, S., Schinco, M. A., Griffen, M. M., Kerwin, A. J., Devin, T.,
& Tepas, J. J. (2006). Urgent airway intervention: Does outcome
change with personnel performing the procedure? Journal of
Trauma, 61(5), 1162–1165.
Van den Heede, K., Sermeus, W., Diya, L., Clarke, S. P., Lesaffre,
E., Vleugels, A., et al (2009). Nurse staffing and patient outcomes
in Belgian acute hospitals: Cross-sectional analysis of administra-
tive data. International Journal of Nursing Studies, 46(7),
928–939.
Veatch, R. (1993). Justice and outcomes research: The ethical limits.
Journal of Clinical Ethics, 4(3), 258–261.
Volinn, E., Diehr, P., Ciol, M. A., & Loeser, J. D. (1994). Why does
geographic variation in health care practices matter (and seven
questions to ask in evaluating studies on geographic variation)?
Spine, 19(18S), 2092S–2100S.
Waltz, C. F., Strickland, O. L., & Lenz, E. R. (2010). Measurement
in nursing and health research (4th ed.). New York, NY: Springer.
Ward, S. E., Wang, K. K., Serlin, R. C., Peterson, S. L., & Murray,
M. E. (2009). A randomized trial of a tailored barriers intervention
for cancer information service (CIS) callers in pain. Pain, 144(1–
2), 49–56.
Weinstein, M. C., Siegel, J. E., Gold, M. R., Kamlet, M. S., &
Russell, L. B. (1996). Recommendations of the Panel on Cost-
Effectiveness in Health and Medicine. Journal of the American
Medical Association, 276(15), 1253–1258.
Westert, G. P., & Groenewegen, P. P. (1999). Medical practice varia-
tions: Changing the theoretical approach. Scandinavian Journal
of Public Health, 27(3), 173–180.
World Health Organization. (2009). The Conceptual Framework for
the International Classification for Patient Safety, Version 1.0,
2007-2008. Geneva, Switzerland: Author. Retrieved from http://
www.who.int/patientsafety/taxonomy/en/.
Wray, N. P., Ashton, C. M., Kuykendall, D. H., Petersen, N. J.,
Souchek, J., & Hollingsworth, J. C. (1995). Selecting disease-
outcome pairs for monitoring the quality of hospital care. Medical
Care, 33(1), 75–89.
   http://evolve.elsevier.com/Grove/practice/
Intervention-Based Research
I Intervention-Based Research
nvestigations of interventions that address prob-
lems or issues of concern to patients, nurses, and
healthcare organizations are vitally important to
expanding the scientific basis for nursing practice.
Intervention-based research verifies the efficacy,
effectiveness, and efficiency of interventions needed
for the development of evidence-based nursing prac-
tice. This type of research requires careful decision
making regarding the optimal study design, controlled
implementation of the treatment protocol and proce-
dures, use of quality outcome measure(s), and appro-
priate analysis of data to ensure the best conditions for
testing a selected intervention. A rigorously developed
and implemented intervention study improves the
likelihood of detecting statistically significant differ-
ences between the experimental and comparison
groups on specified outcomes. The ability to find a
statistical difference with an intervention is also based
on having enough statistical power from an adequately
sized sample and controlling sources of error and bias
that can obscure study findings. Error and bias intro-
duced into a study create threats to the validity of the
study and limit the ability of drawing definitive con-
clusions about the effects of the intervention.
This chapter examines approaches to intervention-
based research and describes various aspects of the
research process that are unique to the study of inter-
ventions. The focus is on highlighting those distinct
features of intervention-oriented research and explain-
ing designs and methods to accomplish such research.
Because the methods employed in testing interven-
tions can be complex and rigorous, additional guide-
lines are provided for critically appraising these
studies. The ability to critically appraise intervention-
based research has wide-spread implications for
examining the scientific merit of studies and translat-
ing findings into practice. This chapter also covers
essential information for nurses conducting interven-
tion studies.
14 CHAPTER
Conducted by Nurses
Thousands of intervention-based studies conducted by
nurses have contributed to the scientific foundations
for nursing practice. Nurse scientists, such as PhD-
prepared nurses, not only have conducted individual
studies but also have summarized the state of the
science in selected areas with meta-analyses and sys-
tematic reviews. This high level of synthesis of
research findings through meta-analyses and system-
atic reviews has enabled the pooling of data from
multiple studies to formulate statistical and evaluative
conclusions about the strength of evidence for inter-
ventions. Data compiled from intervention-based
studies have guided the development and application
of evidence-based practice approaches to care and, in
some instances, have contributed to the development
of national evidence-based guidelines. One of the
most influential meta-analyses that forever changed
nursing practice demonstrated that saline was just as
effective as heparin flush solution in maintaining the
patency of peripheral intravenous locks in adults
(Goode et al., 1991). Investigators pooled data from
15 studies involving a total of 3490 patients to draw
this conclusion. This would not have been possible
without the individual studies conducted to examine
the efficacy and effectiveness of a very common
nursing intervention.
Summaries of intervention-based research are also
valuable in determining how strong an intervention
might be by gauging its effectiveness among multiple
studies. For example, it is vitally important to under-
stand what health promotion interventions work best
and under what conditions. To address this issue,
Conn, Hafdahl, and Mehr (2011) conducted a meta-
analysis of studies focused on physical activity inter-
ventions in adults. They found that behavioral
approaches were more effective than cognitive ones in
323
324 UNIT TWO  The Research Process
improving adults’ activity outcomes. Face-to-face
delivery of the intervention, instead of telephone or
mail, and individualized-focused interventions, rather
than targeting of communities, were more effective in
promoting physical activities. Systematic review,
meta-analysis, meta-synthesis, and mixed-methods
systematic review techniques in the evaluation of
interventions across multiple studies provide much
stronger evidence for implementing them in practice
than one isolated investigation (see Chapter 19 for
conducting these research syntheses).
This chapter introduces you to approaches to
inquiry, ways of interpreting outcomes, and the evalu-
ative aspects of research aimed at testing nursing inter-
ventions and plans of care. You will learn to examine
components of intervention studies and to appreciate
the scientific approaches that are relevant to the study
of interventions. Few investigators do research alone,
and intervention research is no exception. More often
than not, research teams are required to accomplish
studies using interventions. These research teams are
composed of nurse scientists and sometimes investiga-
tors from other disciplines, such as medicine, phar-
macy, psychology, and sociology. Interprofessional
and/or interdisciplinary studies include additional
intellectual perspectives, collective efforts among sci-
entists, and additional resources that promote quality
research. Whether interventions are unique to nursing
or have implications for other disciplines, intervention-
based research involves similar processes consisting
of thoughtful planning, theory- or literature-driven
rationales, appropriate designs for testing the interven-
tion, systematic and consistent methods of implement-
ing the intervention and collecting data, data analyses
using parametric or nonparametric statistics, and dis-
semination of results. Examples of these steps in car-
rying out intervention research are described and
illustrated with examples from published nursing
studies.
Nursing Interventions
The term intervention-based research encompasses
a broad range of investigations that examines the
effects of any intervention or treatment having rele-
vance to nursing. Nursing interventions are defined
as “deliberative cognitive, physical, or verbal activi-
ties performed with, or on behalf of, individuals
and their families [that] are directed toward accom-
plishing particular therapeutic objectives relative to
individuals’ health and well-being” (Grobe, 1996, p.
50). An expansion of this definition includes nursing
interventions that are performed with, or on behalf of,
communities. Sidani and Braden (1998, p. 8) view
interventions as “treatments, therapies, procedures,
or actions implemented by health professionals to and
with clients, in a particular situation, to move the
clients’ condition toward desired health outcomes that
are beneficial to the clients.” Nursing interventions are
nurse-initiated and based on nursing classifications
of interventions unique to clinical problems or
issues addressed by nurses (Bulechek, Butcher, &
Dochterman, 2008; Forbes, 2009).
Historically, nursing interventions have tended to
be viewed as discrete actions, such as “positioning a
limb with pillows,” “raising the head of the bed 30
degrees,” and “assessing a patient’s pain.” Interven-
tions can be described more broadly as all of the
actions required to address a particular nursing
problem or issue. But there is little agreement regard-
ing the conceptualization of interventions and how
these discrete actions fit together (McCloskey &
Bulechek, 2000). Frameworks have been proposed in
moving from isolated action-oriented interventions to
more integrative approaches to care. For example,
dance therapy might be used as part of a falls reduc-
tion initiative (Krampe et al., 2010). In addition, care
bundles, which are combinations of interrelated
nursing actions, might be the basis for comprehensive
approaches to care (Deacon & Fairhurst, 2008; Quigley
et al., 2009). These variations in interventions are dis-
cussed in the following section.
Variations in Nursing Interventions
Nursing intervention research involves quite a few
variations in how the intervention is conceived, deliv-
ered, and evaluated. An intervention can be: (1) a
treatment, protocol, or clinical tools to improve patient
care processes and outcomes; (2) an educational
program to provide support for the patient; (3) an
educational program targeting nurses or other health-
care providers (e.g., a continuing education program);
or (4) a leadership practice to change staffing patterns
or improve the workplace environment.
Sequencing of interventions is also an important
consideration when one is designing and conducting
intervention-based research. A single intervention can
be delivered and studied at one point in time. This
approach is typically reserved for interventions that
require limited exposure to produce a response, such
as a one-time educational intervention of preoperative
teaching on the study participants’ knowledge. A
single intervention might also be delivered over a
series of time points. These interventions include

those that have to be repeated for their maximum
benefits, such as exercise programs, counseling ses-
sions, pain reduction strategies, and medication adher-
ence teaching. The effects of two or more separate
interventions might be examined in the context of a
study at just one point in time or delivered over mul-
tiple time points. Interventions can be combined and
administered together to determine the collective
impact that they may have on outcomes. Thus,
researchers can deliver different variations of inter-
ventions and examine outcomes at one time point or
across time points.
Stage-based interventions are tailored to a spe-
cific phase of recovery, response to treatment, or
change in behavior. These interventions can be deliv-
ered to: (1) study participants as they progress into
different phases along a continuum or recovery trajec-
tory over time that require different approaches or (2)
discrete or selected samples in various settings at
various points in recovery or along a continuum, as
with a cross-sectional design, when the intervention is
tailored to this stage. Stage-based interventions are
used in studies such as testing of smoking cessation
techniques when the intervention strategies differ
according to whether a person is contemplating quit-
ting smoking, has just quit smoking, or has remained
free of smoking (Cahill, Lancaster, & Green, 2010).
As a novice researcher, you might want to implement
a specific intervention and measure a selected outcome
at one point in time. Students are also encouraged to
join research teams to examine the effects of
interventions.
Nursing Intervention Taxonomies
An intervention taxonomy is an organized categori-
zation of the interventions performed by nurses. A
number of classifications of nursing diagnoses and
interventions have been developed, which include the
following:
• The Nursing Interventions Classification (NIC),
containing 542 direct and indirect care interven-
tions, each of which has an assigned numerical
code and that are categorized in seven domains
including a new community domain (McCloskey
& Bulechek, 2000; Bulechek et al., 2008). The
Center for Nursing Classification & Clinical Effec-
tiveness is housed at the University of Iowa College
of Nursing and can be accessed online (http://
www.nursing.uiowa.edu/excellence/nursing_
knowledge/clinical_effectiveness/nic.htm/).
• NANDA International (NANDA-I; formerly
North American Nursing Diagnosis Association)
CHAPTER 14  Intervention-Based Research 325
compiles nursing diagnoses and classifications
(http://www.nanda.org/Home.aspx/). NANDA-I
publishes the International Journal of Nursing Ter-
minologies and Classifications (IJNTC) quarterly,
which is distributed internationally.
• Home Health Care Classification System (HHCC)
has two taxonomies with 20 Care Components
used as a standardized framework to code, index,
and classify home health nursing practice (Saba,
2002).
• The Omaha System provides a list of diagnoses and
an Intervention Scheme designed to describe and
communicate multidisciplinary practice intended
to prevent illness, improve or restore health,
decrease deterioration, and/or provide comfort
before death. There are 75 targets or objects of
action (Martin, 2005). Martin and Bowles (2008)
address the link between practice and research that
is predicated on the Omaha System.
• The Nursing Intervention Lexicon and Taxonomy
(NILT) (Grobe, 1996). Grobe (1996, p. 50) sug-
gested that “theoretically, a validated taxonomy
that describes and categorizes nursing interventions
can represent the essence of nursing knowledge
about care phenomena and their relationship to one
another and to the overall concept of care.”
Although taxonomies may contain brief definitions
of interventions, they do not provide sufficient detail
to allow one to implement an intervention. The actions
identified in taxonomies may be too discrete for testing
and may not be linked to the resolution of a particular
patient problem. Populations and settings for which
nursing interventions are intended are not elucidated
through nursing intervention taxonomies, making
them somewhat ambiguous for guiding studies (Sidani
& Braden, 1998; Forbes, 2009).
Problems Examined by Intervention Studies
Interventions are developed to address problems or
issues in practice. The researcher must make a careful
analysis of the situational or contextual aspects of a
problem or issue prior to designing an intervention.
Box 14-1 outlines a list of questions that researchers
often ask to focus their attention on who, what, where,
and how in designing and implementing an interven-
tion in a study. Upon answering many of these ques-
tions, researchers gain a better idea of the planning
process needed before undertaking an intervention-
based study. Moreover, plans for a grand-scale study
may need to be executed in phases, depending on the
resources, commitment, support, and feasibility of the
study.
326 UNIT TWO  The Research Process
Box 14-1 Problem Analysis Questions
1. Who?
a. For whom is the situation, condition, or
circumstance a problem (e.g., patients,
families, nurses, other healthcare
professionals)?
b. Who (if anyone) would benefit from a
change as it is now?
c. Who should share the responsibility for
“solving” the problem?
d. Who are stakeholders having an invested
interest and stake in wanting positive
resolution to the problem?
e. Who might support change?
f. Who might function as champions for
change?
g. Who should be involved in designing the
intervention?
2. What?
a. What are the negative consequences of the
problem for the affected individuals?
b. What are the negative consequences of the
problem for the community (healthcare
providers or the healthcare system or
agency)?
c. What barriers exist to implementing an
intervention?
d. What has to occur for the problem to be
considered solved?
Questions adapted from Fawcett, S. B., Suarez-Belcazar, Y., Belcazar, F. E., White, G. W., Paine, A. L., Blanchard, K.
A., et al. (1994). Conducting intervention research: The design and development process. In J. Rothman & E. J.
Thomas (Eds.), Intervention research: Design and development for human service (pp. 25–54). New York, NY:
Haworth Press.
Programs of Nursing
Intervention Research
It is becoming increasingly clear that the design and
testing of a nursing intervention require an extensive
program of research rather than a single well-designed
study (Forbes, 2009; Sidani & Braden, 1998). As the
discipline of nursing advances its mission to accumu-
late a strong practice science, it is apparent that a
larger portion of nursing studies must focus on design-
ing and testing interventions. Moreover, intervention
research is central to the role of nurses in practice,
education, and leadership. Vallerand, Musto, and
Polomano (2011), in an extensive review of nurses’
roles in pain management, depict intervention-based
research as an integral component of advocacy, quality
care, and innovation in patient care (see Figure 14-1).
e. What conditions must change to establish
or support a resolution of the problem?
f. What is an acceptable and realistic level of
change that needs to occur?
g. What does each stakeholder have invested
in the status quo?
3. Where?
a. Where would the greatest impact for the
greatest numbers occur?
b. Where would it be easiest to bring about
change (technically, financially, and
politically)?
c. Where are the most resources available?
d. Where are the individuals who would be
most willing to participate in change?
4. How?
a. How should the problem be addressed (e.g.,
small scale or large scale)?
b. How will the complexity of the problem
require multilevel actions to have the
greatest impact for change?
c. How should stakeholders be involved and at
what points in development and execution
of a study might they be best used?
d. How long might it take to accomplish the
goals of the research?
Equally important in advancing nursing science are
replication studies, which mimic the design and pro-
cedures of interventions tested in previous research.
Despite the recognized importance of replicating
intervention studies, few have been repeated to vali-
date and verify their results.
Intervention research is a methodology that holds
great promise as a more effective way of testing inter-
ventions. It shifts the focus from causal connection to
causal explanation. In causal connection, the focus of
a study is to provide evidence that the intervention
contributes to the outcome. In causal explanation, in
addition to demonstrating that the intervention causes
the outcome, the researcher must provide scientific
evidence to explain why the intervention contributes
to changes in outcomes and how it does so. Causal
explanation is theory based. Thus, research focused on
 CHAPTER 14  Intervention-Based Research 327

Phase IV:
Long-term follow-
up and/or
Phase III: replication
Test intervention in
clinical setting
Clinical questions: Phase II: comparing new
patient problems; Efficacy studies to treatment to
clinical observations; establish outcome current best
a new theory (etc) Phase I: effects and optimal treatment (usually
Proof of concept treatment via RCT)
safety conditions
Pre-clinical work estimates of effect
Clinical answers:
change clinical
practice (evidence
implementation
strategies)

Theory Factor Factor Situation Situation

Replication
development isolating relating relating producing

Figure 14-1  Model of the integral components of the role of nurses in pain management.
Note: RCT, randomized controlled trial. Adapted from Vallerand, A. H., Musto, S., & Polomano, R. C. (2011). Nursing’s role in cancer pain management.
Current Pain and Headache Reports, 15(4), 250–562; Vallerand, A. H., Musto, S., & Polomano, R. C. (2011). Nursing’s role in cancer pain management.
Current Pain and Headache Reports, 15(4), 250–562.

causal explanation is guided by theory, and the find-
ings are expressed theoretically. Researchers employ
a broad base of methodologies, including qualitative
studies, to examine the effectiveness of an interven-
tion. Qualitative approaches offer substantial value to
the development, feasibility, and validation of nursing
interventions. For example, Van Hecke et al. (2011)
used qualitative research in the development and vali-
dation of their nursing intervention for promoting life-
style changes in patients with leg ulcers.
Theory-based Interventions
Knowledge obtained through a synthesis of collected
information can be used to develop a middle-range
intervention theory. An intervention theory is explan-
atory and combines characteristics of descriptive,
middle-range theories and prescriptive, practice theo-
ries. A descriptive theory describes the causal pro-
cesses that are occurring. A prescriptive theory
specifies what must be done to achieve the desired
effects, including (1) the components, intensity, and
duration required; (2) the human and material resources
needed; and (3) the procedures to be followed to
produce the desired outcomes.
A theory or model can be used to guide the devel-
opment of an intervention as well as provide direction
in the design of the study and testing procedures. The
theory itself should contain conceptual definitions,
propositions linked to hypotheses, and any empirical
generalizations available from previous studies
(Rothman & Thomas, 1994; Sidani & Braden, 1998).
Theoretical constructs serve as frameworks for many
nursing intervention studies, for example, risk
reductions strategies for cardiovascular disease
(Gholizadeh, Davidson, Salamonson, & Worrall-
Carter, 2010), improving quality of life with pressure
ulcers (Gorecki et al., 2010), and goal-directed thera-
pies for rehabilitative care (Scobbie, Wyke, & Dixon,
2009). Kolanowski, Litaker, Buettner, Moeller, and
Costa (2011) derived their activity interventions for
nursing home residents with dementia from the Need-
Driven Dementia–Compromised Behavior Model for
responding to behavioral symptoms. Their random-
ized double-blind controlled trial involving cogni-
tively impaired residents also linked theory-based
underpinnings to the behavioral outcome measures for
the study.
Not all interventions are examined in the context
of theories and theoretical propositions, but some
researchers contend that they should, especially when
frameworks can be used to help refine interventions
and select outcomes (Li, Melnyk, & McCann, 2004).
328 UNIT TWO  The Research Process

Master’s and doctoral nursing students working in col- Box 14-2 Elements of Intervention
laboration with faculty researchers or project teams
Theory
should recognize the value of theory-driven interven-
tions in providing greater clarity to their studies and Problem or Issue
contributing to the science of intervention theory. Nature of the problem or issue
Deciding on the best theory to guide intervention Manifestations
research requires: (1) a thorough and thoughtful Causative factors
review and synthesis of the literature (see Chapters 6 Level of severity
and 19); (2) scholarly papers that discuss appropriate Variation in different patient populations
theory-based interventions in areas of research inter- Variation in different conditions
est; (3) interactive discussions among faculty, stu-
dents, and other nurses about options for an optimal Critical Inputs
theory to guide interventions; and (4) once a theory Activities to be performed
has been selected, conversations with the project team Procedures to be followed
to determine its application to the study. Amounts of the intervention elements
Theory-based interventions can be developed for (intensity)
broad populations of patients, such as those with Frequency of the intervention
chronic illnesses, and then be more specifically applied Duration of the intervention
to patients with diabetes. For example, Kazer, Bailey, Mediating Processes
and Whittemore (2010) used a theory-based interven-
tion for self-management of the uncertainty associated Stages of change that occur after the
with active surveillance for prostate cancer. These intervention
researchers discussed an uncertainty management Mediating variables that bring about
intervention from a theory-based perspective to formu- treatment effects
late approaches to care for men with prostate cancer. Hypothesized relations among mediating
Research and Theory for Nursing Practice is a specific variables
journal that seeks intervention-based research derived Expected Outcomes
from theoretical frameworks and perspectives. Aspects of health status affected
An intervention theory must include a careful Biopsychosocial (physical, psychological, social,
description of the problem the intervention will and spiritual)
address, the intermediate actions that must be imple- Timing and pattern of changes
mented to address the problem, moderator variables Hypothesized interrelationships among
that might change the impact of the intervention, outcomes
mediator variables that might alter the effect of the
intervention, and expected outcomes of the interven- Extraneous Factors
tion. Box 14-2 lists the elements of an intervention Contextual factors
theory that are applied to research processes. Models Environmental factors
of theories or frameworks help explain the relation- Patient characteristics
ships of concepts, interventions, and outcomes. When
Treatment Delivery System Resources
critically appraising intervention research, students
need to keep in mind that a major threat to construct Setting
study validity arises if a framework or model used to Equipment
guide a study and its intervention has no clear link to Intervener characteristics
the development and implementation of the interven-
tion and the interpretation of study findings. Evidence
as to how the framework has guided the study needs solely on empirical evidence from existing literature
to be threaded throughout the research report (see that provides scientific perspectives for different inter-
Chapter 7 for understanding the inclusion of frame- ventions. A thorough review and synthesis of prior
works in studies). research should let researchers know the basis for an
intervention and justify the study design. Several peer-
Scientific Rationale for Interventions reviewed nursing journals publish studies without
Not all interventions need to be derived from theory theoretical underpinnings or models depicting the
or conceptual frameworks. Interventions can be based relationships of interventions to outcomes. In such

cases, clarity in the scientific basis for the intervention
must be presented throughout the research report.
Some interventions might evolve from experiential
knowledge in clinical practice, nursing education, and
administrative leadership. These interventions might
not have a strong scientific foundation to support their
use. Investigators conducting studies on these types of
interventions need to make a case for the importance
of their research. These types of studies can be impor-
tant in formulating study methodologies, establishing
relevant outcomes for novel interventions, and advanc-
ing nursing science.
An example of an intervention with minimal scien-
tific rationale is evident in the early work of Schmelzer
and Wright (1993). These gastroenterology nurses
began a series of studies that examined the procedures
for administering an enema. At that time, they found
no research in the nursing or medical literature that
tested the effectiveness of various enema procedures.
Without scientific evidence to justify the use of various
procedures for administering enemas—such as the
amount of solution, temperature of solution, speed of
administration, content of the solution (soap suds,
normal saline, or water), positioning of the patient, or
measurement of expected outcomes or possible
complications—they were faced with relying on the
tradition of practice and clinical experience. Their first
study involved telephone interviews with nurses
across the country in an effort to identify patterns in
the methods used to administer enemas; however, this
study was unsuccessful in helping Schmelzer and
Wright (1996) validate any commonly used techniques
to establish guidelines for enema interventions.
In their next study, these researchers developed
their own protocol for enemas and pilot-tested it on
hospitalized patients awaiting liver transplantation. In
their subsequent study with a sample of liver trans-
plant patients, these researchers tested for differences
in the effects of various enema solutions (Schmelzer,
Case, Chappell, & Wright, 2000). Schmelzer (1999-
2001) then conducted a study funded by the National
Institute for Nursing Research to compare the effects
of three enema solutions on the bowel mucosa. Healthy
subjects were paid $100 for each of three enemas, after
which a small biopsy specimen was collected. These
researchers’ experiences illustrate how the lack of sci-
entific evidence for an intervention requires a series
of studies before a large-scale study can be launched
to answer important research questions.
Eventually, these programs of intervention-based
research with enemas led to a safety and effectiveness
study involving healthy volunteers that showed that
enemas using soap suds and tap water produced
CHAPTER 14  Intervention-Based Research 329
greater return in bowel evacuation but were more
uncomfortable than those using a polyethylene glycol
electrolyte solution (Schmelzer, Schiller, Meyer,
Rugari, & Case, 2004). Although the two former inter-
ventions resulted in a more effective response, a higher
degree of surface epithelium loss in the bowel was
confirmed on biopsy. These investigators concluded
that the risks and benefits of selecting an enema solu-
tion must be balanced with the desire to produce a
better response versus safety concerns, for example,
the resultant effects on the bowel and the patient.
Terminology for Intervention-
Based Research
Specific research terms and concepts have importance
to intervention-based research, and it is necessary to
understand these when critically appraising and con-
ducting this type of research. Master’s and Doctor of
Nursing Practice (DNP) students need to be able to
critically appraise individual intervention studies and
those synthesized in meta-analyses and systematic
reviews. PhD students have the preparation to be prin-
cipal investigators for various types of studies, includ-
ing intervention-based research. Thus, this section
discusses relevant terms to increase nurses’ expertise
in critically appraising and conducting intervention
research.
Efficacy versus Effectiveness
Fundamental terms that are often mistakenly inter-
changed with intervention-based research are efficacy
and effectiveness. Both words are used to describe an
effect of a treatment or intervention; however, these
terms are different and should not be used synony-
mously. Efficacy refers to the ability to produce a
desired, beneficial, or therapeutic effect, and it can be
determined only in controlled experimental research
trials in which study criteria are stringent. Random-
ized controlled trials (RCTs) involving the testing of
drugs or procedures are most often efficacy studies
designed to demonstrate, under optimal conditions,
the desired outcomes that a drug or procedure pro-
duces (Compher, 2010). Efficacy studies require strict
control of as many potentially confounding variables
as possible to allow measures of efficacy to be quanti-
fied. A treatment that is deemed “efficacious” is one
that produces good outcomes when subjected to highly
controlled, experimental conditions. Randomized,
placebo-controlled trials are required in drug develop-
ment before the U.S. Federal Drug Administration
(U.S. FDA) can approve the new drug.
330 UNIT TWO  The Research Process

TABLE 14-1  Comparison of Characteristics for inherent risks in applying ES parameters across diverse
Efficacy vs. Effectiveness Studies
fields of study because interpretations of ESs can vary.
Cohen (1988) identifies a small ES as 0.2, a medium
Criteria Efficacy Effectiveness ES as 0.5, and a large ES as 0.8. A small ES of 0.20
Purpose Test a question Assess effectiveness for an intervention means a 20% difference can be
Sample size Smallest adequate Large sample size expected between the intervention and comparison
sample size groups. Likewise, a 0.5 ES represents a 50% differ-
Study cohort Homogeneous Heterogeneous
ence attributed to the intervention. Knowledge of the
Population Study population Study population in
in tertiary care initial care setting
ES for any intervention is helpful and often necessary
setting for calculations of the sample size needed to provide
Eligibility Stringent or strict Minimally restrictive sufficient statistical power to detect a treatment differ-
criteria or relaxed ence. The nature of the ES also varies from one statis-
Outcomes Single or minimal Multiple outcomes tical procedure to the next; it could be the difference
outcomes in cure rates, a standardized mean difference, or a
Duration Short study Long study duration correlation coefficient. However, the ES function in
duration conducting a power analysis is the same in all proce-
Adverse event Minimal Comprehensive dures (Aberson, 2010; see Chapter 15 for a more
recording
detailed discussion of ES and power analysis).
Focus of Internal validity External validity
inference
Analysis Completer-only Intent-to-treat Placebo and Sham Interventions
analysis analysis A placebo is an intervention intended to have no
effect. However, a placebo generally looks, tastes,
Adapted from Piantadosi, S. (2005). Clinical trials: A methodologic smells, and/or feels like the test intervention or is
perspective (2nd ed) (p. 323). Hoboken, NJ: John Wiley & Sons; and experienced like the real study intervention. The
Gartlehner, G., Hansen, R. A., Nissman, D., Lohr, K. N., & Carey, T. S.
(2006). Criteria for distinguishing effectiveness from efficacy trials in
purpose of a placebo is to account for how study par-
systematic reviews. Technical Review, Agency for Healthcare Research and ticipants would respond without actually receiving the
Quality. (AHRQ Publication No. 06-0046.) Rockville, MD: U.S. active intervention. Sham interventions are often used
Department of Health and Human Services. with procedures, and are a variation of a “fake” inter-
vention that omits the essential therapeutic element of
the intervention. A “sham” intervention also attempts
Effectiveness of a treatment or intervention indi- to control for the placebo effect. Rates for placebo
cates that it is capable of producing positive results responses do differ depending on the types of inter-
in a usual or routine care condition. This term is gen- ventions and populations studied. For example, the
erally reserved for interventions studied in situations rate for a placebo response with symptom manage-
in which stringent controls on group assignment are ment research can be as high as 90% (Kwekkeboom,
not possible or the study might not include a placebo, 1997). Very complex psychobiological responses
control, or comparison group. With effectiveness occur in the brain, called the real placebo effect, even
studies, the new intervention may be compared with when an intervention is perceived to be inert or of no
the existing standard of care to ascertain whether the known therapeutic value (Benedetti, Carlino, & Pollo,
new treatment is better, worse, or the same as the 2011). It is important to note that designs using placebo
existing intervention. These studies still require con- or sham interventions are typically carefully evaluated
trols in the design and execution of the interventions by institutional review boards (IRBs) to ensure that
but are not nearly as stringent as occurs in RCTs. patients’ rights are not violated.
Effectiveness studies most often use real-world clini- The placebo effect is extremely complex and often
cians and patients available to researchers. Table attributed to multiple variables. These might include
14-1 contrasts the characteristics of efficacy and attention that a participant receives in a study even if
effectiveness. there is no structured intervention or documented
therapeutic value. The influence of the nurse or physi-
Treatment Effect Size cian demeanor, patient factors such as motivations to
The treatment effect size (ES) refers to the magnitude do better, and the patient-provider relationship can
of effect produced by the intervention. Cohen (1988) have a placebo effect. Other factors, such as
and Aberson (2010) provide parameters for qualifying co-interventions that cannot be controlled, expecta-
and quantifying the ES but caution that there are tions of a treatment effect, the context in which an

intervention is delivered, relief associated with receiv-
ing some treatment, and spontaneous improvement of
problem during a study, are also likely explanations
for placebo effects.
A placebo or sham intervention is intended to spe-
cifically blind study participants and/or research per-
sonnel to the treatment conditions in intervention
trials. Placebos also blind healthcare providers taking
care of or interacting with study participants, who
might introduce bias into the study if study conditions
are known. Placebos are most frequently used with
RCTs involving drug testing.
Sham procedures are often used in research
focused on interventions such as healing touch, relax-
ation therapy, acupuncture, cognitive and behavioral
therapy, or any treatment that might produce an effect
just from the intervener/participant interactions. To
account for the benefits of paying attention to partici-
pants or doing something to them that might in and of
itself lead to a positive response, researchers some-
times use sham techniques to control for the con-
founding effects of a treatment. For example, Baird,
Murawski, and Wu (2010) studied the efficacy of
guided imagery and relaxation on pain reduction,
improvements in mobility, and decreased over-the-
counter medication use in older adults with osteoar-
thritis. These investigators provided a sham condition
with planned relaxation provided to the comparison
group to control for any positive influence on out-
comes that might be caused by simply engaging
patients in a study. With this type of design, the
investigators were able to detect the true effects of
the experimental condition under investigation,
guided imagery and relaxation. It is always important
that the design of a sham technique be similar in
many respects to the experimental treatment, such as
the circumstances surrounding the encounter with the
researcher or intervener, duration of the encounter,
and implementation of procedural steps or sequenc-
ing of the mock treatment.
Blinding versus Open Label
The term blinding refers to preventing disclosure of
group assignment while a study is being conducted to
avoid bias or undesirable influence in the ways par-
ticipants respond and researchers perform. According
to the CONSORT (CONsolidated Standards of Report-
ing Trials, 2011) Transparency Reporting of Trials
statement, the term “blinding” or “masking,” which is
gaining greater acceptance as the more appropriate
designation, “refers to withholding information about
the assigned interventions from people involved in
the trial who may potentially be influenced by this
CHAPTER 14  Intervention-Based Research 331
knowledge.” Procedures for blinding or masking in
the conduct of clinical trials are of utmost importance
to the validity of findings and estimates of treatment
effects. One reason is that participants may respond
differently on study outcomes if they have knowledge
of their exposure to a treatment condition. Along
these same lines, investigators and data collectors
can unknowingly or knowingly encourage certain
responses from participants in favor of one treatment
over another if they are aware of a patient’s treatment
group assignment. Those caring for patients in studies
could inadvertently share information about treat-
ments that might bias participant performance on
outcome measures. Some study conditions even stipu-
late that knowledge of group assignment be kept from
data analysts or statisticians to safeguard against bias
in the analysis and interpretation of data.
Blinding or masking occurs at various levels.
Although this procedure helps to reduce bias in favor
of one intervention over another, only 33% of 199
published RCTs from 2007 to 2009 published in 16
nursing journals reported using blinding procedures
(Polit, Gillespie, & Griffin, 2011). With single-blinded
studies, there are two variations. Study participants are
not aware of their treatment assignment, but investiga-
tors and research staff may have knowledge of who is
in what group. The reverse may be used when partici-
pants have knowledge of their treatment group assign-
ment, but this information is not available to the
investigators and research staff. For example, children
involved in research comparing the calming effects of
music alone with those of music with mother’s voice
delivered by audiotaped recordings would recognize
the contents of the tape if they were alert. Researchers
and data collectors may not be told of participant
group assignment. Double-blinded studies keep
knowledge of study conditions from study partici-
pants, investigators, and research staff. When a study
is triple-blinded, the participants, researchers, and
those involved in data management are unaware of
group assignment.
Open-label extension is a term that usually applies
to RCTs or other clinical trials whereby participants
in the active treatment group are offered an opportu-
nity to continue treatment and those in the placebo or
control group are given the option to transition to the
active treatment at the conclusion of the intervention
period. This phase of a clinical trial is often used in
drug studies as a continuation phase to obtain safety
data about the frequency of adverse events or side
effects over an extended period. This aspect of the
study is required by the FDA for approval of new
drugs. Participants are closely monitored by research
332 UNIT TWO  The Research Process
coordinators and study personnel, and data are often
still collected on the study primary end points to dem-
onstrate whether treatment effectiveness continues.
Important attention needs to be paid to the side effect
profile of the drug that develops over time. Other treat-
ment trials may use open-label extensions to accumu-
late more data about the treatment and its effect on
study outcomes.
Treatment Fidelity
Treatment fidelity has to do with the accuracy, con-
sistency, and thoroughness of how an intervention is
delivered according to the specified protocol, treat-
ment program, or intervention model. Strict adherence
to treatment specifications must be evaluated on an
ongoing basis during the course of a study. Thus, inter-
vention fidelity is “the adherent and competent deliv-
ery of an intervention by the interventionist as set forth
in the research plan” (Santacroce, Maccarelli, & Grey,
2004, p. 63), and this fidelity is of utmost importance
to the inference of the study’s internal validity in
intervention-based research. Stringent controls for
implementation of study procedures are critical to the
study’s integrity. Methodological approaches to treat-
ment fidelity include education and training of all
persons implementing the treatment, periodic moni-
toring or surveillance of the implementation of the
treatment or fidelity checks (either conspicuous or
inconspicuous observation), and retraining and reeval-
uation of study research staff if deviations from the
prescribed protocol for study procedures are found.
Gearing et al. (2011) propose a scientific guide to
treatment fidelity and discuss implications at all phases
of the treatment, including: (1) implementing the
design; (2) training the research staff; and (3) monitor-
ing the delivery and receipt of the intervention.
Treatment fidelity is important in the conduct of all
intervention-based research, but it is especially critical
for complex interventions such as those involving
information dissemination, social support, counseling,
and other interventions intended to bring about
changes in health behaviors and psychosocial out-
comes. Radziewicz et al. (2009) reported their experi-
ences maintaining treatment fidelity in a study testing
a communication support by telephone intervention
for older adults with advanced cancer and their family
caregivers. These researchers contended that fidelity
was maximized by the following: ensuring that the
intervention is congruent with relevant theory, stan-
dardizing the training, using stringent criteria for con-
ducting fidelity checks to determine interventionist
competence, monitoring the delivery of the interven-
tion, and carefully documenting findings. “The degree
to which internal validity can be established affects
the conclusions’ accuracy drawn from the interven-
tion” (Radziewicz et al., p. 194).
Best practices in constructing treatment fidelity cri-
teria and executing sound methodological procedures
have been summarized for behavioral research (Bellg
et al., 2004; Borrelli et al., 2005). According to Bellg
et al. (2004), specific goals should direct fidelity
checks. The goals include: (1) ensuring same treatment
dose within conditions; (2) ensuring equivalent dose
across conditions; and (3) planning for implementa-
tion setbacks. Investigators should make provisions to
accomplish these goals while formulating criteria by
which fidelity or adherence to treatment protocols can
be assessed. Often a sampling parameter is set prior
to initiation of the study that a certain percentage of
study intervention episodes will be evaluated. Manipu-
lation checks are also a critical part of ensuring the
integrity of study procedures. These types of checks
are valuable in gathering information about circum-
stances and study conditions that might interfere with
or impede implementation of the study intervention.
Box 14-3 shows an example of a manipulation check
used by interventionists in a study of activity interven-
tions for nursing home residents with dementia (Kola-
nowski et al., 2011). This type of checklist is also
useful to investigators in determining intervention fre-
quency, dose intensity, and protocol deviations. This
information not only is critical in recording adherence
to the study intervention but also might be included
in the analyses of data to separate out or account for
variations in exposures to the intervention.
Dose-Intensity of the Intervention
The exposure to an intervention can influence the
response to it; therefore, the dose-intensity of the
intervention must be carefully monitored and tracked.
Dose-intensity refers to the amount of the intervention
delivered in terms of the (1) components of the inter-
vention; (2) duration of a single session for the inter-
vention; (3) the frequency with which the intervention
is delivered (e.g., daily, times per week, times per
month); and (4) cumulative intervention intensity
(number of treatments received and duration). The
latter is the product of dose × dose frequency × total
intervention duration (Breit-Smith, Justice, McGinty,
& Kaderavek, 2009). The ability to capture, quantify,
and report the exposure to an intervention in the
context of a single study enhances the ability to draw
conclusions about how effective an intervention might
be on the basis of the quantity delivered. Reed et al.
(2007) provide a comprehensive review of ways to
measure the dose of nursing interventions and offer

Box 14-3 Example of a Manipulation Check for a Study of an Activity Intervention for
Nursing Home Residents with Dementia
Subject Code: _____________
Date: _____________________
Interventionist: _______________
At the completion of each activity session,
please evaluate the extent to which the activity
was implemented by answering the following
questions:
1. Was today’s activity the one selected for this
condition?
a. Yes ___
b. No ____
Explain:
2. Were there any extraneous circumstances that
influenced the delivery of the activity?
_____ a.  Inability to form group activity.
Explain:
_____ b.  Inability of study participant to stay
by himself or herself for independent
activity.
Explain:
Kolanowski, A., Buettner, L., & Moeller, J. (2006). Treatment fidelity plan for an activity intervention designed for
persons with dementia. American Journal of Alzheimer’s Disease and Other Dementias, 21(5), 326–332.
formulas for calculating intervention exposure in
terms of metrics for frequency and duration of the
intervention.
For repeated measures designs involving interven-
tions delivered over time, research staff must maintain
accurate records of all intervention sessions for all
study participants, including completed sessions (the
full intervention was delivered), partial sessions (only
a part or parts of the intervention were administered),
and missed sessions (the intervention was not admin-
istered at the appropriate time point in the study).
Field notes should be kept to explain reasons for the
inability to deliver the total intervention. Maintaining
detailed information regarding dose-intensity is essen-
tial for determining the treatment effect. If known,
varying levels of the exposure to the intervention can
be handled in the statistical analyses to separate out
the magnitude of benefits on the outcome(s) if partici-
pants received full exposure to the intervention or only
partial exposure.
Numbers Needed to Treat
Numbers needed to treat (NNT) is a metric that is
defined as the number of patients who would need to
be treated with the new intervention to avoid one event
CHAPTER 14  Intervention-Based Research 333
Facility: _______________________
Time: __________________________
_____ c.  Interference from staff, other
residents, visitors, etc.
Explain:
_____ d.  Subject uncooperative.
Explain:
_____ e.  Subject ill.
Explain:
_____ f.  Subject engaged in another activity.
Explain:
_____ g.  Subject not available.
Explain:
_____ h.  Interventionist not available.
Explain:
that might have occurred with standard treatment. It
is necessary that NNT be estimated in context of those
deriving a benefit or those who might be harmed by
the intervention (Day, 2007). Intuitively, it is an easy
concept to grasp because it is expressed in a numeric
value. For example, if the NNT is 4 for a benefit from
an intervention, then a prescriber of an intervention
could expect that 1 out of 4 patients treated with the
intervention would derive a positive outcome. The
NNT is determined by an analysis of the data follow-
ing the completion of a study that is adequately
powered (i.e., the sample size is adequate) to detect
treatment differences. The researcher establishes cri-
teria for the primary outcome measure for the study
that is indicative of a positive or undesirable response,
and the NNT is expressed in terms of either outcome.
Researchers are often required to calculate the NNT
for clinical trials prior to submitting a manuscript for
publication to influential research journals. NNT
values are often pooled and analyzed in meta-analyses
and systematic reviews of studies. Moreover, the NNT
is a meaningful parameter for practice because it
informs clinicians of the overall effectiveness of the
intervention and of the expected response when pre-
scribing an intervention.
334 UNIT TWO  The Research Process
When one is interpreting NNT, it is important to
determine whether the NNT value is for a single
administration of the intervention, at a single time
point, or overall across multiple time points. For
example, Hatfield, Gusic, Dyer, and Polomano (2008)
conducted a blinded, placebo-controlled trial of oral
sucrose administration for pain associated with immu-
nizations in infants 2 and 4 months of age. With use
of a repeated measures design, infants receiving 24%
oral sucrose (n = 38) or sterile water (placebo) (n =
45) via a pacifier were assessed for pain at baseline
and 2, 5, 7 and 9 minutes after a series of vaccine
injections. Pain was measured using the University of
Wisconsin Children’s Hospital (UWCH) Pain Scale (0
denoting the absence of any pain response to 5 reflect-
ing the highest degree of the pain response). These
investigators assigned a pain score of 2 or less as the
desirable criterion for a favorable response to the treat-
ment. At 2 minutes the NNT was 4, and at 9 minutes,
the end of the study, it was 2. This can be interpreted
that at 9 minutes after administration of oral sucrose
with a sequence of vaccine injections, 1 in 2 infants
(50%) received a beneficial response to the therapy;
whereas earlier, at 2 minutes, the intervention was not
as effective with only 1 in 4 infants (25%) achieving
the desired level of response.
Analysis of Intention-to-Treat
Intention-to-treat (ITT) is an analysis based on the
principle that participant data are analyzed according
to the groups into which they were randomly assigned
regardless of what happens to them in the study. In
other words, if a participant drops out or changes
therapy after randomization, this individual would be
retained in the data set and in the group to which he
or she was randomly assigned in the analysis. If any
participants are not included in the analysis, then the
equalization of the treatment groups is no longer main-
tained, possibly introducing bias into the results.
Nurses need to understand the concept of intention to
treat, especially because this analysis strategy is
widely used by healthcare researchers.
There are several ways to account for the missing
data for subjects who leave a study earlier than
expected. It is important, however, to understand that
all methods of imputing data values create “made-up”
data. The ideal clinical trial or intervention-based
study must be designed to minimize participant
dropout and missed interventions to limit as much
missing data as possible. There are many ways to
impute data. They have a variety of advantages and
disadvantages, which have been well described in a
National Academy of Science report (National
Research Council, 2010). Two of the more commonly
used methods are baseline observation carried forward
(BOCF) and last observation carried forward (LOCF).
BOCF populates the missing data points for an
outcome following the last data collection value
obtained with the baseline value obtained for that par-
ticipant. Use of this procedure assumes that the par-
ticipant would have returned to baseline without
further treatment. LOCF, on the other hand, assumes
that the last value obtained for a particular outcome
for a participant would continue for the rest of the
study. A variation on these techniques is termed the
worst observation carried forward (WOCF), which
takes the worst value for the participant no longer in
the study and carries this value forward for all the
missing data collection points for the study outcomes.
Sometimes the preferred method(s) for observation
carried forward in drug development studies is stipu-
lated by the FDA drug approval section. For example,
the FDA guidance for studies of pain stipulates that
the more conservative BOCF analysis be used for all
regulatory submissions for drug approval (National
Research Council, 2010). There are no known guide-
lines for nursing research studies using intention-to-
treat designs.
A review of 71 clinical trials published in high-
impact medical journals in the late 1990s found that
63 (89%) had incomplete data sets owing to missing
data, and for 13 of these trials, outcomes data were
missing for more than 20% of participants (Wood,
White, & Thompson, 2004). In 37 trials using repeated
measures to obtain outcome measurements over time,
46% performed the analysis after excluding those with
only some follow-up data, 19% of the trials used
LOCF, and 11% imputed missing data with the worst
case value (WOCF). LOCF is less conservative,
assuming that participants who drop out may have had
some benefit before dropping out, and has been
favored as an analytical strategy for missing data;
however, in studies of pharmacotherapy for chronic
pain, for example, there are circumstances in which
BOCF may be more appropriate and scientifically jus-
tified (Dworkin, et al., 2009). The imputation of
missing data allows the application of intention-to-
treat analysis for clinical trials. It is vital that the
method selected as the strategy of handling missing
data be determined and specified a priori (before the
study is initiated) to maximize the likelihood of
obtaining a valid conclusion from the data analysis.
Responder Analysis
Responder analysis is an additional technique to iden-
tify variables associated with a beneficial response

to an intervention or to determine outcomes of poten-
tial importance to the treatment. This technique can be
performed when one is analyzing the effect of the
intervention or during a secondary analysis of study
data after the primary data have been reported. A
responder analysis can also be conducted on more than
one study using pooled data from investigations with
similar designs and outcomes. This type of analysis
informs researchers and clinicians about the significant
variables contributing to participants’ response or lack
of response to a particular intervention. Responder
analyses provide useful information about sample
characteristics and conditions likely to lead to signifi-
cant results for researchers conducting similar future
studies. Criterion-based parameters on outcome
measures can also be determined to define what is
considered to be a positive or beneficial response to a
treatment.
Clinicians often relish these types of studies to
guide practice in selecting those individuals who
might benefit most from an intervention and to quan-
tify levels of responses on specific instruments.
Weaver, Chasens, and Arora (2009) conducted a
responder analysis with two randomized, double-
blind, placebo-controlled studies examining the effi-
cacy of modafinil, a drug used to treat excessive
sleepiness in individuals with sleep disorders. The
response to treatment with modafinil in these two pre-
vious studies was measured with the Functional Out-
comes of Sleep Questionnaire (FOSQ). The FOSQ
was administered to patients with obstructive sleep
apnea who were experiencing residual excessive
sleepiness with continuous positive airway pressure
(CPAP) use. A secondary data analysis using a covari-
ance model showed that a greater proportion of
responders were in the active drug arm, and functional
outcomes indicative of a beneficial response were
characterized on the basis of improvements for certain
items and domains on the FOSQ. These findings will
help both researchers and clinicians better understand
the items and subscales on the FOSQ that are most
important in capturing improvements with modafinil
in this patient population.
Types of Research Designs
Experimental or quasi-experimental research designs
are required if investigators want to test the efficacy
or effectiveness of nursing interventions. The effects
of interventions can also be examined, evaluated,
described, or observed in the context of nonexperimen-
tal designs such as descriptive, correlational, or obser-
vational research designs. Models and descriptions of
CHAPTER 14  Intervention-Based Research 335
both experimental and nonexperimental quantitative
research designs are presented in Chapter 11. A quali-
tative research design may be used if a researcher is
in the initial phase of developing a model or prototype
for an intervention. This type of research methodology
might also be used in later phases, when the interven-
tion is implemented and the experiences encountered
with the intervention from the perspectives of those
who receive or deliver it are of interest (see Chapter
12, which is focused on qualitative research). Mixed-
methods studies that combine both quantitative and
qualitative research methods are of great value when
the interventions are complex or various aspects of the
interventions must be validated (Brady et al., 2011;
Creswell & Zhang, 2009). Chapter 10 explains differ-
ent types of mixed-method designs.
In the context of interventional research, one way
of examining research designs is to focus on the types
of clinical trials and the design strategies used to test
interventions. From a generic view, a clinical trial
involves the systematic study of the effects of a treat-
ment in human subjects (Day, 2007). There are many
types of clinical trials, and many levels of scientific
rigor are applied in studying treatment effects. Duffy
(2006) provides a basic overview for types of clinical
trial designs and also emphasizes the growing use of
these designs in nursing research. Some of the clinical
trial designs discussed here are gaining greater accep-
tance for studying nursing interventions. Additionally,
nonexperimental approaches can be used to observe
and describe outcomes that result from nursing inter-
ventions, and these also need to be considered as part
of intervention research.
Experimental and
Quasi-experimental Designs
Parallel Group Randomized Controlled Trials
Parallel group RCTs represent the most common or
standard type of RCTs. Subjects are randomly assigned
to receive either one or two or more concurrently
administered treatments (Piantadosi, 2005). This
design is experimental in nature and offers a high
degree of scientific rigor in determining the efficacy
(if placebo control is used) of a treatment or the effec-
tiveness if the comparison group includes another
treatment and/or a control group receiving no treat-
ment, just routine care. RCTs can be costly and diffi-
cult to conduct because of the complexity of the design
and stringent controls that must be in place to ensure
the validity of results. An integrative review of pub-
lished nurse-led RCTs revealed several challenges
encountered by nurse researchers. They included: (1)
 336 UNIT TWO  The Research Process
difficulties with sufficient patient recruitment; (2) non-
adherence to research protocols; and (3) economic and
organizational barriers to conducting these trials
(Vedelø & Lomborg, 2011). Nonetheless, RCTs are
vitally important to advancing nursing science and
evidence-based practice because they yield the stron-
gest level of evidence for single studies. When pos-
sible, both nurses in clinical practice and students need
to participate in nurse-led RCTs as co-investigators
and research assistants.
Crossover Randomized Controlled Trials
A crossover randomized controlled trial exposes each
subject to both treatment conditions, which can be a
placebo and an active intervention or two active inter-
ventions. This type of design is used to determine how
each subject responds to both study conditions, and
it enables investigators to examine each subject’s
responses under these study conditions. As such, par-
ticipants serve as their own controls. This design
maximizes enrollment because fewer subjects are
needed in crossover than in parallel RCTs. This design
is typically executed using a repeated measures design,
whereby time is the variable under which within-
subject variability is examined. Study conditions need
to be delivered in random order so that the influence
of sequence of exposure to the intervention is con-
trolled and does not result in a confounding variable.
This approach lessens the likelihood that exposure to
one study condition first would influence the response
to the second condition, and vice versa.
A washout period (or time) is incorporated into
crossover designs, which occurs between exposures
to study conditions. The purpose of this interval is to
provide ample time for any effects of first exposure to
the active intervention or placebo to dissipate. Conse-
quently, each study condition can be examined sepa-
rately with little risk that diffusion of their effects
oasis-ebl|Rsalles|1476144266
may contaminate the outcomes. During this washout
period, subjects do not receive any intervention;
however, they remain in the study and may also be
monitored or asked to complete study outcome mea-
sures. Overall, crossover designs are thought to offer
greater precision in conducting comparisons of treat-
ments (Matthews, 2000). A major disadvantage is that
subjects are often required to stay in the study longer
and are exposed to two and not just one study
condition.
Cluster Randomized Controlled Trials
Intervention-based research can often be costly, and
gaining access to representative samples in multiple
study sites can be time consuming and difficult.
Further, interventions that may be highly influenced
by existing clinical practices, which cannot be con-
trolled, are challenging to study. Cluster randomized
controlled trials or group randomized trials employ a
randomized sampling method to more efficiently test
interventions using representative groups of samples
randomized to study conditions rather than individu-
als. In this design, the unit of analysis is the research
site, and investigators randomly select research set-
tings from a sampling frame of eligible and willing
sites, which would then be assigned to serve as test
sites receiving either the experimental intervention
or the control condition. This type of random selec-
tion involving clinical settings can more efficiently
accomplish participant accrual, control for confound-
ing variables introduced by site-specific practice
variations, and ensure more representative groups or
clusters. For example, Pellfolk, Gustafson, Bucht,
and Karlsson (2010) selected 40 group dwelling units
of individuals with dementia to examine the impact
of a 6-month educational program in restraint reduc-
tion offered to the nursing staff. Because their study
involved both nurses and patients, nursing units were
randomized for nursing staff members either to
receive a 6-month educational intervention or to
serve as the controls. Nurses working on these units
participated in the study, and outcomes from their
patients were collected and analyzed. Knowledge
and attitudes regarding restraint practices improved
for the educational intervention group, and this
improvement translated into a reduction in restraint
use in comparison with practices by the staff in the
control group.
Preference Clinical Trials
With most clinical trials, participants are randomly
assigned to treatment or control conditions. However,
in some situations, patient preference takes prece-
dence over randomization to maximize enrollment.
Understandably, preference for treatment conditions
plays an important role prior to randomization and has
significant implications for outcomes. A meta-analysis
of clinical trials for musculoskeletal conditions dem-
onstrated that patients who were randomly assigned to
their preferred treatment had a standardized ES of the
treatment that was greater than those who were indif-
ferent to their group assignment (Preference Collab-
orative Review Group, 2009). Although not found to
be statistically significant in this study, a trend emerged
showing that those who received their preferred treat-
ment did better on outcome measures than participants
not assigned to their preferred treatment. Outcomes
were similar for participants allocated to their

undesired treatment and those who were indifferent to
the treatment arm.
One way to handle participant preference or partial-
ity for treatment conditions is through the use of pref-
erence designs that allow an option for patients to
receive the treatment of their choice (Day, 2007).
These types of designs include Wennberg’s design and
Zelen’s randomized consent design. In the latter,
patients are randomly assigned to treatment conditions
and informed of the group assignment prior to consent
to participate. If they decline, patients are offered the
alternative treatment option(s) of their choice. This
procedure is used to prevent the biasing of the results
of clinical trial associated with “resentful demoraliza-
tion” (unhappiness with the group assignment) and the
Hawthorne effect (responses that result when study
participants change their behavior because they are
being observed and are not a true reflection of the
intervention) (Adamson, Cockayne, Puffer, & Torger-
son, 2006). Thus, in preference clinical trials, rather
than being randomly assigned to subject groups,
patients choose among all treatments available. With
preference clinical trials, researchers blend both
experimental and quasi-experimental design strategies
in the context of a single study.
Treatment Matching Designs
Treatment matching or client treatment matching
(CTM) compares the relative effectiveness of various
treatments. Treatment matching designs are used
when the following conditions are met: (1) there is no
clearly superior treatment for all individuals with a
given problem; (2) a number of treatments with some
proven efficacy have comparable effectiveness for
undifferentiated groups of subjects; and (3) there is
evidence of differential outcomes, either within or
among treatments, for defined subtypes of patients
(Donovan & Mattson, 1994; Donovan et al., 1994).
Unlike sample matching techniques to ensure equal
distribution of sample characteristics across treatment
groups, treatment matching designs align interven-
tions with predetermined criteria that will ensure the
best possible benefit to subjects. Thus far, CTM inter-
vention studies have been successful, albeit limited to
studies involving treatment approaches for substance
abuse and addiction (Melnick, De Leon, Thomas, &
Kressel, 2001). Faul et al. (2011) have suggested
that this strategy be used in studies of exercise therapy
for patients with cancer in order to consider pre-
intervention activity levels as criteria for matching
patients to the most appropriate type of exercise pro-
grams to improve participants’ functionality and
overall quality of life.
CHAPTER 14  Intervention-Based Research 337
Open-Label Designs
Open-label designs refers to the absence of any blind-
ing or masking procedures for the intervention. For
obvious reasons, open-label clinical trials, or other
intervention-based designs in which the intervention
groups are known, must be used when it is impossible
to blind or mask the intervention. In such cases,
researchers, research participants, and healthcare pro-
fessionals caring for research participants have knowl-
edge of the intervention. Research participants can
still be randomly assigned to groups of intact cohorts
receiving an intervention as part of routine clinical
care. Regardless of the design, it is critical for research-
ers to take measures to ensure that investigator or
participant bias does not interfere with the validity or
credibility of the results. Safeguards to control for bias
and improve internal validity of the study include: (1)
clear specifications for the implementation of the
study protocol or treatment; (2) education and training
of research assistants or nurses implementing the
interventions; (3) establishing interrater reliability for
the implementation of study procedures and interob-
server reliability for measurement of the study
outcome(s); and (4) separation of study participants to
avoid experimental diffusion of the treatment effect,
which may occur when study participants interact with
one another. As with all intervention-based research,
any violations in execution of study procedures or data
collection should be immediately communicated to
the investigators, however minor these may be.
Nonexperimental Designs
Observational Studies
The effects of nursing interventions can be studied
through observation. Although observation is also
used in experimental research designs, nonexperimen-
tal observational designs can be used to assess patient
responses and outcomes to the implementation of new
or existing nursing practices; new technology or
equipment; or specific prevention protocols to reduce,
for example, occurrences of falls, hospital-acquired
pressure ulcers, and catheter-associated bloodstream
infections. Observational studies are generally carried
out in the naturalistic setting in the course of routine
clinical care. Methods for data collection might
include surveillance, extraction of clinical data from
electronic health information systems or paper health
records, compliance monitoring by observers, and
data storage or output from monitoring technology or
devices.
For example, Shever (2011) conducted an observa-
tional study of the impact of nursing surveillance on
338 UNIT TWO  The Research Process
failure to rescue, a nursing-sensitive indicator related
to surgical death from preventable complications.
Data were extracted from various electronic clinical
information repositories and a nursing documentation
system based on the Nursing Interventions Classifica-
tion (NIC). A high level of nursing surveillance of
patients was quantified as an average of 12 times per
day or more. Patients receiving low and high nursing
surveillance were evaluated, and when nursing
surveillance was high, there was a significant decrease
(p = 0.006) in the likelihood of experiencing failure to
rescue.
Retrospective Descriptive Studies
Not all interventions occur in prospective studies, so
retrospective designs are employed to examine the
effects of interventions. In a retrospective study of
nursing practice across three community nursing sites,
Schoneman (2002) compiled data on nursing interven-
tions using the computerized Omaha Classification
System (OCS) to examine “surveillance” as a nursing
intervention. The practice of surveillance was identi-
fied as a significant nursing intervention accounting
for 27.1% of all nursing interventions. Of the 1506
clients studied, those most likely to receive surveil-
lance were more than 40 years of age (p = 0.001) and
female (p = 0.001). The primary reasons for this inter-
vention were to promote health and prevent disease.
Specifically, subjects with a nursing diagnosis of cir-
culation and nutrition deficits seemed to be assessed
more frequently with surveillance nursing practices.
Planning Intervention Research
Project Planning
Because an intervention research project often involves
a study extended over a long time, sometimes years,
nurse researchers are advised to give careful thought
to the composition of their research teams. Research-
ers need to determine (1) who will be included on the
project team; (2) what level of expertise is required
for team members; (3) how team members will func-
tion together; and (4) the roles team members will
assume.
Forming a Project Team
Because of the nature of intervention research, you
may need to gather a multidisciplinary project team to
facilitate distribution of the work and a broader gen-
eration of ideas. If both quantitative and qualitative
data will be gathered during the research project, your
team should include members experienced in various
qualitative and quantitative data collection and analy-
sis approaches. Including a team member with market-
ing expertise will be beneficial, because the final step
of the project will be to market the intervention. Teams
are enhanced by the inclusion of undergraduate, mas-
ter’s, and doctoral nursing students.
Recruiting colleagues located in other areas of the
country or the world for the research team can add an
important dimension by permitting multisite interven-
tion studies. To achieve this goal, investigators need
to: (1) contact researchers with similar interests; (2)
attend specialty conferences related to the research
area, during which they can speak with researchers
and possibly extend an invitation to participate in the
project; (3) invite colleagues to join the project after
presentations at a professional meeting; (4) develop a
project website that invites other researchers to par-
ticipate; and (5) develop or participate in an internet
mailing list (Listserv) or a blog related to the topic.
The process of developing a team is dynamic rather
than static, with changes occurring as development of
the research program continues.
Work of the Project Team
There is almost always a core group in a project team
that carries on most of the work, maintains group
activities, and encourages the achievement of tasks.
However, other people can contribute in lesser ways
to benefit the project. For example, you may want to
establish liaison groups from the clinical facilities in
which the intervention will be studied. In some cases,
the addition of other advisory groups can be helpful.
The initial focus of the team is to clarify the
problem or issue of interest. In analyzing identified
problems or issues, the team should answer the ques-
tions listed in Box 14-1. Considering these questions
may provide new insights that redefine the problem or
issue and may lead to a more effective intervention.
Sidani and Braden (1998) have cautioned the project
team to be alert to the risk of making a Type III error.
A Type III error is the risk of asking the wrong
question—a question that does not address the problem
of concern. This error is most likely to occur when the
researchers do not thoroughly analyze the problem
and, as a result, have a fuzzy or inaccurate understand-
ing of the issue of concern. The solution, then, does
not fit the problem. A study conducted on the basis of
a Type III error provides the right answer to the wrong
question, leading to the incorrect conclusion that the
newly designed intervention will resolve the problem.
Once the problem or issue to be examined is clari-
fied, you must establish your goals and objectives.
Project team tasks include: (1) gathering information;
 CHAPTER 14  Intervention-Based Research 339

(2) developing an intervention theory; (3) designing Box 14-4 Topics for Information
the intervention; (4) establishing an observation
Gathering
system;, (5) testing the intervention; (6) collecting and
analyzing data; and (7) disseminating the intervention. Nature of the Problem (Actual or Potential)
Seeking funding for the various studies of the project or Issue
will be an ongoing effort. Manifestations
Causative factors
Gathering Information Level of severity
Once a clinical problem or issue has been identified, Variation in different patient populations
the investigators and members of the research team Variation in different conditions
can begin by conducting an extensive search for
information related to the project. This search is con- Intervention
siderably more focused and encompassing than the How people who have actually experienced
traditional literature review. Box 14-4 lends some the problem or issue have addressed it
organization and clarity to topics that need to be Previous interventions designed to address the
addressed as the basis for developing a plan for inter- problem or issue
vention research. Several sources of information can Unsuccessful interventions
be accessed to gain knowledge of existing nursing Value to target population
interventions, extant research conducted on these Sensitivity to cultural diversity
interventions, and evidence-based practice guidelines Biases or prejudices
that incorporate these interventions and guide their Processes underlying the intervention effects
application to practice (Box 14-5). It is important to Intervention actions
keep in mind that researchers should not “reinvent the Components
wheel.” For many nursing interventions, prescriptive Mode of delivery
procedures are already published in credible scientific Strength of dosage
sources that can be adapted for the purposes of inter- Amount
vention testing. Undergraduate, master’s, and doctoral Frequency
nursing students as well as clinicians working with Duration
the project team can be very helpful in retrieving lit- Mediating Processes
erature and resources and assisting with synthesizing
information. Patient characteristics
Setting characteristics
Intervener characteristics
Design and Testing of Expected Outcomes

Interventions Contextual factors

Researchers either develop their own procedures for
implementing an intervention or replicate an interven-
tion and its protocol from an existing study. As men-
tioned earlier, theory-driven intervention research is
conceived from theoretical or conceptual perspectives
that form the basis of the intervention and its per-
ceived effectiveness or benefit. Theory-based inter-
vention research can begin with a concept analysis of
the problem or the intervention intended to address it.
By dissecting the components of problems and plau-
sible solutions, investigators can gain clarity of the
issues at hand. At the other end of the spectrum are
replication studies that involve similar intervention
designs, approaches, and outcomes used by other
investigators. These types of studies are especially
important to validate or refute results from published
research. Replication studies add to the strength of
Environmental factors
Patient characteristics
Provider factors
Healthcare system factors
using specific interventions in selected populations
and settings.
A study by Forbes (2009) provides an excellent
model of the four phases of testing clinical interven-
tions. The model in Figure 14-2 reflects the testing of
interventions on a continuum from preclinical work to
the final phase of study replication. This model is
particularly useful in guiding students through the
phases of intervention-based research. Initially, ideas
for interventions and the need for intervention research
340 UNIT TWO  The Research Process

Box 14-5 Sources of Information begin with problems and questions that arise in prac-
tice. You might be involved in any of these four phases
about Interventions
of intervention research as an investigator, consultant,
Peer-reviewed journal articles on nursing or research assistant (see Figure 14-2).
interventions
Nursing intervention taxonomies Designing the Intervention
Computerized databases containing data on An intervention to be tested must be carefully devel-
nursing interventions oped and well thought out to accomplish the purpose
Nursing textbooks of the study. The study framework should guide the
Previous intervention studies (theses, identification of the research hypotheses and the
dissertations, and other publications) development of the study design and intervention.
Clinical guidelines: http://www.guidelines.gov Investigators may need to consult other researchers
(see Chapter 19 for additional websites) who have designed and tested similar interventions,
Critical pathways the research project team who will carry out the
Intervention protocols research, and clinical experts who might use the inter-
Interviews with patients who have vention in practice. Aranda (2008) recommends taking
experienced the problem and related a similar approach to that used in pharmacological
interventions research, which is illustrated by a progressive path
Interviews with providers who have addressed from phase I to phase IV studies in determining safety,
the problem efficacy, and effectiveness.
Interviews with researchers who have tested During the design period, and guided by the inter-
previous interventions vention theory, the project team specifies the proce-
Probing of personal experiences dural elements of the intervention and develops an
Observations of care provided to patients observation system (Fawcett et al., 1994). The inter-
with the problem vention may be (1) a strategy; (2) a technique; (3) a
program; (4) informational or training materials;
(5) environmental design variables; (6) a motivational
system; or (7) a new or modified policy. The interven-
tion must be specified in sufficient detail to allow
interveners to implement it consistently.

Pain Promoting
assessment self-management

Practice Education
Advocacy
Pain Empowering
Qu

on

management patients
alit

ati
ov
yc

Inn
are

Evidence-based Overcoming
Research
practice barriers to care

Influencing
Intervention- Descriptive
research
based research research
agendas

Figure 14-2  Phases of intervention-based research from theory development to study replication.

Box 14-6 Criteria for Intervention
Design
1. The intervention is effective.
2. The intervention is replicable by typical
interveners.
3. The intervention is simple to use.
4. The intervention is practical.
5. The intervention is adaptable to various
contexts and settings.
6. The intervention is compatible with patient
preferences and values.
Adapted from Fawcett, S. B., Suarez-Belcazar, Y.,
Belcazar, F. E., White, G. W., Paine, A .L., Blanchard,
K. A., et al. (1994). Conducting intervention
research: The design and development process. In J.
Rothman & E. J. Thomas (Eds.), Intervention
research: Design and development for human
service (pp. 25–54). New York: Haworth Press.
During the design process, the intervention emerges
in stages as it is repeatedly tested, redesigned, and
retested. Training materials and programs for inter-
veners also must be developed and repeatedly tested
and revised. Design criteria are established to evaluate
the implementation of the intervention and outcomes
(Box 14-6). In addition to the detailed development of
the intervention, an operational development of the
design guided by the theory should include the follow-
ing actions:
• Define the target population.
• List acceptable strategies for selecting a sample.
• Identify subgroups that might show differential
effects of the intervention.
• Specify essential characteristics of interventionists.
• Determine study variables.
• Indicate appropriate measures of variables.
• Specify the appropriate time or times to measure
outcomes.
• Indicate what analyses to perform and what rela-
tionships to test on the basis of the relationships
among the treatment and the mediating and outcome
variables specified by the intervention theory
guiding the study.
Study Outcomes
Outcomes are determined by the problem and purpose
of the intervention and reflect the various effects of
the intervention. Selecting study endpoints to capture
the outcomes of an intervention must reflect changes
that occur as a consequence of the intervention.
CHAPTER 14  Intervention-Based Research 341
Typically, in an experimental or quasi-experimental
study in which intervention efficacy and effectiveness
are determined, there is a primary end point or outcome
of greatest interest to investigators and secondary out-
comes that add value to the impact of the intervention.
Regardless of whether they are primary or secondary,
outcome measures have to represent the concept or
concepts within the measurement domain(s) for the
study and must be sensitive and specific to the desired
effects brought about by the intervention.
Outcomes can be obtained using self-reported
questionnaires, survey instruments, or scales admin-
istered by the investigators or research personnel.
These outcome measures can be related to experi-
ences categorized as physical, psychological, social,
spiritual, or any combination of these types. Physio-
logical outcomes can be measured by specialized
equipment that records patients’ physical status, such
as heart rate, blood pressure, respiratory rate, and
oxygen saturation. Outcomes can also be measured by
criteria-based observational instruments or scales that
capture observed phenomena relevant to the interven-
tion. Box 14-7 lists the steps of the observation
process for intervention research. For example, a
study of an intervention to reduce hospital-acquired
pressure ulcers would require a grading scale to assess
the presence and severity of pressure ulcers that can
only be determined by a rater. Study outcomes can be
clinical indicators relating to events, sequences of
care, and resultant effects of care. Length of hospital-
ization, transfer to a critical care unit, rates of com-
plications, and disposition at discharge as defined by
the need for home health care, long-term care, or
rehabilitative therapy all represent clinical indicators
of care.
The timing and expected pattern of changes must
also be considered and specified for the frequency and
sequencing of collecting outcome measures. Timing is
the point in time after the intervention that a change
is expected to occur. Some changes occur immediately
after an intervention, whereas others may not appear
for some time. The sequencing of data collection on
outcomes should be done at the early and most optimal
points in which that intervention may show an effect.
However, the selected outcomes measures must be
able to demonstrate these changes at the particular
time points. For example, studies involving interven-
tions focused on quality of life need to allow ample
time during and following the intervention to detect
improvements in quality of life. Changes in symp-
toms, on the other hand, may occur more rapidly with
an intervention, so more frequent surveillance of these
outcomes is often warranted. Interventions that have
342 UNIT TWO  The Research Process
Box 14-7 Steps of the Observation
Process
1. Determine elements that must be observed
on the basis of the intervention theory.
2. Develop methods of measuring essential
elements.
3. Develop criteria for determining whether
or not the event to be observed has
occurred.
4. Select observers.
5. Train observers.
6. Develop scoring instructions to guide
recording of desired behaviors or
outcomes.
7. Develop a schedule of observations to
include the following:
a. What is happening before the
intervention is implemented?
b. What is happening during the
intervention?
c. What changes occur after the
intervention?
8. Perform preliminary analysis of pre-
intervention data.
9. Apply preliminary analysis results to
further develop the intervention.
10. Analyze changes in environment and
behaviors before, during, and after the
intervention.
11. Refine intervention theory.
no anticipated or appreciable carry-over effect follow-
ing implementation do require study outcomes col-
lected at the time or shortly after the intervention.
Typically, the literature and psychometric properties
of instruments and scales can be used to guide sched-
ules for collection of outcomes.
If theory-guided intervention research is used, there
may be mediating variables, referred to in some
studies as intermediate outcomes, that have rele-
vance to understanding the intervention. Hypothesized
interrelationships among the outcomes must be indi-
cated in the theory and must be represented in the
study through either controlling for these variables in
the design and sampling procedures, or measuring
them as part of the study in order to manage them in
the analysis and interpretations of results.
Extraneous Factors
Extraneous factors or variables are elements of the
environment or characteristics of the patient that
significantly affect the problem, the treatment process,
or the outcomes. Unlike mediator variables, extrane-
ous factors tend not to be well understood and may
not be identified until a study has been initiated. They
are seldom included in explanations of the causal links
between intervention and outcomes. Thus, they are
extraneous to existing theoretical explanations of
cause. They are sometimes referred to as confounding
variables. If the researcher recognizes their potential
effect, extraneous factors may be held constant or
measured so that they can be statistically controlled
during analysis. Careful analyses may indicate that
some variables defined as extraneous are actually
moderator or mediator variables. (For additional infor-
mation on types of variables refer to Chapter 8.)
Process of Testing
the Intervention
Intervention Validation
For some interventions, before the pilot test, it is
useful to test prototypes in situations and environ-
ments that are similar to those planned for the actual
study. Van Meijel, Gamel, van Swieten-Duijfjes, and
Grypdonck (2004) contend that too little time is spent
on the development of an intervention, since more
time is spent in the actual testing of the intervention.
Before one proceeds with a formal study, an interven-
tion should be examined in the setting in which the
problem occurs or in a situation resembling the natural
setting. This examination helps investigators and
research staff, as well as nurses caring for study par-
ticipants, to experience how an intervention may play
out under actual study conditions. Logistical aspects
of the intervention may not be fully developed or
foreseeable until a “real world” trial can be realized,
and the problems addressed. This method of prelimi-
nary validation of the intervention is quite helpful for
studies involving sensitive issues, such as sexual
abuse, alcohol abuse, and drug use. Intervention vali-
dation is important when there is uncertainty about
how research staff will perform and/or participants
will react. It helps researchers determine the feasibility
of the intervention and its components as well as how
accepting nurses and potential participants might be
to the intervention.
Another strategy is to conduct intervention
validation under simulated conditions. Conducting
simulation exercises in an artificial environment close
to the natural study environment might identify issues
in delivering the intervention that can be corrected

before the study is conducted. Role-playing tech-
niques are often used to deliver the intervention.
Members of the project team, staff from the settings
to be used for the project, or nursing students might
perform the roles. Simulation scenarios are con-
structed following the exact steps prescribed by the
intervention protocol. Video recording may be used to
allow investigators to carefully examine any issues or
procedures that require modifications prior to imple-
menting the pilot or final study. Observers also make
notes during the testing of the intervention to identify
missing elements, gain insights, or develop questions
the project team must explore. Some degree of infer-
ence is gleaned from validating an intervention in an
artificial environment and being proactive in refining
study procedures and measurement outcomes on the
basis of simulated experiences.
Pilot Testing of an Intervention
For multiple reasons, pilot or field testing of an inter-
vention is recommended. In some cases, pilot testing
is required when a researcher is applying for grant
funding, seeking approval from an IRB, or garnering
support from an institution to conduct the study.
Feeley et al. (2009) underscore the importance of pilot
testing in developing RCTs for the following reasons:
(1) to assess the feasibility and acceptability of the
intervention; (2) to determine the feasibility and
acceptability of the design and procedures; and (3) to
facilitate the determination of treatment ES to use in
power analysis calculations of sample size. In general,
pilot studies are very useful to:
1. Determine whether the intervention as a prototype
will work and whether it is feasible.
2. Guide the refinement of the intervention and to
work out any problems or issues with implement-
ing it in clinical or research settings. Pilot studies
are especially valuable when interventions are
delivered in the context of usual or routine clinical
care. Investigators can determine any barriers to
as well as facilitators for the execution of study
procedures, and seek support and resources to
administer the intervention in the manner intended.
3. Test and refine instructions, manuals, or training
programs to ensure that study procedures are
understandable and the preparation of study staff
is adequate.
4. Estimate how long it takes to recruit the study
population, implement the intervention, and
collect the desired data.
5. Assess whether the intervention has been
described in sufficient detail to allow clinicians
and other researchers to replicate the work. During
CHAPTER 14  Intervention-Based Research 343
pilot tests, clinicians and research staff should
continually question any concerns or issues that
arise.
6. Verify whether there are problems in gaining
access to study participants and how participants
are reacting to study conditions and procedures.
7. Ensure that study participants understand what is
required of the study and can interpret and com-
plete outcome measures.
8. Test the feasibility of the design.
9. Obtain reliability and validity estimates for mea-
surement methods in the target population.
10. Ascertain whether there are any unanticipated
effects that may be of concern for participants or
detrimental to the internal and external validity of
the study.
Importantly, pilot testing allows the generation of
data or point estimates (such as means and standard
deviations) for the study outcomes. These estimates
can be used to calculate sufficient statistical power in
terms of study sample size needed to detect a statisti-
cally significant treatment effect. Pilot data are often
a prerequisite for determining adequate sample sizes
when there are no available studies to calculate the ES
for the intervention. Typically, only 10 to 20 partici-
pants are needed to obtain estimates of variance in the
primary outcome measure(s) needed to determine
intervention ES (Hertzog, 2008; see Chapter 15 for
calculation of ESs for interventions).
Pilot studies should be conducted with the target
population and in settings similar to those planned for
the proposed study. These studies must also be
approved by the organization’s IRB in compliance
with all regulations and mandates for acceptable
conduct of research and protection of human subjects.
The manner in which all study procedures are imple-
mented must also be exactly the same as described in
the study proposal. Strict adherence to study proce-
dures enables investigators to use pilot data as the
starting point for the study if no amendments or
changes are made. The pilot testing phase can move
right into the larger-scale study if this is a condition
stipulated in the approved IRB application.
Often pilot studies are separate studies, and inves-
tigators may alter protocols on the basis of the experi-
ence of the pilot study. Keeping field notes is
particularly useful during pilot studies so relevant
issues can be addressed prior to implementing the
larger-scale study. Pilot data may be published,
although with controlled trials, findings must be mean-
ingful to the body of research (Watson, Atkinson, &
Rose, 2007). Thus, pilot studies need to be published
if they provide estimates of variance in the outcomes,
344 UNIT TWO  The Research Process
help determine the quality of the intervention protocol,
and/or identify issues relevant to future studies.
However, it is imperative that researchers avoid any
inference to hypothesis testing using pilot studies.
Some researchers have cautioned that problems remain
in the interpretation of pilot studies’ findings in pub-
lications (Arain, Campbell, Cooper, & Lancaster,
2010).With the limited sample sizes, researchers run
the risk of misinterpreting any significant or nonsig-
nificant results. Type II errors are common when the
sample size is insufficient to provide acceptable sta-
tistical power to detect differences from interventions.
Pilot samples are nonrepresentative of the target popu-
lation, and data tend to be nonnormally distributed.
Assumptions of normally distributed data are often
violated if parametric statistical tests were used in the
analyses of pilot data. Nonparametric statistics are
more appropriate for conducting group comparisons
with pilot data and should be considered. Sometimes
only descriptive statistics (frequencies, means, stan-
dard deviations, ranges, scatter plots or diagrams, and
medians) are most appropriate in reporting the results
from pilot studies.
Graduate students should consider opportunities to
work on pilot studies because they are generally of
short duration and instructive about the research
process. Investigators often try to recruit graduate stu-
dents for their studies because they have knowledge
of research and can provide useful feedback about the
experience as well as flexible daytime schedules to
accommodate the work.
Formal Testing of an Intervention
The most desirable formal test of an intervention is a
conventional experimental design to determine
whether the intervention causes the intended effects.
The design should be as rigorous as possible. Research-
ers need to conduct a power analysis to determine a
sample size sufficient to avoid a Type II error (see
Chapter 15 for discussion of power analysis). Analy-
ses must be performed to ensure that the treatment and
control groups are comparable on important demo-
graphic variables. The measurement methods used in
studies must have documented reliability and validity
(see Chapter 16). The ES for each outcome needs to
be examined and reported. The observation system
established before the initiation of testing must be
continued, and patient characteristics, intervener
characteristics, and setting characteristics need to be
measured and described (Sidani & Braden, 1998).
Researchers also must develop plans for statistical
analyses to detect differences between the intervention
and comparison groups (see Chapter 25).
Advanced Testing
Advanced testing of the intervention occurs after suf-
ficient evidence is available that the intervention is
effective in achieving the desired outcomes. The
model proposed by Forbes (2009) places this type of
testing in phases III and IV, during which the interven-
tion is compared with standard of care interventions,
followed by replication studies with long-term
follow-up of the intervention’s effectiveness (see
Chapter 11 for discussion of phases I to IV of clinical
trials). Phase III of intervention testing might involve
a single, well-designed study that indicates a satisfac-
tory ES for an intervention, but often an intervention
requires modification through additional studies. Rep-
lication studies ensure that an intervention is refined
to create the desired effect in practice. Sidani and
Braden (1998) emphasize that the advanced testing
phase must include variations in the studies conducted
to ensure the intervention is ready for delivery in prac-
tice to a variety of patients.
Testing Variations in Effectiveness Based on
Variations in Patient Characteristics
Intervention effects that have been determined through
the use of a sample of only white, middle-class Ameri-
cans may not have the same effect with other groups.
The intervention should be tested among various
ethnic groups and vulnerable populations, such as the
socioeconomically and educationally disadvantaged.
Pilot tests may reveal a need to refine the intervention
to make it more culturally appropriate and adaptive to
diverse groups. Minority and disadvantaged popula-
tions may respond differently to interventions because
they: (1) have different views of health and preventive
behaviors; (2) do not fully understand the purpose of
the study and their role in participation; (3) are more
vulnerable to persuasion from researchers and health-
care professionals; and (4) encounter difficulties in
comprehending patient-reported outcome measures
used in the study. For these reasons, researchers need
to design studies and select outcomes that are tailored
to diverse patients or screen potential participants for
variables that might interfere with study participation.
However, eliminating diversity as a requisite for a
study can limit the generalizability of findings and
hamper efforts to better understand and reduce health-
care disparities.
Studies also must be conducted to examine
the effect of an intervention on patients with comor-
bidities or differing levels of illness severity. Other
variations in patient characteristics, such as age,
gender, and diagnosis, must be examined. Character-
istics specific to the intervention may be identified as

important for determining differential effects. When
sufficiently large samples are obtained in initial
studies, these patient characteristics can be examined
in secondary analyses of available data.
Testing Variations in Effectiveness Based on Setting
If the research setting is similar for the testing of an
intervention with similar circumstances, conditions,
and subjects, then the effects of a different setting can
potentially confound the treatment effects. Therefore,
one component of testing the intervention is to set up
multisite projects, in which the settings are varied and
the effects of the settings on outcomes are examined
(Sidani & Braden, 1998).
Testing Variations in Effectiveness Based on
Variations in Intervener Characteristics
The initial study examining the effectiveness of an
intervention is usually conducted under ideal condi-
tions. Ideal conditions involve the selection of highly
educated interveners judged to be experts in the field
of practice related to the intervention. However,
after the intervention is found to be effective, ques-
tions arise regarding the use of less well-prepared
Box 14-8 Factors Affecting the Integrity of an Intervention and Study Validity
Lack of Stringent Study Controls
Poorly defined study intervention
Lack of attention to detail in planning the
intervention(s)
Lack of or inadequate use of written guidelines
and protocols for the intervention(s)
Unreported protocol violations or deviations
from standardized procedures
Inadequate supervision and guidance during
study from the principal investigator and/or
research coordinator
Variations in dose intensity of study
intervention(s) across study participants
Lack of reliable and valid outcome measures
Complexity of Intervention Procedures
Level of complexity of the intervention(s)
Variation in elements of intervention provided
Difficulties in implementing intervention
activities
Number of sites involved in implementing the
intervention
CHAPTER 14  Intervention-Based Research 345
interveners to provide the intervention. Studies need
to be conducted to determine variations in the effec-
tiveness of the intervention based on the competencies
of interveners.
Testing Variations in Effectiveness Based on
Strength of an Intervention
The strategy of testing the variations of an interven-
tion’s strength is used to determine the amount of
treatment needed to achieve the desired outcome in
practice. To test this issue, the researcher must provide
varying doses of the intervention. Researchers might
vary the intensity of the intervention, the length of
time of a single treatment, the frequency of an inter-
vention, and/or the span of time over which the inter-
vention is continued or repeated.
Data Collection
Data collection in intervention research consists of not
only accurate collection and recording of study out-
comes but also accurate documentation of the study
intervention. Researchers need to document the timing
of the intervention, dose-intensity of the intervention,
Study Intervener and Staff Variables
Too many individuals delivering the
intervention(s)
Level of education and research experience of
study staff
Failure to obtain intraobserver and
interobserver reliability
Lack of initial and ongoing preparation and
competency verification assessments
Level of compliance of staff with treatment
protocol
Competing priorities and responsibilities of
study staff
Turnover of study staff during the investigation
Level of staff interest and commitment to the
intervention
Study Site Variables
Lack of human, professional, and financial
resources to support the study
Changes in organizational policies, procedures,
and practices after initiation of study
Adoption of new technology or equipment as
part of the intervention
346 UNIT TWO  The Research Process
Box 14-9 Step by Step Approach to Critically Apprising Intervention-Based Research:
A Focus on the Intervention
Step 1: Background and Review of Literature
Examine the scientific basis for the study
intervention:
• Is the scientific rationale for the intervention
supported by the study literature review?
• Do authors cite variations in ways the
intervention has been studied?
• Has the intervention been tested in different
study populations?
• Is the magnitude or effect size of the
intervention provided from previous studies?
Step 2: Theoretical Framework
Establish whether the intervention is theory-
based or conceptually derived and the level to
which theory drives the intervention:
• Does the intervention evolve from a theory or
conceptual framework?
• How has the intervention been developed
using a theory?
• Are the relationships of the study framework
to the intervention explicit?
• Is the intervention represented in a model of
the study framework?
Step 3: Design
Determine the appropriateness of the design to
test the intervention:
• Is the design appropriate to test study
hypotheses?
• Are confounding variables controlled by the
design?
• Is an experimental design with a control,
placebo, or standard care group used to test
the efficacy or effectiveness of an
intervention?
• In a study with a quasi-experimental design, is
a comparison group used in testing the
intervention?
Step 4: Sample
Examine the appropriateness of the sample and
sampling process:
• Is the sample representative of the population
that might benefit most from the
intervention?
• How are subjects allocated to treatment and,
if appropriate, control or comparison groups?
• Was a power analysis performed to determine
an adequate sample size to yield sufficient
statistical power to detect statistically
significant differences between or among
study treatment conditions?
• Do inclusion and exclusion sample criteria
control for sample characteristics that might
influence the outcomes?
• To what extent does the sample allow for
generalizations to similar populations, thereby
improving the study’s external validity?
Step 5: Study Outcomes
Evaluate the study outcomes:
• Are reliability and validity (psychometric
properties) of all instruments and outcome
measures reported, and are they acceptable?
• Do investigators plan to conduct psychometric
testing of instruments or outcomes measures
if existing ones are applied to different
populations or if new ones are designed by
the investigators for the purpose of the
study?
• Has interrater or intrarater reliability been
established for data collectors if observational
methods are used to obtain outcome
measurements?
• What limitations are evident for the outcome
measures, and are these addressed?
Step 6: Study Procedures
Evaluate the integrity for the implementation of
the intervention:
• Are the study procedures clearly explained to
the degree that other researchers might be
able to replicate the study?
• How were study personnel or clinical staff
prepared to execute study procedures?
• Were fidelity or manipulation checks
conducted during training or in the course
of the study?
• Was information collected on protocol
deviations and dose intensity of the
intervention?
Step 7: Statistical Analyses
Assess the appropriateness of statistical
methods:
• Are appropriate statistical methods applied
given the level of measurement for outcome
variables?

Box 14-9 Step by Step Approach to Critically Apprising Intervention-Based Research:
A Focus on the Intervention—cont’d
• Have the investigators violated any statistical
assumptions in the selection of analytical
procedures raising statistical conclusion
validity issues?
subject responses to the intervention, missed treat-
ments, and any protocol deviations that occur even if
these are beyond the intervener’s control. Investiga-
tors also must keep track of any unexpected circum-
stances that occur in the implementation of the study
intervention or at the time the outcomes are measured.
These types of field notes can be vital to the interpreta-
tion of study participants’ responses during data analy-
sis and the development of findings (see Chapter 20
for more details on the data collection process).
Threats to Study Validity
Like other types of research, intervention research is
subject to threats to validity. Box 14-8 summarizes
aspects of the research design, intervention, and analy-
ses that need to be examined for threats to validity.
Determining threats to internal validity in the execu-
tion of a study is critically important to examining
study results and interpreting findings. Studies must
be evaluated for sources of bias from systematic error
that might occur with problems of treatment fidelity
and uncontrolled confounding variables. Researchers
also need to evaluate the bias and possible threats to
design validity occurring from random error intro-
duced by low statistical power, imprecise instruments,
and inappropriate statistical methods (Gerhard, 2008;
Shadish, Cook, & Campbell, 2002; see Chapter 10).
Critical Appraisal of
Intervention-Based Research
Criteria for critically appraising intervention-based
research are very similar to those for the appraisal of
any type of study. However, a more thorough and
systematic evaluation of the integrity of the interven-
tion, study procedures for implementing the interven-
tion, and outcomes for measuring its efficacy or
effectiveness is required. The validity of research find-
ings involving interventions depends on sound meth-
odological practices including accuracy, precision,
and consistency in the delivery of a study intervention.
CHAPTER 14  Intervention-Based Research 347
• Did the intervention show a statistically
significant difference in the study outcomes?
• Were the study results clinically important?
Equally important are controls to reduce bias and error
that pose threats to study validity. Many of these issues
have been discussed previously. Guidelines for evalu-
ating components of the intervention are outlined in
Box 14-9. Important questions are posed to help
nurses identify aspects of studies specific to the inter-
vention and to assist them in critically appraising the
intervention and its implementation in a study.
KEY POINTS
• Intervention research holds great promise for
designing and testing nursing interventions and
advancing nursing science.
• Nursing interventions are defined as “deliberative
cognitive, physical, or verbal activities performed
with, or on behalf of, individuals and their families
that are directed toward accomplishing particular
therapeutic objectives relative to individuals’ health
and well-being” (Grobe, 1996, p. 50).
• Nursing interventions are nurse-initiated and can
be based on nursing classifications of interventions
unique to clinical problems addressed by nurses.
• Intervention research programs consist of multiple
studies conducted over a period of years by a
project team that may include undergraduate and
graduate nursing students.
• Some teams use a participatory research method
that involves community groups.
• An intervention theory must include: (1) a careful
description of the problem to be addressed; (2) the
intervening actions that must be implemented to
address the problem; (3) moderator variables that
might change the impact of the intervention; (4)
mediator variables that might alter the effect of the
intervention; and (5) expected outcomes of the
intervention.
• The intervention theory guides the design and
development of an intervention, which is then
extensively tested, refined, and retested using quan-
titative, qualitative, and mixed-methods research
designs.
348 UNIT TWO  The Research Process
• Experimental and quasi-experimental research
designs are particularly useful in determining inter-
vention effectiveness through the use of such
methods as parallel groups, crossover, cluster, pref-
erence, client treatment matching, and open-label
designs.
• Advanced testing of the intervention occurs after
sufficient evidence is available to determine that
the intervention is effective in achieving desired
outcomes.
• When the intervention is sufficiently refined and
evidence of effectiveness has been determined, the
intervention is field-tested to ensure that it can be
effectively implemented in clinical settings.
• An observation system is developed for use
throughout the intervention design and develop-
ment process. This system allows the researchers
to observe events related to the intervention natu-
ralistically and to analyze these observations.
• Dissemination efforts are more extensive than in
traditional experimental studies and involve choos-
ing a brand name, establishing a price, and setting
standards for the intervention’s use.
• As with any study, intervention-based research
needs to be critically appraised to ascertain the
integrity of the intervention, its implementation
fidelity, and reported outcomes.
REFERENCES
Aberson, C. L. (2010). Applied power analysis for the behavioral
sciences. New York, NY: Routledge Taylor & Francis Group.
Adamson, J., Cockayne, S., Puffer, S., & Torgerson, D. J. (2006).
Review of randomised trials using the post-randomised consent
(Zelen’s) design. Contemporary Clinical Trials, 27(4), 305–319.
Arain, M., Campbell, M. J., Cooper, C. L., & Lancaster, G. A.
(2010). What is a pilot or feasibility study? A review of current
practice and editorial policy. BMC Medical Research Methodol-
ogy, 10, 67.
Aranda, S. (2008). Designing nursing interventions. Collegian,
15(1), 19–25.
Baird, C. L., Murawski, M. M., & Wu, J. (2010). Efficacy of guided
imagery with relaxation for osteoarthritis symptoms and medica-
tion intake. Pain Management Nursing, 11(1), 56–65.
Bellg, A. J., Borrelli, B., Resnick, B., Hecht, J., Minicucci, D. S.,
Ory, M., et al.; Treatment Fidelity Workgroup of the NIH Behav-
ior Change Consortium. (2004). Enhancing treatment fidelity in
health behavior change studies: Best practices and recommenda-
tions from the NIH Behavior Change Consortium. Health Psy-
chology, 23(5), 443–451.
Benedetti, F., Carlino, E., & Pollo, A. (2011). How placebos change
the patient’s brain. Neuropsychopharmacology, 36(1), 339–354.
Borrelli, B., Sepinwall, D., Ernst, D., Bellg, A. J., Czajkowski, S.,
Breger, R., et al. (2005). A new tool to assess treatment fidelity
and evaluation of treatment fidelity across 10 years of health
behavior research. Journal of Consulting and Clinical Psychol-
ogy, 73(5), 852–860.
Brady, M. C., Stott, D. J., Norrie, J., Chalmers, C., St. George, B.,
Sweeney, P. M., et al. (2011). Developing and evaluating the
implementation of a complex intervention: Using mixed methods
to inform the design of a randomised controlled trial of an oral
healthcare intervention after stroke. Trials, 12, 168.
Breit-Smith, A., Justice, L. M., McGinty, A., & Kaderavek, J.
(2009). How often and how much? Intensity of print referencing
intervention. Topics in Language Disorders, 29(4), 360–369.
Bulechek, G., Butcher, H., & Dochterman, J. (2008). Nursing inter-
ventions classification (NIC) (5th ed.). St. Louis, MO: Elsevier.
Cahill, K., Lancaster, T., & Green, N. (2010). Stage-based interven-
tions for smoking cessation. Cochrane Database Systematic
Review, 11, CD004492.
Cohen, J. (1988). Statistical power analysis for the behavioral sci-
ences (2nd ed.). Hillsdale, NJ: Lawrence Earlbaum Associates.
Compher, C. (2010). Efficacy vs. effectiveness. Journal of Paren-
teral and Enteral Nutrition, 34(6), 598–599.
Conn, V. S., Hafdahl, A. R., & Mehr, D. R. (2011). Interventions to
increase physical activity among healthy adults: Meta-analysis of
outcomes. American Journal of Public Health, 101(4), 751–758.
Consolidated Standards of Reporting Trials (CONSORT, 2011).
Transparency reporting of trials. Retrieved from http://www
.consort-statement.org/consort-statement/3-12-methods/item11a_
blinding/.
Creswell, J. W. & Zhang, W. (2009). The application of mixed
methods designs to trauma research. Journal of Traumatic Stress,
22(6), 612–621.
Day, S. (2007). Dictionary for clinical trials (2nd ed.). Hoboken,
NJ: John Wiley & Sons.
Deacon, M. & Fairhurst, E. (2008). The real-life practice of acute
inpatient mental health nurses: An analysis of “eight interrelated
bundles of activity.” Nursing Inquiry, 15(4), 330–340.
Donovan, D. M. & Mattson, M. E. (1994). Alcoholism treatment
matching research: Methodological and clinical issues. Journal of
Studies of Alcohol, 12(Suppl), 5–14.
Donovan, D. M., Kadden, R. M., DiClemente, C. C., Carroll,
K. M., Longabaugh, R., Zweben, A., et al. (1994). Issues in the
selection and development of therapies in alcoholism treatment
matching research. Journal of Studies of Alcohol, 12(Suppl),
138–148.
Duffy, M. (2006). The randomized controlled trial: Basic consider-
ations. Clinical Nurse Specialist, 20(2), 62–64.
Dworkin, R. H., Turk, D. C., McDermott, M. P., Peirce-Sandner, S.,
Burke, L. B., Cowan, P., et al. (2009). Interpreting the clinical
importance of group differences in chronic pain clinical trials:
IMMPACT recommendations. Pain, 146(3), 238–244.
Faul, L. A., Jim, H. S., Minton, S., Fishman, M., Tanvetyanon, T.,
& Jacobsen, P. B. (2011). Relationship of exercise to quality of
life in cancer patients beginning chemotherapy. Journal of Pain
& Symptom Management, 41(5), 859–869.
Fawcett, S. B., Suarez-Belcazar, Y., Balcazar, F. E., White, G. W.,
Paine, A. L., Blanchard, K. A., et al. (1994). Conducting interven-
tion research: The design and development process. In: J. Rothman
& E. J. Thomas (Eds.), Intervention research: Design and devel-
opment for human service (pp. 25–54). New York, NY: Haworth
Press.

Feeley, N., Cossette, S., Côté, J., Héon, M., Stremler, R., Martorella,
G., et al. (2009). The importance of piloting an RCT intervention.
Canadian Journal of Nursing Research, 41(2), 85–99.
Forbes, A. (2009). Clinical intervention research in nursing. Inter-
national Journal of Nursing Studies, 46(4), 557–568.
Gartlehner, G., Hansen, R. A., Nissman, D., Lohr, K. N., & Carey,
T. S. (2006). Criteria for distinguishing effectiveness from efficacy
trials in systematic reviews. Technical Review, Agency for Health-
care Research and Quality. (AHRQ Publication No. 06-0046.)
Rockville, MD: U.S. Department of Health and Human
Services.
Gearing, R. E., El-Bassel, N., Ghesquiere, A., Baldwin, S., Gillies,
J., & Ngeow, E. (2011). Major ingredients of fidelity: A review
and scientific guide to improving quality of intervention research
implementation. Clinical Psychological Reviews, 31(1), 79–88.
Gerhard, T. (2008). Bias: Considerations for research practice.
American Journal of Health System Pharmacists, 65(22),
2159–2168.
Gholizadeh, L., Davidson, P., Salamonson, Y., & Worrall-Carter, L.
(2010). Theoretical considerations in reducing risk for cardiovas-
cular disease: Implications for nursing practice. Journal of Clini-
cal Nursing, 19(15-16), 2137–2145.
Goode, C. J., Titler, M., Rakel, B., Ones, D. S., Kleiber, C., Small,
S., et al. (1991). A meta-analysis of effects of heparin flush and
saline flush: Quality and cost implications. Nursing Research,
40(6), 324–330.
Gorecki, C., Lamping, D. L., Brown, J. M., Madill, A., Firth, J., &
Nixon, J. (2010). Development of a conceptual framework of
health-related quality of life in pressure ulcers: A patient-focused
approach. International Journal of Nursing Studies, 47(12),
1525–1534.
Grobe, S. J. (1996). The nursing intervention lexicon and taxonomy:
Implications for representing nursing care data in automated
patient records. Holistic Nursing Practice, 11(1), 48–63.
Hatfield, L. A., Gusic, M. E., Dyer, A. M., & Polomano, R. C.
(2008). Analgesic properties of oral sucrose during routine
immunizations at 2 and 4 months of age. Pediatrics, 121(2),
e327–e334.
Hertzog, M. A. (2008). Considerations in determining sample
size for pilot studies. Research in Nursing & Health, 31(2),
180–191.
Kazer, M. W., Bailey, D. E. Jr., & Whittemore, R. (2010). Out of
the black box: Expansion of a theory-based intervention to self-
manage the uncertainty associated with active surveillance (AS)
for prostate cancer. Research & Theory in Nursing Practice,
24(2), 101–112.
Kolanowski, A., Buettner, L., & Moeller, J. (2006). Treatment fidel-
ity plan for an activity intervention designed for persons with
dementia. American Journal of Alzheimer’s Disease and Other
Dementias, 21(5), 326–332.
Kolanowski, A., Litaker, M., Buettner, L., Moeller, J., & Costa,
P. T. Jr. (2011). A randomized clinical trial of theory-based activi-
ties for the behavioral symptoms of dementia in nursing home
residents. Journal of the American Geriatric Society, 59(6),
1032–1041.
Krampe, J., Rantz, M. J., Dowell, L., Schamp, R., Skubic, M.,
& Abbott, C. (2010). Dance-based therapy in a program of
all-inclusive care for the elderly: An integrative approach to
CHAPTER 14  Intervention-Based Research 349
decrease fall risk. Nursing Administrative Quarterly, 34(2),
156–161.
Kwekkeboom, K. L. (1997). The placebo effect in symptom man-
agement. Oncology Nursing Forum, 24(8), 1393–1399.
Li, H., Melnyk, B. M., & McCann, R. (2004). Review of interven-
tion studies of families with hospitalized elderly relatives. Journal
of Nursing Scholarship, 36(1), 54–59.
Martin, K. S. (2005). The Omaha System: A key to practice, docu-
mentation, and information management (2nd ed.). Omaha, NE:
Health Connections Press.
Martin, K. S. & Bowles, K. H. (2008). Using a standardized lan-
guage to increase collaboration between research and practice.
Nursing Outlook, 56(3), 138–139.
Matthews, J. N. S. (2000). An introduction to randomized controlled
clinical trials. London, England: Arnold, Hodder Headline Group.
McCloskey, J. C. & Bulechek, G. M. (2000). Nursing interventions
Classification (NIC) (3rd ed.). St. Louis: Mosby-Year.
Melnick, G., De Leon, G., Thomas, G., & Kressel, D. (2001). A
client-treatment matching protocol for therapeutic communities:
First report. Journal of Substance Abuse and Treatment, 21(3),
119–128.
National Research Council. (2010). The prevention and treatment
of missing data in clinical trials. Panel on Handling Missing Data
in Clinical Trials. Committee on National Statistics, Division of
Behavioral and Social Sciences and Education. Washington, DC:
The National Academies Press.
Pellfolk, T. J., Gustafson, Y., Bucht, G., & Karlsson, S. (2010).
Effects of a restraint minimization program on staff knowledge,
attitudes, and practice: A cluster randomized trial. Journal of the
American Geriatric Society, 58(1), 62–69.
Piantadosi, S. (2005). Clinical trials: A methodologic perspective
(2nd ed.). Hoboken, NJ: John Wiley & Sons.
Polit, D. F., Gillespie, B. M., & Griffin, R. (2011). Deliberate igno-
rance: A systematic review of blinding in nursing clinical trials.
Nursing Research, 60(1), 9–16.
Preference Collaborative Review Group. (2009). Patients’ prefer-
ences within randomised trials: Systematic review and patient
level meta-analysis. BMJ, 337, a1864.
Quigley, P. A., Hahm, B., Collazo, S., Gibson, W., Janzen, S.,
Powell-Cope, G., et al. (2009). Reducing serious injury from falls
in two veterans’ hospital medical-surgical units. Journal of
Nursing Care Quality, 24(1), 33–41.
Radziewicz, R. M., Rose, J. H., Bowman, K. F., Berila, R. A.,
O’Toole, E. E., & Given, B. (2009). Establishing treatment fidelity
in a coping and communication support telephone intervention for
aging patients with advanced cancer and their family caregivers.
Cancer Nursing, 32(3), 193–202.
Reed, D., Titler, M. G., Dochterman, J. M., Shever, L. L., Kanak,
M., & Picone, D. M. (2007). Measuring the dose of nursing inter-
vention. International Journal of Nursing Terminology Classifica-
tion, 18(4), 121–130.
Rothman, J., & Thomas, E. J. (1994). Intervention research: Design
and development for human service. New York, NY: Haworth
Press.
Saba, V. (2002). Nursing classifications: Home Health Care Clas-
sification System (HHCC): An overview. Online Journal of Issues
in Nursing, 7(3), 9. Retrieved from http://www.nursingworld.org/
MainMenuCategories/ANAMarketplace/ANAPeriodicals/OJIN/
350 UNIT TWO  The Research Process
TableofContents/Volume72002/No3Sept2002/ArticlesPrevious
Topic/HHCCAnOverview.asp/.
Santacroce, S. J., Maccarelli, L. M., & Grey, M. (2004). Interven-
tion fidelity. Nursing Research, 53(1), 63–66.
Schmelzer, M. (1999-2001). Safety and effectiveness of large
volume enema solutions. (Unpublished paper.) National Institutes
of Health Research Enhancement Award (AREA) NIH grant
number R15NR04867-01.
Schmelzer, M. & Wright, K. (1993). Risky enemas: What’s the ideal
solution? American Journal of Nursing, 93(7), 16.
Schmelzer, M. & Wright, K. (1996). Enema administration tech-
niques used by experienced registered nurses. Gastroenterology
Nursing, 19(5), 171–175.
Schmelzer, M., Case, P., Chappell, S., & Wright, K. (2000). Colonic
cleansing, fluid absorption, and discomfort following tap
water and soapsuds enemas. Applied Nursing Research, 13(2),
83–91.
Schmelzer, M., Schiller, L. R., Meyer, R., Rugari, S. M., & Case,
P. (2004). Safety and effectiveness of large-volume enema solu-
tions. Applied Nursing Research, 17(4), 265–274.
Schoneman, D. (2002). The intervention of surveillance across clas-
sification systems. International Journal of Nursing Terminology
in Classification, 13(4), 137–147.
Scobbie, L., Wyke, S., & Dixon, D. (2009). Identifying and apply-
ing psychological theory to setting and achieving rehabilitation
goals. Clinical Rehabilitation, 23(4), 321–333.
Shadish, W. R., Cook, T. D., & Campbell, D. T. (2002). Experimen-
tal and quasi-experimental designs for generalized causal infer-
ence. Boston, MA: Houghton Mifflin.
Shever, L. L. (2011). The impact of nursing surveillance on failure to
rescue. Research & Theory in Nursing Practice, 25(2), 107–126.
Sidani, S. & Braden, C. J. (1998). Evaluating nursing interventions:
A theory-driven approach. Thousand Oaks, CA: Sage.
Vallerand, A. H., Musto, S., & Polomano, R. C. (2011). Nursing’s
role in cancer pain management. Current Pain and Headache
Reports, 15(4), 250–562.
Van Hecke, A., Verhaeghe, S., Grypdonck, M., Beele, H., Flour, M.,
& Defloor, T. (2011). Systematic development and validation of
a nursing intervention: The case of lifestyle adherence promotion
in patients with leg ulcers. Journal of Advanced Nursing, 67(3),
662–676.
van Meijel, B., Gamel, C., van Swieten-Duijfjes, B., & Grypdonck,
M. H. (2004). The development of evidence-based nursing inter-
ventions: Methodological considerations. Journal of Advanced
Nursing, 48(1), 84–92.
Vedelø, T. W. & Lomborg, K. (2011). Reported challenges in nurse-
led randomised controlled trials: An integrative review of the litera-
ture. Scandinavian Journal of Caring Science, 25(1), 194–200.
Watson, R., Atkinson, I., & Rose, K. (2007). Pilot studies: To
publish or not? Journal of Clinical Nursing, 16(4), 619–620.
Weaver, T. E., Chasens, E. R., & Arora, S. (2009). Modafinil
improves functional outcomes in patients with residual excessive
sleepiness associated with CPAP treatment. Journal of Clinical
Sleep Medicine, 5(6), 499–505.
Wood, A. M., White, I. R., & Thompson, S. G. (2004). Are missing
outcome data adequately handled? A review of published random-
ized controlled trials in major medical journals. Clinical Trials,
1(4), 368–376.
   http://evolve.elsevier.com/Grove/practice/
M
any of us have preconceived notions about
samples and sampling, which we acquired
from television commercials, polls of public
opinion, market researchers, and newspaper reports of
research findings. The advertiser boasts that four of
five doctors recommend its product; the newscaster
announces that John Jones is predicted to win the
senate election by a margin of 3 to 1; the newspaper
reports that scientists’ studies have found that taking
a statin drug, such as atorvastatin (Lipitor), signifi-
cantly reduces the risk of coronary artery disease.
All of these examples use sampling techniques.
However, some of the outcomes are more valid than
others, partly because of the sampling techniques
used. In most instances, television, newspapers, and
advertisements do not explain their sampling tech-
niques. You may hold opinions about the adequacy of
these techniques, but there is not enough information
to make a judgment.
The sampling component is an important part of the
research process that needs to be carefully thought out
and clearly described. To achieve these goals, research-
ers need to understand the techniques of sampling and
the reasoning behind them. With this knowledge, you
can make intelligent judgments about sampling when
you are critically appraising studies or developing a
sampling plan for your own study. This chapter exam-
ines sampling theory and concepts; sampling plans;
probability and nonprobability sampling methods for
quantitative, qualitative, outcomes, and intervention
research; sample size; and settings for conducting
studies. The chapter concludes with a discussion of the
process for recruiting and retaining subjects or partici-
pants for study samples in various settings.
Sampling Theory
Sampling involves selecting a group of people, events,
behaviors, or other elements with which to conduct a
15 CHAPTER
Sampling
study. A sampling plan defines the process of making
the sample selections; sample denotes the selected
group of people or elements included in a study. Sam-
pling decisions have a major impact on the meaning
and generalizability of the findings.
Sampling theory was developed to determine math-
ematically the most effective way to acquire a sample
that would accurately reflect the population under
study. The theoretical, mathematical rationale for
decisions related to sampling emerged from survey
research, although the techniques were first applied to
experimental research by agricultural scientists. One
of the most important surveys that stimulated improve-
ments in sampling techniques was the U.S. census.
Researchers have adopted the assumptions of sam-
pling theory identified for the census surveys and
incorporated them within the research process
(Thompson, 2002).
Key concepts of sampling theory are (1) popula-
tions, (2) elements, (3) sampling criteria, (4) represen-
tativeness, (5) sampling errors, (6) randomization,
(7) sampling frames, and (8) sampling plans. The fol-
lowing sections explain these concepts; later in the
chapter, these concepts are used to explain various
sampling methods.
Populations and Elements
The population is a particular group of people, such
as people who have had a myocardial infarction, or
type of element, such as nasogastric tubes, that is
the focus of the research. The target population is the
entire set of individuals or elements who meet the
sampling criteria, such as women who have experi-
enced a myocardial infarction in the past year. Figure
15-1 shows the relationships among the population,
target population, and accessible populations. An
accessible population is the portion of the target
population to which the researchers have reasonable
access. The accessible population might be elements
351
352 UNIT TWO  The Research Process
Population
Target Population
(Determined by sampling
criteria)
Accessible Population
(Available to researcher)
Sample
(Selected with sampling
method)
Subject, Research
Participant,
or Informant
Figure 15-1  Population, sample, and subject selected for a study.
within a country, state, city, hospital, nursing unit, or
clinic, such as the adults with diabetes in a primary
care clinic in Fort Worth, Texas. The sample is
obtained from the accessible population by a particu-
lar sampling method, such as simple random sam-
pling. The individual units of the population and
sample are called elements. An element can be a
person, event, behavior, or any other single unit of
study. When elements are persons, they are usually
referred to as subjects or research participants or
informants (see Figure 15-1). The term used by
researchers depends of the philosophical paradigm
that is reflected in the study and the design. The term
subject, and sometimes research participant, is used
within the context of the postpositivist paradigm of
quantitative research (see Chapter 2). The term study
or research participant or informant is used in the
context of the naturalistic paradigm of qualitative
research (Fawcett & Garity, 2009; Munhall, 2012).
In quantitative, intervention, and outcomes research,
the findings from a study are generalized first to the
accessible population and then, if appropriate, more
abstractly to the target population.
Generalizing means that the findings can be
applied to more than just the sample under study
because the sample is representative of the target pop-
ulation. Because of the importance of generalizing,
there are risks to defining the accessible population
too narrowly. For example, a narrow definition of the
accessible population reduces the ability to generalize
from the study sample to the target population and
diminishes the meaningfulness of the findings. Biases
may be introduced that make generalization to the
broader target population difficult to defend. If the
accessible population is defined as individuals in a
white, upper-middle-class setting, one cannot general-
ize to nonwhite or lower income populations. These
biases are similar to biases that may be encountered
in a nonrandom sample (Thompson, 2002).
In some studies, the entire population is the target
of the study. These studies are referred to as popula-
tion studies (Barhyte, Redman, & Neill, 1990). Many
of these studies use data available in large databases,
such as the census data or other government-maintained
databases. Epidemiologists sometimes use entire
populations for their large database studies. In other
studies, the entire population of interest in the study
is small and well defined. For example, one could
conduct a study in which the defined population was
all living recipients of heart and lung transplants.
In some cases, a hypothetical population is defined
for a study. A hypothetical population assumes the
presence of a population that cannot be defined accord-
ing to sampling theory rules, which require a list of all
members of the population. For example, individuals
who successfully lose weight would be a hypothetical
population. The number of individuals in the popula-
tion, who they are, how much weight they have lost,
how long they have kept the weight off, and how they
achieved the weight loss are unknown. Some popula-
tions are elusive and constantly changing. For example,
identifying all women in active labor in the United
States, all people grieving the loss of a loved one,
or all people coming into an emergency department
would be impossible.
Sampling or Eligibility Criteria
Sampling criteria, also referred to as eligibility cri-
teria, include a list of characteristics essential for
membership or eligibility in the target population. The
criteria are developed from the research problem, the
purpose, a review of literature, the conceptual and
operational definitions of the study variables, and the
design. The sampling criteria determine the target
population, and the sample is selected from the acces-
sible population within the target population (see
Figure 15-1). When the study is complete, the findings
are generalized from the sample to the accessible
population and then to the target population if the
study has a representative sample (see the next
section).
You might identify broad sampling criteria for a
study, such as all adults older than 18 years of age able
to read and write English. These criteria ensure a large
target population of heterogeneous or diverse poten-
tial subjects. A heterogeneous sample increases your
ability to generalize the findings to a larger target
population. In descriptive or correlational studies, the
sampling criteria may be defined to ensure a hetero-
geneous population with a broad range of values for

the variables being studied. However, in quasi-
experimental or experimental studies, the primary
purpose of sampling criteria is to limit the effect of
extraneous variables on the particular interaction
between the independent and dependent variables. In
these types of studies, the sampling criteria need to be
specific and designed to make the population as
homogeneous or similar as possible to control for the
extraneous variables. Subjects are selected to maxi-
mize the effects of the independent variable and mini-
mize the effects of variation in other extraneous
variables so that they have a limited impact on the
dependent variable scores.
Sampling criteria may include characteristics such
as the ability to read, to write responses on the data
collection instruments or forms, and to comprehend
and communicate using the English language. Age
limitations are often specified, such as adults 18 years
and older. Subjects may be limited to individuals who
are not participating in any other study. Persons who
are able to participate fully in the procedure for obtain-
ing informed consent are often selected as subjects. If
potential subjects have diminished autonomy or are
unable to give informed consent, consent must be
obtained from their legal representatives. Thus, persons
who are legally or mentally incompetent, terminally
ill, or confined to an institution are more difficult to
access as subjects (see Chapter 9). However, sampling
criteria should not become so restrictive that the
researcher cannot find an adequate number of study
participants.
A study might have inclusion or exclusion sam-
pling criteria (or both). Inclusion sampling criteria
are characteristics that a subject or element must
possess to be part of the target population. Exclusion
sampling criteria are characteristics that can cause a
person or element to be excluded from the target popu-
lation. Researchers need to provide logical reasons for
their inclusion and exclusion sampling criteria, and
certain groups should not be excluded without justifi-
cation. In the past, some groups, such as women,
ethnic minorities, elderly adults, and poor people,
were unnecessarily excluded from studies (Larson,
1994). Today, federal funding for research is strongly
linked to including these populations in studies. Exclu-
sion criteria limit the generalization of the study find-
ings and should be carefully considered before being
used in a study.
Twiss et al. (2009) conducted a quasi-experimental
study to examine the effects of strength and weight
training (ST) exercises on muscle strength, balance,
and falls of breast cancer survivors (BCSs) with
bone loss (population). This study included clearly
CHAPTER 15  Sampling 353
identified inclusion and exclusion sampling or eligibil-
ity criteria that are presented in the following excerpt.
“Women were included if they were 35-77 years of
age, had a history of stage 0 (in situ), I, or II breast
cancer, a BMD [bone mineral density] T-score of −1.0
or less at any of three sites (hip, spine, forearm), were
at least 6 months post breast-cancer treatment and
12 months postmenopausal, resided within 100 miles
of one of four research sites (Omaha, Lincoln,
Kearney, and Scottsbluff, NE), and had their physi-
cians’ permission to participate [inclusion sampling
criteria]. Women were excluded if they (a) had a
recurrence of breast cancer; (b) were currently taking
hormone therapy, bisphosphonates, glucocorticoste-
roids, or other drugs affecting bone; (c) were cur-
rently engaging in ST exercises; (d) had a body mass
index (BMI) of 35 or greater; (e) had serum calcium,
creatinine, or thyroid stimulating hormone (if on
thyroid therapy) outside normal limits; or (f) had
active gastrointestinal problems or other conditions
that prohibited ST exercises, risedronate, calcium, or
vitamin D intake [exclusion sampling criteria].” (Twiss
et al., 2009, p. 72)
Twiss et al. (2009) identified specific inclusion and
exclusion sampling criteria to designate the subjects
in the target population precisely. These sampling cri-
teria probably were narrowly defined by the research-
ers to promote the selection of a homogeneous sample
of postmenopausal BCSs with bone loss. These inclu-
sion and exclusion sampling criteria were appropriate
for the study to reduce the effect of possible extrane-
ous variables that might have an impact on the treat-
ment (ST exercises) and the measurement of the
dependent variables (muscle strength, balance, and
falls). Because this is a quasi-experimental study that
examined the impact of the treatment on the dependent
or outcome variables, the increased controls imposed
by the sampling criteria strengthened the likelihood
that the study outcomes were caused by the treatment
and not by extraneous variables. Twiss et al. (2009)
found significant improvement in muscle strength and
balance for the treatment group but no significant dif-
ference in the number of falls between the treatment
and comparison groups.
Sample Representativeness
For a sample to be representative, it must be similar
to the target population in as many ways as possible.
It is especially important that the sample be
354 UNIT TWO  The Research Process
representative in relation to the variables you are
studying and to other factors that may influence the
study variables. For example, if your study examines
attitudes toward acquired immunodeficiency syn-
drome (AIDS), the sample should represent the distri-
bution of attitudes toward AIDS that exists in the
specified population. In addition, a sample must rep-
resent the demographic characteristics, such as age,
gender, ethnicity, income, and education, which often
influence study variables.
The accessible population must be representative
of the target population. If the accessible population
is limited to a particular setting or type of setting, the
individuals seeking care at that setting may be differ-
ent from the individuals who would seek care for the
same problem in other settings or from individuals
who self-manage their problems. Studies conducted in
private hospitals usually exclude poor patients, and
other settings could exclude elderly or undereducated
patients. People who do not have access to care are
usually excluded from health-focused studies. Sub-
jects and the care they receive in research centers are
different from patients and the care they receive in
community clinics, public hospitals, veterans’ hospi-
tals, and rural health clinics. Obese individuals who
choose to enter a program to lose weight may differ
from obese individuals who do not enter a program.
All of these factors limit representativeness and limit
our understanding of the phenomena important in
practice.
Representativeness is usually evaluated by compar-
ing the numerical values of the sample (a statistic
such as the mean) with the same values from the target
population. A numerical value of a population is called
a parameter. We can estimate the population param-
eter by identifying the values obtained in previous
studies examining the same variables. The accuracy
with which the population parameters have been esti-
mated within a study is referred to as precision. Preci-
sion in estimating parameters requires well-developed
methods of measurement that are used repeatedly in
several studies. You can define parameters by conduct-
ing a series of descriptive and correlational studies,
each of which examines a different segment of the
target population; then perform a meta-analysis to
estimate the population parameter (Thompson, 2002).
Sampling Error
The difference between a sample statistic and a popu-
lation parameter is called the sampling error (Figure
15-2). A large sampling error means that the sample
is not providing a precise picture of the population; it
is not representative. Sampling error is usually larger
Sampling error
Population Sample
Population Sample
mean mean
Figure 15-2  Sampling error.
with small samples and decreases as the sample size
increases. Sampling error reduces the power of a
study, or the ability of the statistical analyses con-
ducted to detect differences between groups or to
describe the relationships among variables (Aberson,
2010; Cohen, 1988). Sampling error occurs as a result
of random variation and systematic variation.
Random Variation
Random variation is the expected difference in
values that occurs when one examines different sub-
jects from the same sample. If the mean is used to
describe the sample, the values of individuals in that
sample will not all be exactly the same as the sample
mean. Values of individual subjects vary from the
value of the sample mean. The difference is random
because the value of each subject is likely to vary in
a different direction. Some values are higher and
others are lower than the sample mean. The values are
randomly scattered around the mean. As the sample
size becomes larger, overall variation in sample values
decreases, with more values being close to the sample
mean. As the sample size increases, the sample mean
is also more likely to have a value similar to that of
the population mean.
Systematic Variation
Systematic variation, or systematic bias, is a conse-
quence of selecting subjects whose measurement
values are different, or vary, in some specific way from
the population. Because the subjects have something
in common, their values tend to be similar to the
values of others in the sample but different in some
way from the values of the population as a whole.
These values do not vary randomly around the popula-
tion mean. Most of the variation from the mean is in
the same direction; it is systematic. All the values in
the sample may tend to be higher or lower than the
mean of the population (Thompson, 2002).

For example, if all the subjects in a study examin-
ing some type of healthcare knowledge have an intel-
ligence quotient (IQ) higher than 120, many of their
scores will likely be higher than the mean of a popula-
tion that includes individuals with a wide variation in
IQ, such as IQs that range from 90 to 130. The IQs of
the subjects have introduced a systematic bias. This
situation could occur, for example, if all the subjects
were college students, which has been the case in
the development of many measurement methods in
psychology.
Because of systematic variance, the sample mean
is different from the population mean. The extent of
the difference is the sampling error (see Figure 15-2).
Exclusion criteria tend to increase the systematic bias
in the sample and increase the sampling error. An
extreme example of this problem is the highly restric-
tive sampling criteria used in some experimental
studies that result in a large sampling error and greatly
diminished representativeness.
If the method of selecting subjects produces a
sample with a systematic bias, increasing the sample
size would not decrease the sampling error. When a
systematic bias occurs in an experimental study, it can
lead the researcher to believe that a treatment has
made a difference when, in actuality, the values would
be different even without the treatment. This situation
usually occurs because of an interaction of the system-
atic bias with the treatment.
Refusal and Acceptance Rates in Studies
Systematic variation or bias is most likely to occur
when the sampling process is not random. However,
even in a random sample, systematic variation can
occur if potential subjects decline participation. Sys-
tematic bias increases as the subjects’ refusal rate
increases. A refusal rate is the number and percent-
age of subjects who declined to participate in the
study. High refusal rates to participate in a study
have been linked to individuals with serious physical
and emotional illnesses, low socioeconomic status,
and weak social networks (Neumark, Stommel,
Given, & Given, 2001). The higher the refusal
rate, the less the sample is representative of the
target population. The refusal rate is calculated by
dividing the number of potential subjects refusing
to participate by the number of potential subjects
meeting sampling criteria and multiplying the results
by 100%.
Refusal rate formula = number potential subjects
refusing to participate ÷ number potential
subjects meeting sample criteria × 100%
CHAPTER 15  Sampling 355
For example, if 200 potential subjects met the sam-
pling criteria, and 40 refused to participate in the
study, the refusal rate would be 20%.
Refusal rate = 40 (number refusing) ÷
200 (number meeting sampling criteria) =
0.2 × 100% = 20%
Sometimes researchers provide an acceptance
rate, or the number and percentage of the subjects
who agree to participate in a study, rather than a
refusal rate. The acceptance rate is calculated by
dividing the number of potential subjects who agree
to participate in a study by the number of potential
subjects who meet sampling criteria and multiplying
the result by 100%.
Acceptance rate formula = number potential
subjects agreeing to participate ÷ number
potential subjects meeting sample criteria
× 100%
If you know the refusal rate, you can also subtract
the refusal rate from 100% to obtain the acceptance
rate. Usually researchers report either the acceptance
rate or the refusal rate but not both. In the example
mentioned earlier, 200 potential subjects met the sam-
pling criteria; 160 agreed to participate in the study,
and 40 refused.
Acceptance rate = 160 (number accepting) ÷
200 (number meeting sampling criteria) =
0.8 × 100% = 80%
Acceptance rate = 100% - refusal rate or
100% - 20% = 80%
Sample Attrition and Retention
Rates in Studies
Systematic variation can also occur in studies with
high sample attrition. Sample attrition is the with-
drawal or loss of subjects from a study. Systematic
variation is greatest when a high number of subjects
withdraw from the study before the data have been
collected or when a large number of subjects withdraw
from one group but not the other in the study (Ker-
linger & Lee, 2000; Thompson, 2002). In studies
involving a treatment, subjects in the control group
who do not receive the treatment may be more likely
to withdraw from the study. Sample attrition should
be reported in the published study to determine if the
final sample represents the target population. Research-
ers also need to provide a rationale for subjects with-
drawing from the study and to determine if they are
356 UNIT TWO  The Research Process
different from the subjects who complete the study.
The sample is most like the target population if the
attrition rate is low (<10% to 20%) and the subjects
withdrawing from the study are similar to the subjects
completing the study. Sample attrition rate is calcu-
lated by dividing the number of subjects withdrawing
from a study by the sample size and multiplying the
results by 100%.
Sample attrition rate formula = number subjects
withdrawing ÷ sample size × 100%
For example, if a study had a sample size of 160, and
40 people withdrew from the study, the attrition rate
would be 25%.
Attrition rate = 40 (number withdrawing) ÷
160 (sample size) = 0..25 × 100% = 25%
The opposite of the attrition rate is the retention
rate, or the number and percentage of subjects com-
pleting the study. The higher the retention rate, the
more representative the sample is of the target popula-
tion, and the more likely the study results are an accu-
rate reflection of reality. Often researchers identify
either the attrition rate or the retention rate but not
both. It is better to provide a rate in addition to the
number of subjects withdrawing or completing a
study. In the example just presented with a sample size
of 160, if 40 subjects withdrew from the study, then
120 subjects were retained or completed the study.
The retention rate is calculated by dividing the number
of subjects completing the study by the initial sample
size and multiplying by 100%.
Sample retention rate formula = number subjects
completing study ÷ sample size × 100%
Retention rate = 120 (number retained) ÷
160 (sample size) = 0.75 × 100% = 75%
The study by Twiss et al. (2009) of the effects of
ST exercises on muscle strength, balance, and falls of
BCSs with bone loss was introduced earlier in this
chapter with the discussion of sampling criteria; the
following excerpt presents the acceptance rate and
sample attrition for this study.
“A sample of 249 participants met the screening cri-
teria and they were enrolled in the study.… Of the
249 women, 223 completed the 24-month testing
and were included in the analysis (exercise [treatment
group] = 110; comparison = 113). The remaining 26
women (exercise = 14; comparison = 12) withdrew
from the study before 24 months. Reasons for with-
drawal included the desire for a different exercise
program (n = 7); insufficient time (n = 6); intolerance
to meds (n = 5); cancer recurrence (n = 5); health
problems (n = 2); and relocation (n = 1).” (Twiss et al.,
2009, p. 22)
Twiss et al. (2009) identified that 249 participants
or subjects met the sampling criteria and 249 were
enrolled in the study indicating that the acceptance
rate for the study was 100%. The sample retention
was 223 women for a retention rate of 90% (223 ÷
249 × 100% = 89.6% = 90%), and the sample attrition
rate was 26 women for an attrition rate of 10% (100%
− 90% = 10%). The treatment group retention was
110 women with a retention rate of 89% (110 ÷ 124
× 100% = 88.7% = 89%). The comparison group
retention was 113 women with a retention rate of
90% (113 ÷ 125 = 90.4% = 90%). This study has an
excellent acceptance rate (100%) and a very strong
sample retention rate of 90% for a 24-month-long
study. The retention rates for both groups were very
strong and comparable (treatment group 89% and
comparison group 90%). Twiss et al. (2009) also pro-
vided a rationale for the subjects’ attrition, and the
reasons were varied and seemed appropriate and
typical for a study lasting 24 months. The acceptance
rate, the sample and group retention rates, and the
reasons for subjects’ attrition indicate limited poten-
tial for systematic variation in the study sample. The
likelihood is increased that the sample is representa-
tive of the target population and the results are an
accurate reflection of reality. The study would have
been strengthened if the researchers would have
included not only the numbers but also the sample
and group retention rates.
Randomization
From a sampling theory point of view, randomization
means that each individual in the population should
have a greater than zero opportunity to be selected for
the sample. The method of achieving this opportunity
is referred to as random sampling. In experimental
studies that use a control group, subjects are randomly
selected and randomly assigned to either the control
group or the experimental group. The use of the
term control group—the group not receiving the
treatment—is usually limited to studies using random
sampling and random assignment to the treatment and
control groups. The control group usually receives no
care. If nonrandom sampling methods are used for
sample selection, the group not receiving a treatment

receives usual or standard care and is generally
referred to as a comparison group. With a compari-
son group, there is an increase in the possibility of
preexisting differences between that group and the
experimental group receiving the treatment.
Random sampling increases the extent to which
the sample is representative of the target population.
However, random sampling must take place in an
accessible population that is representative of the
target population. Exclusion criteria limit true ran-
domness. Thus, a study that uses random sampling
techniques may have such restrictive sampling criteria
that the sample is not truly random. In any case, it is
rarely possible to obtain a purely random sample for
nursing studies because of informed consent require-
ments. Even if the original sample is random, persons
who volunteer or consent to participate in a study may
differ in important ways from persons who are unwill-
ing to participate. All samples with human subjects
must be volunteer samples, which includes individu-
als willing to participate in the study, to protect the
rights of the individuals (Fawcett & Garity, 2009).
Methods of achieving random sampling are described
later in the chapter.
Sampling Frame
For each person in the target or accessible population
to have an opportunity to be selected for the sample,
each person in the population must be identified. To
accomplish this goal, the researcher must acquire a list
of every member of the population through the use of
the sampling criteria to define membership. This
listing of members of the population is referred to as
the sampling frame. The researcher selects subjects
from the sampling frame using a sampling plan.
Djukic, Kovner, Budin, and Norman (2010) studied
the effect of nurses’ perceived physical work environ-
ment on their job satisfaction and described their sam-
pling frame in the following excerpt.
“The study was conducted at a large urban hospital
in the U.S. northeast region that is a nongovernment,
not-for-profit, general medical and surgical major
teaching hospital. About 1,300 staff RNs [population]
were employed at the hospital at the time of the
study.… A total of 746 RNs who met eligibility criteria
were invited to participate in the study [sampling
frame of target population]. The eligible RNs were
those who had a functioning work e-mail account
and who worked fulltime, on inpatient units, provid-
ing direct patient care.” (Djukic et al., 2010,
pp. 444-445)
CHAPTER 15  Sampling 357
The sampling frame in this study included the
names of the 746 RNs who were asked to participate
in the study.
Sampling Plan
A sampling plan describes the strategies that will be
used to obtain a sample for a study. The plan is devel-
oped to enhance representativeness, reduce systematic
bias, and decrease the sampling error. Sampling strate-
gies have been devised to accomplish these three tasks
and to optimize sample selection. The sampling plan
may use probability (random) sampling methods or
nonprobability (nonrandom) sampling methods.
A sampling method is the process of selecting a
group of people, events, behaviors, or other elements
that represent the population being studied. A sam-
pling method is similar to a design; it is not specific
to a study. The sampling plan provides detail about the
application of a sampling method in a specific study.
The sampling plan must be described in detail for
purposes of critical appraisal, replication, and future
meta-analyses. The sampling method implemented
in a study varies with the type of research being
conducted. Quantitative, outcomes, and intervention
research apply a variety of probability and nonprob-
ability sampling methods. Qualitative research usually
includes nonprobability sampling methods. The sam-
pling methods to be included in this text are identified
in Table 15-1 and are linked to the types of research
that most commonly incorporate them. The following
sections describe the different types of probability and
nonprobability sampling methods most commonly
used in quantitative, qualitative, outcomes, and inter-
vention research in nursing.
Probability (Random)
Sampling Methods
Probability sampling methods have been developed
to ensure some degree of precision in estimations of
the population parameters. Probability samples reduce
sampling error. The term probability sampling
method refers to the fact that every member (element)
of the population has a probability higher than zero of
being selected for the sample. Inferential statistical
analyses are based on the assumption that the sample
from which data were derived has been obtained ran-
domly. Thus, probability sampling methods are often
referred to as random sampling methods. These
samples are more likely to represent the population
than samples obtained with nonprobability sampling
methods. All subsets of the population, which may
differ from one another but contribute to the
358 UNIT TWO  The Research Process

TABLE 15-1  Probability and Nonprobability randomly selected from the sampling frame. Accord-
Sampling Methods Commonly
ing to sampling theory, it is impossible to select a
sample randomly from a population that cannot be
Applied in Nursing Research
clearly defined. Four sampling designs have been
Sampling Method Common Applications developed to achieve probability sampling: simple
Probability Sampling Methods random sampling, stratified random sampling, cluster
sampling, and systematic sampling.
Simple random Quantitative, outcomes, and
sampling intervention research
Stratified random Quantitative, outcomes, and Simple Random Sampling
sampling intervention research Simple random sampling is the most basic of the
Cluster sampling Quantitative, outcomes, and probability sampling methods. To achieve simple
intervention research random sampling, elements are selected at random
Systematic sampling Quantitative, outcomes, and from the sampling frame. This goal can be accom-
intervention research plished in various ways, limited only by the imagina-
Nonprobability Sampling Methods tion of the researcher. If the sampling frame is small,
the researcher can write names on slips of paper, place
Convenience Quantitative, qualitative, outcomes,
sampling and intervention research
the names in a container, mix well, and draw out one
Quota sampling Quantitative, outcomes, and at a time until the desired sample size has been reached.
intervention research Another technique is to assign a number to each name
Purpose or purposeful Qualitative and sometimes in the sampling frame. In large population sets,
sampling quantitative research elements may already have assigned numbers. For
Network or snowball Qualitative and sometimes example, numbers are assigned to medical records,
sampling quantitative research organizational memberships, and professional licenses.
Theoretical sampling Qualitative research The researcher can use a computer to select these
numbers randomly to obtain a sample.
There can be some differences in the probability
for the selection of each element, depending on
parameters of the population, have a chance to be whether the name or number of the selected element
represented in the sample. Probability sampling is replaced before the next name or number is selected.
methods are most commonly applied in quantitative, Selection with replacement, the most conservative
outcomes, and intervention research. random sampling approach, provides exactly equal
There is less opportunity for systematic bias if sub- opportunities for each element to be selected (Thomp-
jects are selected randomly, although it is possible for son, 2002). For example, if the researcher draws
a systematic bias to occur by chance. Using random names out of a hat to obtain a sample, each name must
sampling, the researcher cannot decide that person X be replaced before the next name is drawn to ensure
would be a better subject for the study than person Y. equal opportunity for each subject.
In addition, a researcher cannot exclude a subset of Selection without replacement gives each element
people from selection as subjects because he or she different levels of probability for selection. For
does not agree with them, does not like them, or finds example, if the researcher is selecting 10 subjects from
them hard to deal with. Potential subjects cannot be a population of 50, the first name has a 1 in 5 chance
excluded just because they are too sick, not sick (10 draws, 50 names), or a 0.2 probability, of being
enough, coping too well, or not coping adequately. selected. If the first name is not replaced, the remain-
The researcher, who has a vested interest in the study, ing 49 names have a 9 in 49 chance, or a 0.18 probabil-
could (consciously or unconsciously) select subjects ity, of being selected. As further names are drawn, the
whose conditions or behaviors are consistent with the probability of being selected decreases.
study hypothesis. It is tempting to exclude uncoopera- There are many ways to achieve random selection,
tive or assertive individuals. Random sampling leaves such as with the use of a computer, a random numbers
the selection to chance and decreases sampling error table, drawing names out of a hat, or a roulette wheel.
and increases the validity of the study (Thompson, The most common method of random selection is
2002). the computer, which can be programmed to select a
Theoretically, to obtain a probability sample, the sample randomly from the sampling frame with
researcher must develop a sampling frame that includes replacement. However, some researchers still use a
every element in the population. The sample must be table of random numbers to select a random sample.
 CHAPTER 15  Sampling 359

TABLE 15-2  Section from a Random

patients.… Because some patients accepted the
Numbers Table
intervention before the operation, but changed their
06 84 10 22 56 72 25 70 69 43 mind after the operation (3 patients in total), not all
07 63 10 34 66 39 54 02 33 85 patients participated in the study.” (Degirmen et al.,
03 19 63 93 72 52 13 30 44 40 2010, p. 154)
77 32 69 58 25 15 55 38 19 62
20 01 94 54 66 88 43 91 34 28

Degirmen et al. (2010) clearly identified their target

Table 15-2 shows a section from a random numbers
table. To use a table of random numbers, the researcher
places a pencil or a finger on the table with the eyes
closed. The number touched is the starting place.
Moving the pencil or finger up, down, right, or left,
the researcher uses the numbers in order until the
desired sample size is obtained. For example, the
researcher places a pencil on 58 in Table 15-2, which
is in the fourth column from the left and fourth row
down. If five subjects are to be selected from a popula-
tion of 100 and the researcher decides to go across the
column to the right, the subject numbers chosen are
58, 25, 15, 55, and 38. Table 15-2 is useful only if the
population number is less than 100. However, tables
are available for larger populations, such as the
random numbers table provided in the online resources
for this textbook or the Thompson (2002, pp. 14-15)
sampling text.
Degirmen, Ozerdogan, Sayiner, Kosgeroglu, and
Ayranci (2010, p. 153) conducted a pretest-posttest
randomized controlled experimental study to deter-
mine the effect of hand and foot massage and foot
massage only interventions on the postoperative
pain of women who had a cesarean operation. These
researchers obtained their sample using a simple
random sampling method that is described in the fol-
lowing excerpt from their study.
“The study was conducted in obstetric intensive care
units and services of all the public and university hos-
pitals in the province of Eskisehir, Turkey.… During
the 4 month study, 281 patients attended for the
cesarean operations to the obstetric intensive care
units and services of all hospitals concerned [target
population and settings]. The total 75 study patients
[sample] out of the 281 were selected by random
sampling method from the patients’ presenting orders
[sampling frame] and evenly divided into three groups;
a control group, a foot and hand massage group, and
a foot massage group, each of which included 25
population as women needing cesarean operations, and
the 281 women with presenting orders provided the
sampling frame for the study. The sample of 75 women
was randomly selected, but the researchers did not
indicate the process for the random selection. The use
of a computer to select a sample randomly is usually
the most efficient and unbiased process. The subjects
were evenly divided with 25 in each group, but the
researchers do not indicate if the assignment to groups
was random or based on the convenience of the sub-
jects or researchers. Application of simple random
sampling and the attrition of only three (4%) subjects
from the study seem to provide a sample representative
of the target population. However, the study would
have been strengthened by a discussion of the process
for random sampling and a clarification of how the
subjects were assigned to groups. The outcomes of the
study were that foot and hand massage interventions
significantly reduced postoperative pain experienced
by the women and that foot and hand massage was
significantly more effective than foot massage only.
Stratified Random Sampling
Stratified random sampling is used when the
researcher knows some of the variables in the popula-
tion that are critical to achieving representativeness.
Variables commonly used for stratification are age,
gender, ethnicity, socioeconomic status, diagnosis,
geographical region, type of institution, type of care,
care provider, and site of care. The variable or vari-
ables chosen for stratification need to be correlated
with the dependent variables being examined in the
study. Subjects within each stratum are expected to be
more similar (homogeneous) in relation to the study
variables than they are to be similar to subjects in
other strata or the total sample. In stratified random
sampling, the subjects are randomly selected on the
basis of their classification into the selected strata.
For example, if in conducting your research you
selected a stratified random sample of 100 adult sub-
jects using age as the variable for stratification, the
sample might include 25 subjects in the age range 18
to 39 years, 25 subjects in the age range 40 to 59 years,
25 subjects in the age range 60 to 79 years, and 25
360 UNIT TWO  The Research Process
subjects 80 years or older. Stratification ensures that
all levels of the identified variable, in this example
age, are adequately represented in the sample. With a
stratified random sample, you could use a smaller
sample size to achieve the same degree of representa-
tiveness as a large sample acquired through simple
random sampling. Sampling error decreases, power
increases, data collection time is reduced, and the cost
of the study is lower if stratification is used (Fawcett
& Garity, 2009; Thompson, 2002).
One question that arises in relation to stratification
is whether each stratum should have equivalent
numbers of subjects in the sample (termed dispropor-
tionate sampling) or whether the numbers of subjects
should be selected in proportion to their occurrence in
the population (termed proportionate sampling). For
example, if stratification is being achieved by ethnicity
and the population is 45% white non-Hispanic, 25%
Hispanic nonwhite, 25% African American, and 5%
Asian, your research team would have to decide
whether to select equal numbers of each ethnic group
or to calculate a proportion of the sample. Good argu-
ments exist for both approaches. Stratification is not
as useful if one stratum contains only a small number
of subjects. In the aforementioned situation, if propor-
tions are used and the sample size is 100, the study
would include only five Asians, hardly enough to be
representative. If equal numbers of each group are
used, each group would contain at least 25 subjects;
however, the white non-Hispanic group would be
underrepresented. In this case, mathematically weight-
ing the findings from each stratum can equalize the
representation to ensure proportional contributions of
each stratum to the total score of the sample. Most
textbooks on sampling describe this procedure (Levy
& Lemsbow, 1980; Thompson, 2002; Yates, 1981).
Ulrich et al. (2006) used a stratified random sam-
pling method to obtain their sample of nurse practitio-
ners (NPs) and physician assistants (PAs) for the
purpose of studying the ethical conflict of these health-
care providers associated with managed care. The fol-
lowing excerpt from this study describes the sampling
method used to obtain the final sample of 1536 provid-
ers (833 NPs and 689 PAs).
“A self-administered questionnaire was mailed to
an initial stratified random sample [sampling method]
of 3,900 NPs and PAs practicing in the United
States. The sample was selected from the national
lists provided by Medical Marketing Services, an inde-
pendently owned organization that manages medical
industry lists (www.mmslists.com/main.asp). The list
for PAs was derived from the American Academy of
Physicians Assistants (AAPA), and a comprehensive
list of NPs was derived from the medical and nursing
boards of the 50 states and the District of Columbia
[sampling frames for NPs and PAs].… After undeliv-
erable (1.9%) and other disqualified respondents
(13.2%, i.e., no longer practicing, non-primary-care
practitioner) were removed, the overall adjusted
response rate was 50.6%.” (Ulrich et al., 2006, p.
393)
The study sampling frames for the NPs and PAs are
representative of all 50 states and the District of
Columbia, and the lists for the sampling frames were
from quality sources. The study has a strong response
rate of 50.6% for a mailed questionnaire, and the
researchers identified why certain respondents were
disqualified. The final sample was large (1536 sub-
jects) with strong representation for both NPs (833
subjects) and PAs (689 subjects). The study sample
might have been stronger with a more equal number
of NP and PA subjects. The 833 NPs and 689 PAs add
to 1522 subjects and it is unclear why the sample size
is identified as 1536 unless there are missing data from
subjects. However, the sample was a great strength of
this study and appeared to represent the target popula-
tion of NPs and PAs currently practicing in primary
care in the United States.
Cluster Sampling
Cluster sampling is a probability sampling method
applied when the population is heterogeneous; it is
similar to stratified random sampling but takes advan-
tage of the natural clusters or groups of population
units that have similar characteristics (Fawcett &
Garity, 2009). Cluster sampling is used in two situa-
tions. The first situation is when a simple random
sample would be prohibitive in terms of travel time
and cost. Imagine trying to arrange personal meetings
with 100 people, each in a different part of the United
States. The second situation is in cases in which the
individual elements making up the population are
unknown, preventing the development of a sampling
frame. For example, there is no list of all the heart
surgery patients who complete rehabilitation programs
in the United States. In these cases, it is often possible
to obtain lists of institutions or organizations with
which the elements of interest are associated.
In cluster sampling, the researcher develops a sam-
pling frame that includes a list of all the states, cities,

institutions, or organizations with which elements of
the identified population would be linked. States,
cities, institutions, or organizations are selected ran-
domly as units from which to obtain elements for the
sample. In some cases, this random selection contin-
ues through several stages and is referred to as mul-
tistage cluster sampling. For example, the researcher
might first randomly select states and next randomly
select cities within the sampled states. Hospitals
within the randomly selected cities might then be ran-
domly selected. Within the hospitals, nursing units
might be randomly selected. At this level, either all
the patients on the nursing unit who fit the criteria for
the study might be included, or patients could be ran-
domly selected.
Cluster sampling provides a means for obtaining a
larger sample at a lower cost. However, it has some
disadvantages. Data from subjects associated with the
same institution are likely to be correlated and not
completely independent. This correlation can cause a
decrease in precision and an increase in sampling
error. However, such disadvantages can be offset to
some extent by the use of a larger sample.
Fouladbakhsh and Stommel (2010, p. E8) used
multistage cluster sampling in their study of the
“complex relationships among gender, physical and
psychological symptoms, and use of specific CAM
[complementary and alternative medicine] health
practices among individuals living in the United States
who have been diagnosed with cancer.” These
researchers described their sampling method in the
following excerpt from their study.
“The NHIS [National Health Interview Survey] meth-
odology employs a multistage probability cluster sam-
pling design [sampling method] that is representative
of the NHIS target universe, defined as ‘the civilian
noninstitutionalized population’ (Botman, Moore,
Moriarty, & Parsons, 2000, p. 14; National Center for
Health Statistics). In the first stage, 339 primary sam-
pling units were selected from about 1,900 area sam-
pling units representing counties, groups of adjacent
counties, or metropolitan areas covering the 50 states
and the District of Columbia [1st stage cluster sam-
pling]. The selection included all of the most populous
primary sampling units in the United States and strati-
fied probability samples (by state, area poverty level,
and population size) of the less populous ones. In a
second step, primary sampling units were partitioned
into substrata (up to 21) based on concentrations of
African American and Hispanic populations [2nd stage
CHAPTER 15  Sampling 361
cluster sampling]. In a third step, clusters of dwelling
units form the secondary sampling units selected
from each substratum [3rd stage cluster sampling].
Finally, within each secondary sampling unit, all African
American and Hispanic households were selected for
interviews, whereas other households were sampled
at differing rates within the substrata. Therefore, the
sampling design of the NHIS includes oversampling of
minorities.” (Fouladbakhsh & Stommel, 2010, pp.
E8-E9)
These researchers detailed their use of multistage
cluster sampling and clearly identified the three stages
implemented and the rationale for each stage. The
study had a large, national sample that seemed repre-
sentative of all 50 states and the District of Columbia
with an oversampling of minorities to accomplish the
purpose of the study. The complex cluster sampling
method used in this study provided a representative
sample, which decreases the likelihood of sampling
error and increases the validity of the study findings.
The findings reported by Fouladbakhsh and Stommel
(2010, p. E7) indicated that “CAM practice use was
more prevalent among female, middle-aged, Cauca-
sian, and well-educated subjects. Pain, depression, and
insomnia were strong predictors of practice use, with
differences noted by gender and practice type.”
Systematic Sampling
Systematic sampling can be conducted when an
ordered list of all members of the population is avail-
able. The process involves selecting every kth indi-
vidual on the list, using a starting point selected
randomly. If the initial starting point is not random,
the sample is not a probability sample. To use this
design in your research, you must know the number
of elements in the population and the size of the
sample desired. Divide the population size by the
desired sample size, giving k, the size of the gap
between elements selected from the list. For example,
if the population size is N = 1200 and the desired
sample size is n = 100, then you could calculate the
value of k:
k = population size ÷ sample size desired
Example: k = 1200 (population size) ÷
100 (sample size desired) = 12
Thus, k = 12, which means that every 12th person on
the list would be included in the sample. Some authors
362 UNIT TWO  The Research Process
argue that this procedure does not truly give each
element an opportunity to be included in the sample;
it provides a random but unequal chance for inclusion
(Thompson, 2002).
Researchers must be careful to determine that the
original list has not been set up with any ordering that
could be meaningful in relation to the study. The
process is based on the assumption that the order of
the list is random in relation to the variables being
studied. If the order of the list is related to the study,
systematic bias is introduced. In addition to this risk,
it is difficult to compute sampling error with the use
of this design (Floyd, 1993).
Li and Mukamel (2010, p. S256) conducted a sec-
ondary data analysis of National Nursing Home Survey
(NNHS) data to identify any “racial disparities in the
receipt and documentation of influenza and pneumo-
coccus vaccinations among nursing-home residents.”
The NNHS data were obtained using stratified random
sampling and systematic sampling for selection of the
nursing homes and random sampling of the residents
for interviews. The researchers described the sampling
plan for the NNHS and the participants for their study
in the following study excerpt.
“We obtained the public-use resident file of the
Centers for Disease Control and Prevention’s 2004
NNHS, which includes a nationally representative
sample of nursing homes, their residents, and the
services the residents received.… The 2004 NNHS
involved a stratified 2-stage probability design. The
first stage was the selection of nursing homes strati-
fied by geographical location (state, county, and zip
code), bed-size category (number of beds), and own-
ership status (profit vs. nonprofit). Nursing homes
were finally selected by systematic sampling, with the
probability proportional to bed-size category. The
second stage, sampling of current residents, was
carried out by interviewers at the time of their visits
to the facilities. Individuals were randomly selected
from patient rosters, and a sample of up to 12 current
residents per facility was selected for the final inter-
view. The final NNHS comprised 13,507 residents in
1174 nursing homes, with an overall response rate of
78%.… Our analyses were limited to 11,448 nonHis-
panic Whites and 1384 Blacks, excluding 765 resi-
dents (5.7%) of other races/ethnicities.… The final
sample comprised 10,562 residents for analysis of
influenza and 12,134 residents for the analysis of
pneumococcus vaccinations.” (Li & Mukamel, 2010,
p. S256)
Li and Mukamel (2010) used a national database
developed with strong probability sampling methods
(stratified random sampling and systematic sampling).
The database included a nationally representative
sample of nursing homes and their residents, which
decreases the potential for sampling error and supports
the validity of the findings. The study design was
limited to an examination of data from white and black
subjects, who were the largest groups represented in
the nursing homes surveyed. The sample size was
extremely large, which increases the potential to gen-
eralize the findings to these two ethnic groups. The
study would have been strengthened by a more detailed
discussion of the systematic sampling of the nursing
homes and the random sampling of the residents. The
researchers found that disparities existed in vaccina-
tion coverage of white and black nursing home
residents and recommended interventions to improve
coverage.
Nonprobability (Nonrandom)
Sampling Methods
Commonly Applied in
Quantitative Research
In nonprobability sampling, not every element of the
population has an opportunity to be included in the
sample. Nonprobability sampling methods increase
the likelihood of obtaining samples that are not repre-
sentative of their target populations. However, most
nursing studies use nonprobability sampling, espe-
cially convenience sampling, to select study samples.
In conducting studies in nursing and other health dis-
ciplines, limited subjects are available, and it is often
impossible to obtain a random sample. Researchers
often include any subjects willing to participate who
meet the eligibility criteria.
There are several types of nonprobability (nonran-
dom) sampling designs. Each addresses a different
research need. The five nonprobability sampling
designs described in this textbook are (1) convenience
sampling, (2) quota sampling, (3) purposive or pur-
poseful sampling, (4) network or snowball sampling,
and (5) theoretical sampling. These sampling methods
are applied in both quantitative and qualitative
research. However, convenience sampling and quota
sampling are applied more often in quantitative, out-
comes, and intervention research than in qualitative
studies and are discussed in this section (see Table
15-1). Purposive sampling, network sampling, and
theoretical sampling are more commonly applied in

qualitative studies than in quantitative studies and are
discussed later in this chapter.
Convenience Sampling
In convenience sampling, subjects are included in the
study because they happened to be in the right place at
the right time. Researchers simply enter available sub-
jects into the study until they have reached the desired
sample size. Convenience sampling, also called acci-
dental sampling, is considered a weak approach to
sampling because it provides little opportunity to
control for biases. Multiple biases may exist in conve-
nience sampling; these biases range from minimal to
serious. Researchers need to identify and describe
known biases in their samples. You can identify biases
by carefully thinking through the sample criteria used
to determine the target population and taking steps to
improve the representativeness of the sample. For
example, in a study of home care management of
patients with complex healthcare needs, educational
level would be an important extraneous variable. One
solution for controlling this extraneous variable would
be to redefine the sampling criteria to include only
patients with a high school education. Doing so would
limit the extent of generalization but decrease the bias
created by educational level. Another option would be
to select a population known to include individuals
with a wide variety of educational levels. Data could
be collected on educational level so that the descrip-
tion of the sample would include information on edu-
cational level. With this information, one could judge
the extent to which the sample was representative
with respect to educational level (Thompson, 2002).
Decisions related to sample selection must be care-
fully described to enable others to evaluate the pos-
sibility of biases. In addition, data need be gathered to
allow a thorough description of the sample that can
also be used to evaluate for possible biases. Data on
the sample can be used to compare the sample with
other samples and to estimate the parameters of popu-
lations through meta-analyses.
Many strategies are available for selecting a con-
venience sample. A classroom of students might be
used. Patients who attend a clinic on a specific day,
subjects who attend a support group, patients currently
admitted to a hospital with a specific diagnosis, and
every fifth person who enters the emergency depart-
ment are examples of types of commonly selected
convenience samples.
Convenience samples are inexpensive and acces-
sible, and they usually require less time to acquire than
other types of samples. Convenience samples provide
means to conduct studies on topics that could not be
CHAPTER 15  Sampling 363
examined through the use of probability sampling.
Convenience sampling enables researchers to acquire
information in unexplored areas. According to Ker-
linger and Lee (2000), a convenience sample is prob-
ably not that bad when it is used with reasonable
knowledge and care in implementing a study. Health-
care studies are usually conducted with particular
types of patients experiencing varying numbers of
health problems; these patients often are reluctant to
participate in research. Thus, researchers often find it
very difficult to recruit subjects for their studies and
frequently must use a sample of convenience versus
random sampling to obtain their sample.
O’Shea, Wallace, Griffin, and Fitzpatrick (2011,
p. 35) used convenience sampling to determine the
“effectiveness of a spiritual educational session on
pediatric nurses’ perspectives concerning the provi-
sion of pediatric spiritual care.” The following excerpt
describes their sampling method.
“The setting for the data collection was a large
university-affiliated children’s hospital located in the
northeast. Participants represented a convenience
sample [sampling method] from a potential 355 reg-
istered neonatal and pediatric staff nurses employed
at the hospital [target population]. Forty-one nurses
voluntarily consented to participate. Number of par-
ticipants per session varied from approximately two
to five, depending on the day and time of the session.”
(O’Shea et al., 2011, p. 37)
O’Shea et al. (2011) clearly identified their sam-
pling method, target population, and sample size. The
acceptance rate for the study was 41 neonatal and
pediatric nurses, which is only 12% of the 355 nurses
in the target population. The nurses volunteering to
participate in the study might be different in some way
from the nurses refusing to participate. The small
sample size (N = 41) and low acceptance rate increase
the chance for sampling error and decrease the repre-
sentativeness of the sample. However, all 41 partici-
pants completed this quasi-experimental study (0%
attrition rate), which decreases the potential for bias.
The sample was homogeneous for registered nurse
status, employed on pediatric units, and gender (female
n = 40). The sample was heterogeneous for education
ranging from associate’s degree to master’s in nursing
and for years of experience ranging from less than 2
years to more than 20 years. In a quasi-experimental
study, a homogeneous sample decreases the extrane-
ous variables that might influence the findings. Based
on the sampling method (nonprobability convenience
364 UNIT TWO  The Research Process
sample), small sample size, high refusal rate, and dif-
ferences in the education and years of experience of
the nurses, the findings from this study are best gen-
eralized to the sample and not the accessible or target
populations. The researchers found that the educa-
tional sessions had a positive effect on the nurses’
perspectives toward providing spiritual care, but addi-
tional research is needed to confirm the effect of this
intervention. Additional studies with large conve-
nience samples that have similar results would indicate
the effectiveness of this intervention for practice.
Quota Sampling
Quota sampling uses a convenience sampling tech-
nique with an added feature, a strategy to ensure the
inclusion of subject types or strata in a population that
are likely to be underrepresented in the convenience
sample, such as women, minority groups, elderly
adults, poor people, rich people, and undereducated
adults. This method may also be used to mimic the
known characteristics of the target population or to
ensure adequate numbers of subjects in each stratum
for the planned statistical analyses. The technique is
similar to the technique used in stratified random sam-
pling, but the initial sample is not random. If neces-
sary, mathematical weighting can be used to adjust
sample values so that they are consistent with the
proportion of subgroups found in the population.
Quota sampling offers an improvement over conve-
nience sampling and tends to decrease potential biases.
In most studies in which convenience samples are
used, quota sampling could be used and should be
considered (Thompson, 2002).
Pieper, Templin, Kirsner, and Birk (2010) used
quota sampling to examine the impact of vascular leg
disorders, such as chronic venous disorders and
peripheral arterial disease, on the physical activity
levels of opioid-addicted adults in a methadone-
maintained program. The following excerpt describes
their sampling process.
“The sample (N = 713) was obtained from September
2005 to December 2007 from 12 methadone treat-
ment clinics located in a large urban area [conve-
nience sampling]. The sample was stratified on four
variables: age (25-39 years, 40-49 years, 50-65 years);
gender (male, female); ethnicity (African American,
White); and drug use (nonIDU [injection drug use],
arm/upper body injection only, or legs ± upper body
injection; Pieper, Templin, Kirsner, & Birk, 2009)
[quota sampling]. The purpose of the stratification
was to allow comparisons of type of drug use with
minimal confounding by age, gender, or ethnicity.
Additional inclusion criteria included presence of both
legs, able to walk, and able to speak and understand
English. The analyses reported here are on the 569
participants who completed the revised LDUQ [Legs
in Daily Use Questionnaire], which were edited after
examining the test-retest data from 104 participants,
not included in the 569, who were tested first.”
(Pieper et al., 2010, p. 429)
Pieper et al. (2010) clearly identified that the origi-
nal sample was one of convenience because it included
people attending 12 methadone treatment clinics who
were willing to participate in the study. The quota
sampling involved stratification of the sample on four
variables with a clear rationale for the variables
selected for stratification. The study was completed by
569 participants, but 104 participants were used for
examining the test-retest reliability of the LDUQ and
not included in the final data collection. The study had
an attrition of 40 participants (attrition rate = 40 ÷ 569
× 100% = 7%).
The study by Pieper et al. (2010) has several
strengths in the sampling process. The use of quota
sampling ensured that the study sample was more
representative of the target population than using con-
venience sampling only. In addition, the participants
were obtained from 12 different clinics and the sample
size was large (N = 713 − 104 [used only for instru-
ment reliability testing] = 569) with a small attrition
rate (7%). The sample appeared to be representative
of the target population with limited potential for sam-
pling error. The findings indicated that motivation was
the strongest predictor of physical activity and that
healthcare professionals need to evaluate the vascular
health of legs of drug injection users before encourag-
ing exercise.
Nonprobability Sampling
Methods Commonly Applied
in Qualitative Research
Qualitative research is conducted to gain insights and
discover meaning about a particular experience, situa-
tion, cultural element, or historical event. The intent is
an in-depth understanding of a purposefully selected
sample and not the generalization of the findings from
a randomly selected sample to a target population, as
in quantitative, outcomes, and intervention research. In

qualitative research, experiences, events, and incidents
are more the focus of sampling than people (Marshall
& Rossman, 2011; Munhall, 2012; Patton, 2002). The
researcher attempts to select participants or informants
who can provide extensive information about the expe-
rience or event being studied. For example, if the goal
of your study was to describe the phenomenon of
living with chronic pain, you would purposefully
select participants who were articulate and reflective,
had a history of chronic pain, and were willing to share
their chronic pain experience (Coyne, 1997).
The three common sampling methods applied in
qualitative research are purposive or purposeful sam-
pling, network or snowball sampling, and theoretical
sampling (see Table 15-1). These sampling methods
enable the researcher to select the specific participants
who would provide the most extensive information
about the phenomenon, event, or situation being
studied (Marshall & Rossman, 2011). The sample
selection process can have a profound effect on the
quality of the research and should be described in
enough depth to promote the interpretation of the find-
ings and the replication of the study (Munhall, 2012;
Patton, 2002).
Purposive Sampling
In purposive sampling, sometimes referred to as
purposeful, judgmental, or selective sampling, the
researcher consciously selects certain participants,
elements, events, or incidents to include in the study.
In purposive sampling, qualitative researchers select
information-rich cases, or cases that can teach them
a great deal about the central focus or purpose of the
study (Green & Thorogood, 2004; Patton, 2002).
Efforts might be made to include typical and atypical
participants or situations. Researchers also seek criti-
cal cases, or cases that make a point clearly or are
extremely important in understanding the purpose of
the study (Munhall, 2012). The researcher might select
participants or informants of various ages, participants
with differing diagnoses or illness severity, or partici-
pants who received an ineffective treatment versus an
effective treatment for their illness.
This sampling plan has been criticized because it
is difficult to evaluate the precision of the researcher’s
judgment. How does one determine that the patient or
element was typical or atypical, good or bad, effective
or ineffective? Researchers need to indicate the char-
acteristics that they desire in participants and provide
a rationale for selecting these types of participants to
obtain essential data for their study. Purposive sam-
pling method is used in qualitative research to gain
insight into a new area of study or to obtain in-depth
CHAPTER 15  Sampling 365
understanding of a complex experience or event
(Munhall, 2012).
Sternberg and Barry (2011, p. 64) applied purpo-
sive and snowball sampling methods in conducting
their phenomenological study of “the experiences of
transnational Latina mothers who immigrated to the
United States without legal documentation or their
children.” Snowball sampling, discussed in the next
section, involves current study participants identifying
additional potential study participants who are similar
to them. Sternberg and Barry describe their sampling
methods in the following study excerpt.
“A purposive sample of Latina mothers who immi-
grated to the United States without their children was
selected. Interviews were conducted and analyzed
until saturation and redundancy was achieved
(Munhall, 2007). Eight women in total participated in
the study. Three women visiting a free community
clinic in southeast Florida were invited to participate
in the study by the nurse researcher; the other five
were recruited using snowball sampling. Snowball
sampling was chosen over random sampling because
of the difficulty in gaining access to a population with
so many undocumented members (Munhall, 2007).
Further, it enabled the researcher to establish a trust-
ing relationship with the participants and obtain a
more heterogeneous sample group. To be included in
this study, participants had to be mothers who were
18 years of age or older, Spanish or English speaking,
and immigrants to the United States from Latin
America who had left their child or children in their
country of origin.” (Sternberg & Barry, 2011, pp.
65-66)
Sternberg and Barry (2011) clearly identified their
sampling methods that were appropriate for the quali-
tative study they conducted. The initial three partici-
pants were identified through purposive sampling so
that they could achieve a group reflective of their
sampling criteria. Individuals without legal documen-
tation are hard to locate, so use of snowball sampling
was appropriate to identify five additional partici-
pants. The eight participants provided an adequate-
sized sample because the researchers were able to
reach saturation and redundancy of themes during
their data analysis. The findings from the study
included seven essential themes: “living in extreme
poverty, having hope, choosing to walk from poverty,
suffering through the trip to and across the U.S.-
Mexican border, mothering from afar, valuing family,
366 UNIT TWO  The Research Process
and changing personally” (Sternberg & Barry, 2011,
p. 67).
Network (Snowball) Sampling
Network sampling, sometimes referred to as snow-
ball or chain sampling, holds promise for locating
samples difficult or impossible to obtain in other ways
or that had not been previously identified for study.
Network sampling takes advantage of social networks
and the fact that friends tend to have characteristics in
common. When you have found a few participants
with the necessary criteria, you can ask for their assis-
tance in getting in touch with others with similar char-
acteristics. The first few participants are often obtained
through convenience or purposive sampling methods,
and the sample size is expanded using network or
snowball sampling. This sampling method is occa-
sionally used in quantitative studies, but it is more
commonly used in qualitative studies. In qualitative
research, network sampling is an effective strategy for
identifying participants who know other potential par-
ticipants who can provide the greatest insight and
essential information about an experience or event that
is identified for study (Marshall & Rossman, 2011;
Munhall, 2012; Patton, 2002).
This strategy is also particularly useful for finding
participants in socially devalued populations, such as
alcoholics, child abusers, sex offenders, drug addicts,
and criminals. These individuals are seldom willing to
make themselves known. Other groups, such as
widows, grieving siblings, or individuals successful at
lifestyle changes, can be located using this strategy.
These individuals are outside the existing healthcare
system and are difficult to find. Biases are built into
the sampling process because the participants are not
independent of one another. However, the participants
selected have the expertise to provide the essential
information needed to address the study purpose. The
study by Sternberg and Barry (2011) presented in the
previous section applied snowball or network sam-
pling to identify additional study participants who had
immigrated to the United States without legal docu-
mentation. These researchers clearly identified their
use of snowball sampling and their rationale for using
this method in their study.
Theoretical Sampling
Theoretical sampling is usually applied in grounded
theory research to advance the development of a
selected theory throughout the research process
(Munhall, 2012). The researcher gathers data from any
individual or group that can provide relevant data for
theory generation. The data are considered relevant if
they include information that generates, delimits, and
saturates the theoretical codes in the study needed for
theory generation. A code is saturated if it is complete
and the researcher can see how it fits in the theory. The
researcher continues to seek sources and gather data
until the codes are saturated and the theory evolves
from the codes and the data. Diversity in the sample
is encouraged so that the theory developed covers a
wide range of behavior in varied situations (Marshall
& Rossman, 2011; Patton, 2002).
Beaulieu, Kools, Kennedy, and Humphreys (2011,
p. 41) conducted a qualitative study using grounded
theory methods to “explore and better understand the
reasons for the apparent underuse of emergency con-
traceptive pills (ECPs) in young people in coupled
relationships.” These researchers applied three sam-
pling methods: (1) convenience sampling, (2) snow-
ball sampling, and (3) theoretical sampling. They
described their sampling methods in the following
study excerpt.
“A convenience sample was recruited via public
notices and snowball sampling (Fain, 2004). Inclusion
criteria were women 18 to 25 years of age, English
speaking, with basic knowledge of ECPs, and cur-
rently involved in a sexual relationship with a partner
who was also willing to participate in the study.…
Analysis began simultaneously with data collection as
dictated by the tenets of grounded theory.… The
initial analysis of interviews and field notes consisted
of strategies of open coding and memoing (Glaser &
Strauss, 1967).… As new categories emerged, the
original interview guide was revised and additional
couples were recruited to allow for theoretical sam-
pling, that is, sampling specifically to fill in theoretical
gaps, strengthen categories and their relationships,
and verify or challenge emerging conceptualizations
(Strauss & Corbin, 1998) and forced coding (Charmaz,
2006).” (Beaulieu et al., 2011, p. 43)
Beaulieu et al. (2011) clearly identified their sam-
pling methods that were appropriate for a qualitative
study conducted with grounded theory methodology.
Both convenience and snowball sampling methods
were applied because the researchers wanted an ade-
quate number of couples to participate in their study
and discuss their decision making regarding ECPs
use. Beaulieu et al. also provided a detailed rationale
for their use of theoretical sampling to develop a
theory about decision making of young couples
related to ECPs use. The sampling methods provided
a strong sample of 22 couples, who provided the

essential information for grounded theory develop-
ment (Glaser & Strauss, 1967; Strauss & Corbin,
1998). More details on this study are presented later
in this chapter in the discussion of sample size in
qualitative studies.
Sample Size in
Quantitative Research
One of the questions beginning researchers commonly
ask is, “What size sample should I use?” Historically,
the response to this question has been that a sample
should contain at least 30 subjects for each study vari-
able measured. Statisticians consider 30 subjects as
the minimum number for data on a single variable to
approach a normal distribution. So if a study includes
4 variables, researchers need at least 120 subjects in
their final sample. Researchers are encouraged to
determine the possible attrition rate for their study to
ensure an adequate sample size at the completion of
their study. For example, researchers might anticipate
a 10-15% attrition rate in their study and need to
obtain a sample of 130 to 140 subjects to ensure the
final sample size after attrition is 120. The best method
of determining sample size is a power analysis but if
information is not available to conduct a power analy-
sis, this recommendation of 30 subjects per study vari-
able might be used.
The deciding factor in determining an adequate
sample size for correlational, quasi-experimental, and
experimental studies is power. Power is the capacity
of the study to detect differences or relationships that
actually exist in the population. Expressed another
way, power is the capacity to reject a null hypothesis
correctly. The minimum acceptable power for a
study is commonly recommended to be 0.80 (80%)
(Aberson, 2010; Cohen, 1988; Kraemer & Thiemann,
1987). If you do not have sufficient power to detect
differences or relationships that exist in the popula-
tion, you might question the advisability of conduct-
ing the study. You determine the sample size needed
to obtain sufficient power by performing a power
analysis. Power analysis includes the standard power
(usually 80%), level of significance (usually set at
0.05 in nursing studies), effect size (discussed in the
next section), and sample size.
An increasing number of nurse researchers are
using power analysis to determine sample size, but it
is essential that the results of the power analyses be
included in the published studies. Not conducting a
power analysis for a study and omitting the power
analysis results in a published study are significant
CHAPTER 15  Sampling 367
problems if the study failed to detect significant dif-
ferences or relationships, which might be due to an
inadequate sample size. The calculation of the power
analysis varies with the types of statistical analyses
used to analyze study data. Statistical programs are
available to conduct a power analysis for a study (see
Chapter 21). However, you can get a general idea
about sample size using the power tables in Appendix
F in this textbook.
The adequacy of sample sizes must be evaluated
more carefully in future nursing studies before data
collection. Studies with inadequate sample sizes
should not be approved for data collection unless they
are preliminary pilot studies conducted before a
planned larger study. If it is impossible for you to
obtain a larger sample because of time or numbers of
available subjects, you should redesign your study so
that the available sample is adequate for the planned
analyses. If you cannot obtain a sufficient sample size,
you should not conduct the study.
Large sample sizes are difficult to obtain in nursing
studies, require long data collection periods, and are
costly. In developing the methodology for a study,
you must evaluate the elements of the methodology
that affect the required sample size. Kraemer and
Thiemann (1987) identified the following factors that
must be taken into consideration in determining
sample size:
1. The more stringent the significance level (e.g.,
0.001 versus 0.05), the greater the necessary sample
size. Most nursing studies include a level of sig-
nificance or alpha (α) = 0.05.
2. Two-tailed statistical tests require larger sample
sizes than one-tailed tests. (Tailedness of statistical
tests is explained in Chapters 21 and 25.)
3. The smaller the effect size, the larger the necessary
sample size. The effect size is a determination of
the effectiveness of a treatment on the outcome
(dependent) variable or the strength of the relation-
ship between two variables.
4. The larger the power required, the larger the neces-
sary sample size. Thus, a study requiring a power
of 90% requires a much larger sample than a study
with power set at 80%.
5. The smaller the sample size, the smaller the
power of the study (Aberson, 2010; Cohen, 1988;
Kraemer & Thiemann, 1987).
The factors that must be considered in decisions
about sample size (because they affect power) are
effect size, type of study, number of variables, sensi-
tivity of the measurement methods, and data analysis
techniques. These factors are discussed in the follow-
ing sections.
368 UNIT TWO  The Research Process
Effect Size
Effect is the presence of a phenomenon. If a phenom-
enon exists, it is not absent, and the null hypothesis is
in error. However, effect is best understood when not
considered in a dichotomous way—that is, as either
present or absent. If a phenomenon exists, it exists to
some degree. Effect size (ES) is the extent to which a
phenomenon is present in a population. In this case,
the term effect is used in a broader sense than the term
cause and effect. For example, you might examine the
impact of distraction on the experience of pain during
an injection. To examine this question, you might
obtain a sample of subjects receiving injections and
measure the perception of pain in a group of subjects
who were distracted during injection and a group of
subjects who were not distracted. The null hypothesis
would be: “There is no difference in the level of pain
perceived by the treatment group receiving distraction
than the comparison group receiving no distraction.”
If this were so, you would say that the effect of distrac-
tion on the perception of pain was zero, and the null
hypothesis would be accepted. In another study, you
might be interested in using the Pearson product
moment correlation r to examine the relationship
between coping and anxiety. Your null hypothesis is
that the population r would be zero, or coping is not
related to anxiety (Cohen, 1988).
In a study, it is easier to detect large differences
between groups than to detect small differences.
Strong relationships between variables in a study are
easier to detect than weak relationships. Thus, smaller
samples can detect large ESs; smaller ESs require
larger samples. ESs can be positive or negative because
variables are positively and negatively correlated. A
negative ES is calculated if a treatment causes a
decrease in the study mean, such as an exercise
program that decreases the weight of subjects. Broadly
speaking, the definitions for ES strengths might be as
follows:
Small ES would be <0.3 or <−0.3
Medium ES would be about 0.3 to 0.5 or −0.3 to
−0.5
Large ES would be >0.5 or >−0.5
These broad ranges are provided because the ES
definitions of small, medium, and large vary based on
the analysis being conducted. For example, the ESs
for comparing two means, such as the treatment group
mean with the comparison group mean (expressed as
d), are small = 0.2 or −0.2, medium = 0.5 or −0.5, and
large = 0.8 or −0.8. The ESs for relationships (expressed
as r) might be defined as small = 0.1 or −0.1, medium
= 0.3 or −0.3, and large = 0.5 or −0.5 (Aberson, 2010;
Cohen, 1988).
Extremely small ESs (e.g., <0.1) may not be clini-
cally important because the relationships between the
variables are small or the differences between the treat-
ment and comparison groups are limited. Knowing the
ES that would be regarded as clinically important
allows us to limit the sample to the size needed to
detect that level of ES (Kraemer & Thiemann, 1987).
A result is clinically important if the effect is large
enough to alter clinical decisions. For example, in
comparing glass thermometers with electronic ther-
mometers, an effect size of 0.1° F in oral temperature
is probably not important enough to influence selection
of a particular type of thermometer in clinical practice.
The clinical importance of an ES varies on the basis
of the variables being studied and the population.
ESs vary according to the population being studied.
Researchers must determine the ES for the particular
relationship or effect being studied in a selected popu-
lation. The most desirable source of this information
is evidence from previous studies (Aberson, 2010;
Melnyk & Fineout-Overholt, 2011). The correlation
value (r) is equal to the ES for the relationship between
two variables. For example, if depression is correlated
with anxiety at r = 0.45, then the ES = r = 0.45, a
medium ES.
ES formula for relationships = r
Example: ES = r = 0.45
In published studies with treatments, means and
standard deviations can be used to calculate the ES
(Grove, 2007). For example, if the mean weight loss
for the treatment or intervention group is 5 pounds per
month with a standard deviation (SD) = 4.5, and the
mean weight loss of the control or comparison group
is 1 pound per month with SD = 6.5, you can calculate
the ES, which is usually expressed as d.
ES formula for group differences = d = mean of
the treatment group - mean of the control group
÷ standard deviation of control group
Example: ES = d = 5 − 1 ÷ 6.5 = 4 ÷ 6.5 = 0.615 = 0.62
This calculation can be used only as an estimate of
ES for the study. If the researcher changes the mea-
surement method used, the design of the study, or the
population being studied, the ES will be altered. When
estimating ES based on previous studies, you might
note the ESs vary from 0.33 to 0.45; it is best to choose
the lower ES of 0.33 to calculate a sample size for a
study. The best estimate of a population parameter of

ES is obtained from a meta-analysis in which an esti-
mated population ES is calculated through the use of
statistical values from all studies included in the analy-
sis (Aberson, 2010; Cohen, 1988).
If few relevant studies have been conducted in the
area of interest, a small pilot study can be performed,
and data analysis results can be used to calculate the
ES. If pilot studies are not feasible, dummy power
table analysis can be used to calculate the smallest
ES with clinical or theoretical value. Yarandi (1991)
described the process of calculating a dummy power
table. If all else fails, ES can be estimated as small,
medium, or large. Numerical values would be assigned
to these estimates and the power analysis performed.
Cohen (1988) and Aberson (2010) indicated the
numerical values for small, medium, and large effects
on the basis of specific statistical procedures. In new
areas of research, ESs for studies are usually set as
small (<0.3) (Aberson, 2010).
Jones, Duffy, and Flanagan (2011) conducted a ran-
domized clinical trial to test the efficacy of a nurse-
coached telephone intervention on the distress
symptoms and functional health status of ambulatory
arthroscopic surgery patients. These researchers con-
ducted a power analysis to identify the sample size for
their study, and it is described in the following excerpt.
“The inclusion criteria were as follows: adults (18
years or older) who were able to read and write
English, would undergo ambulatory arthroscopic
surgery under general anesthesia, had telephone
access at home, and were discharged home on the
day of surgery [target population]. According to
power calculations, a sample size of 102 participants
would assure a power >.80, given a significance level
[alpha] of .05, three measurement times, a minimum
correlation of repeated measure of .30, and a low to
moderate effect size (d = .75).” (Jones et al., 2011,
p. 94)
Jones et al. (2011) clearly identified their target
population and the process for determining their
sample size using power analysis. The standard
power of 0.80 or 80% was used, and alpha was set at
0.05, which is common in nursing studies. The focus
of the study was determining differences between the
treatment and comparison groups, so the ES was
expressed as d = 0.75, which is a moderate ES for
examining differences between groups (see previous
discussion of ESs). Jones et al. (2011, p. 92) found
the “intervention participants had significantly less
CHAPTER 15  Sampling 369
symptom distress at 72 hours and 1 week postsurgery
and significantly better overall physical and mental
health at 1 week postsurgery than those who received
usual practice.” The significant results indicate the
study had an adequate sample size to determine dif-
ferences between the intervention or experimental
group and usual practice or comparison group. If the
study findings had been nonsignificant, the research-
ers would have needed to conduct a power analysis
to determine the power achieved in the study.
Type of Study
Descriptive case studies tend to use small samples.
Groups are not compared, and problems related to
sampling error and generalization have little relevance
for such studies. A small sample size may better serve
the researcher who is interested in examining a
situation in depth from various perspectives. Other
descriptive studies, particularly studies using survey
questionnaires, and correlational studies often require
large samples. In these studies, multiple variables may
be examined, and extraneous variables are likely to
affect subject responses to the variables under study.
Statistical comparisons are often made among mul-
tiple subgroups in the sample, requiring that an ade-
quate sample be available for each subgroup being
analyzed. In addition, subjects are likely to be hetero-
geneous in terms of demographic variables, and mea-
surement tools are sometimes not adequately refined.
Although target populations may have been identified,
sampling frames may be unavailable, and parameters
have not usually been well defined by previous studies.
All of these factors decrease the power of the study
and require increases in sample size (Aberson, 2010;
Kraemer & Thiemann, 1987).
In the past, quasi-experimental and experimental
studies often used smaller samples than descriptive
and correlational studies. As control in the study
increases, the sample size can decrease and still
approximate the population. Instruments in these
studies tend to be refined, improving precision.
However, sample size must be sufficient to achieve an
acceptable level of power (0.8) and reduce the risk of
a type II error (indicating the study findings are non-
significant, when they are really significant) (Aberson,
2010; Kraemer & Thiemann, 1987).
The study design influences power, but the design
with the greatest power may not always be the most
valid design to use. The experimental design with
the greatest power is the pretest-posttest design with
a historical control or comparison group. However,
this design may have questionable validity because
of the historical control group. Can the researcher
370 UNIT TWO  The Research Process
demonstrate that the historical control group is com-
parable to the experimental group? The repeated mea-
sures design increases power if the trait being assessed
is relatively stable over time. Designs that use block-
ing or stratification usually require an increase in the
total sample size. The sample size increases in propor-
tion to the number of cells included in the data analy-
sis. Designs that use matched pairs of subjects have
greater power and require a smaller sample (see
Chapter 11 for a discussion of these designs). The
higher the degree of correlation between subjects on
the variable on which the subjects are matched, the
greater the power (Kraemer & Thiemann, 1987).
Kraemer and Thiemann (1987) classified studies
as exploratory or confirmatory. According to their
approach, confirmatory studies should be conducted
only after a large body of knowledge has been
gathered through exploratory studies. Confirmatory
studies are expected to have large samples and to use
random sampling techniques. These expectations are
lessened for exploratory studies. Exploratory studies
are not intended for generalization to large popula-
tions. They are designed to increase the knowledge of
the field of study. For example, pilot or preliminary
studies to test a methodology or provide estimates of
an ES are often conducted before a larger study. In
other studies, the variables, not the subjects, are the
primary area of concern. Several studies may examine
the same variables using different populations. In
these types of studies, the specific population used
may be incidental. Data from these studies may be
used to define population parameters. This informa-
tion can be used to conduct confirmatory studies using
large, randomly selected samples.
Confirmatory studies, such as studies testing the
effects of nursing interventions on patient outcomes
or studies testing the fit of a theoretical model, require
large sample sizes. Clinical trials are being conducted
in nursing for these purposes. The power of these
large, complex studies must be carefully analyzed
(Leidy & Weissfeld, 1991). For the large sample sizes
to be obtained, subjects are acquired in numerous
clinical settings, sometimes in different parts of the
United States. Kraemer and Thiemann (1987) believed
that these studies should not be performed until exten-
sive information is available from exploratory studies.
This information should include meta-analysis and the
definition of a population ES.
Number of Variables
As the number of variables under study grows, the
needed sample size may also increase. Adding vari-
ables such as age, gender, ethnicity, and education to
the analysis plan (just to be on the safe side) can
increase the sample size by a factor of 5 to 10 if the
selected variables are uncorrelated with the dependent
variable. In this case, instead of a sample of 50, you
may need a sample of 250 to 500 if you plan to use
the variables in the statistical analyses. (Using them
only to describe the sample does not cause a problem
in terms of power.) If the variables are highly corre-
lated with the dependent variable, however, the effect
size will increase, and the sample size can be reduced.
Variables included in the data analysis must be
carefully selected. They should be essential to the
research purpose or should have a documented strong
relationship with the dependent variable (Kraemer
& Thiemann, 1987). Sometimes researchers have
obtained sufficient sample size for the primary analy-
ses but failed to plan for analyses involving subgroups,
such as analyzing the data by age categories or by
ethnic groups, which require a larger sample size. A
larger sample size is also needed if multiple dependent
variables have been included.
Measurement Sensitivity
Well-developed instruments measure phenomena with
precision. For example, a thermometer measures body
temperature precisely. Instruments measuring psycho-
social variables tend to be less precise. However, a
scale with strong reliability and validity tends to
measure more precisely than an instrument that is less
well developed. Variance tends to be higher in a less
well-developed tool than in one that is well developed.
An instrument with a smaller variance is preferred
because the power of a test always decreases when
within-group variance increases (Kraemer & Thie-
mann, 1987). If you were measuring anxiety and the
actual anxiety score for several subjects was 80, the
subjects’ scores on a less well-developed scale might
range from 70 to 90, whereas a well-developed scale
would tend to show a score closer to the actual score
of 80 for each subject. As variance in instrument
scores increases, the sample size needed to gain an
accurate understanding of the phenomenon under
study increases.
The range of measured values influences power.
For example, a variable might be measured in 10
equally spaced values, ranging from 0 to 9. ESs vary
according to how near the value is to the population
mean. If the mean value is 5, ESs are much larger in
the extreme values and lower for values near the mean.
If you decided to use only subjects with values of 0
and 9, the ES would be large, and the sample could be
small. The credibility of the study might be question-
able, however, because the values of most individuals

would not be 0 or 9 but rather would tend to be in the
middle range of values. If you decided to include
subjects who have values in the range of 3 to 6,
excluding the extreme scores, the ES would be small,
and you would require a much larger sample. The
wider the range of values sampled, the larger the ES
(Kraemer & Thiemann, 1987). If you had a heteroge-
neous group of study participants, you would expect
them to have a wide range of scores on a depression
scale, which would increase the ES. A strong measure-
ment method has validity and reliability and measures
variables at the interval or ratio level. The stronger the
measurement methods used in a study, the smaller the
sample that is needed to identify significant relation-
ships among variables and differences between groups.
Data Analysis Techniques
Data analysis techniques vary in their ability to detect
differences in the data. Statisticians refer to this as the
power of the statistical analysis. For your data analy-
sis, choose the most powerful statistical test appropri-
ate to the data. Overall, parametric statistical analyses
are more powerful than nonparametric techniques in
detecting differences and should be used if the data
meet criteria for parametric analysis. However, in
many cases, nonparametric techniques are more
powerful if your data do not meet the assumptions
of parametric techniques. Parametric techniques vary
widely in their capacity to distinguish fine differences
and relationships in the data. Parametric and nonpara-
metric analyses are discussed in Chapter 21.
There is also an interaction between the measure-
ment sensitivity and the power of the data analysis
technique. The power of the analysis technique
increases as precision in measurement increases.
Larger samples must be used when the power of the
planned statistical analysis is low.
For some statistical procedures, such as the t-test
and analysis of variance (ANOVA), having equal
group sizes increases power because the effect size
is maximized. The more unequal the group sizes
are, the smaller the effect size. In unequal groups,
the total sample size must be larger (Kraemer &
Thiemann, 1987).
The chi-square (χ2) test is the weakest of the sta-
tistical tests and requires very large sample sizes to
achieve acceptable levels of power. As the number of
categories (cells in the chi-square analysis) in a study
grows, the sample size needed increases. Also, if there
are small numbers in some of the categories, you must
increase the sample size. Kraemer and Thiemann
(1987) recommended that the chi-square test be used
only when no other options are available. In addition,
CHAPTER 15  Sampling 371
the categories should be limited to those essential to
the study.
Sample Size in
Qualitative Research
In quantitative research, the sample size must be large
enough to describe variables, identify relationships
among variables, or determine differences between
groups. However, in qualitative research, the focus is
on the quality of information obtained from the person,
situation, event, or documents sampled versus the size
of the sample (Marshall & Rossman, 2011; Munhall,
2012; Patton, 2002; Sandelowski, 1995). The sample
size and sampling plan are determined by the purpose
and philosophical basis of the study. The sample size
required is determined by the depth of information
needed to gain insight into a phenomenon, explore
and describe a concept, describe a cultural element,
develop a theory, or describe a historical event. The
sample size can be too small when the data collected
lack adequate depth or richness. An inadequate sample
size can reduce the quality and credibility of the
research findings. Many qualitative researchers use
purposive or purposeful sampling methods to select
the specific participants, events, or situations that they
believe would provide them the rich data needed to
gain insights and discover new meaning in an area of
study (Sandelowski, 2000).
The adequacy of the sample size in a study should
be justified by the researchers. Often the number of
participants in a qualitative study is adequate when
saturation of information is achieved in the study area
(Fawcett & Garity, 2009). Saturation of data occurs
when additional sampling provides no new informa-
tion, only redundancy of previously collected data.
Important factors that must be considered in determin-
ing sample size to achieve saturation of data are (1)
scope of the study, (2) nature of the topic, (3) quality
of the data, and (4) study design (Marshall & Rossman,
2011; Morse, 2000; Munhall, 2012; Patton, 2002).
Scope of the Study
If the scope of a study is broad, researchers need
extensive data to address the study purpose, and it
takes longer to reach data saturation. A study with a
purpose that has a broad scope requires more sampling
of participants, events, or documents than a study with
a narrow scope (Morse, 2000). A study that has a clear
focus and provides focused data collection usually has
richer, more credible findings. The depth of a study’s
scope and its clarity of focus influence the number
372 UNIT TWO  The Research Process
of participants needed for the study sample. For
example, fewer participants would be needed to
describe the phenomenon of chronic pain in adults
with rheumatoid arthritis than would be needed to
describe the phenomenon of chronic pain in elderly
adults. A study of chronic pain experienced by elderly
adults has a much broader focus, with less clarity, than
a study of chronic pain experienced by adults with a
specific medical diagnosis of rheumatoid arthritis.
Nature of the Topic
If the topic of your study is clear and the participants
can easily discuss it, fewer individuals are needed to
obtain the essential data. If the topic is difficult to
define and awkward for people to discuss, you will
probably need a larger number of participants or infor-
mants to saturate the data (Morse, 2000; Patton, 2002).
For example, a phenomenological study of the experi-
ence of an adult living with a history of child sexual
abuse is a sensitive, complex topic to investigate. This
type of topic would probably require a greater number
of participants and increased interview time to collect
the essential data.
Quality of the Data
The quality of information obtained from an interview,
observation, or document review influences the sample
size. The higher the quality and richness of the data,
the fewer the research participants needed to saturate
data in the area of study. Quality data are best obtained
from articulate, well-informed, and communicative
participants. These participants are able to share more
rich data in a clear and concise manner. In addition,
participants who have more time to be interviewed
usually provide data with greater depth and breadth.
Qualitative studies require that you critically appraise
the quality of the richness of communication elicited
from the participants, the degree of access provided to
events in a culture, or the number and quality of docu-
ments studied. These characteristics directly affect the
richness of the data collected and influence the sample
size needed to achieve quality study findings (Fawcett
& Garity, 2009; Munhall, 2012).
Study Design
Some studies are designed to increase the number
of interviews with participants. The more interviews
conducted with a participant, the greater the quality of
the data collected. For example, a study design that
includes an interview both before and after an event
would produce more data than a single interview.
Designs that involve interviewing a family or a group
of individuals produce more data than an interview
with a single study participant. In critically appraising
a qualitative study, determine if the sample size is
adequate for the design of the study.
Beaulieu et al. (2011) provided a detailed discus-
sion of how they determined their final sample size.
This qualitative study conducted with grounded theory
methodology was introduced earlier in the discussion
of theoretical sampling. The study focused on devel-
oping a theory about decision making of young adult
couples regarding their use of ECPs. The sample was
obtained with convenience, snowball, and theoretical
sampling and resulted in a sample size of 22 couples.
The following study excerpt provides the researchers’
rationale for the final sample size of their study.
“A convenience sample was recruited via public
notices and snowball sampling.… All interested
young women initiated the first contact with the
researcher by e-mail or telephone.… At the first
meeting, which also included partners, study proce-
dures were reviewed with participants, after which
written consent and demographic information were
obtained.…
Analysis began simultaneously with data collection
as dictated by the tenets of grounded theory.… As
these processes progressed, axial coding was per-
formed to identify core categories and their relation-
ships. As new categories emerged, the original
interview guide was revised and additional couples
were recruited to allow for theoretical sampling, that
is sampling specifically to fill in theoretical gaps,
strengthen categories and their relationships, and
verify or challenge emerging conceptualizations
(Strauss & Corbin, 1998) and focused coding
(Charmaz, 2006). Member checking occurred
throughout the analysis by sharing the preliminary
findings with subsequent couples to meet the require-
ments of confirmability of developing conceptualiza-
tions (Denzin & Lincoln, 2000). Saturation—when no
new categories emerge (Strauss & Corbin, 1998)—
was reached after interviewing 18 couples, but five
more couples were included to ensure comprehen-
sive analysis as well as theoretical verification. As the
analysis continued through the processes of grounded
theorizing, salient categories consistent with contem-
porary grounded theory principles (Clarke, 2005)
were constructed to characterize the experience of
young couples regarding ECPs. A theoretical model
was developed to describe and explain the process of
emergency contraceptive decision making in young
couples.” (Beaulieu et al., 2011, p. 43)

The study by Beaulieu et al. (2011) has many
strengths in the area of sampling, including quality
sampling methods (convenience, snowball, and
theoretical), strong sample size (N = 22 couples),
and conscientious participants. The investigators
provide extensive details of the theoretical sampling
conducted to ensure saturation was achieved with no
new categories emerging. The saturation occurred
after 18 couples, but the researchers interviewed 5
more couples to ensure depth and breadth in the data
for theoretical verification. Beaulieu et al. (2011)
described how they were able to develop successfully
a theoretical model of experiences of young couples
regarding use of ECPs. The study would have been
strengthened by knowing how many study participants
were obtained by each of the sampling methods (con-
venience, snowball, and theoretical). Also the research-
ers mentioned saturation was obtained with 18 couples
but 5 more couples were included or N = 23 but the
sample size identified was N = 22. A rationale is needed
for the attrition of one of the couples from the study.
Research Settings
The setting is the location where a study is conducted.
There are three common settings for conducting
nursing research: natural, partially controlled, and
highly controlled. A natural setting, or field setting,
is an uncontrolled, real-life situation or environment
(Kerlinger & Lee, 2000). Conducting a study in a
natural setting means that the researcher does not
manipulate or change the environment for the study.
Descriptive and correlational quantitative studies and
qualitative studies are often conducted in natural set-
tings. For example, in the study by Beaulieu et al.
(2011) discussed previously, the investigators made no
attempt to manipulate the settings when they con-
ducted the interviews for their qualitative study. All
the data were conducted in natural settings as the
interviews of the couples took place in various public
settings or in their homes. The intent of the study was
to develop a theory of the decision making of young
adult couples regarding ECP use in a natural
environment.
A partially controlled setting is an environment
that the researcher manipulates or modifies in some
way. An increasing number of nursing studies, usually
correlational, quasi-experimental, and experimental
studies, are being conducted in partially controlled
settings. In a study that was introduced earlier in the
discussion of sampling criteria, Twiss et al. (2009)
conducted a quasi-experimental study to determine the
effects of a strength training (ST) intervention on
CHAPTER 15  Sampling 373
muscle strength, balance, and falls of breast cancer
survivors (BCSs) with bone loss. These researchers
used both partially controlled and natural settings in
their studies, which are described in the following
excerpt.
“This was a multisite, randomized controlled trial of
a 24-month multi-component intervention with
follow-up data collection at 36 months.…
“Setting
“Exercise activities were performed in participants’
homes [natural setting] or at investigator-approved
fitness centers [partially controlled setting], and
Biodex and balance testing were performed by physi-
cal therapists at hospitals or rehabilitation centers
[partially controlled settings] at each of the four
sites.” (Twiss et al., 2009, p. 22)
Twiss et al. (2009) partially controlled the fitness
centers by approving them but did not try to control
the specific activities of each center. The researchers
also ensured that the measurements were taken in a
precise and accurate way by an expert (physical thera-
pist) in partially controlled settings of hospitals and
rehabilitation centers. The natural and partially con-
trolled settings seemed appropriate in this study for
the implementation of the ST intervention and for the
precise and accurate measurement of the outcome
variables.
A highly controlled setting is an artificially con-
structed environment developed for the sole purpose
of conducting research. Laboratories, research or
experimental centers, and test units in hospitals or
other healthcare agencies are highly controlled set-
tings where experimental studies are often conducted.
This type of setting reduces the influence of extrane-
ous variables, which enables the researcher to examine
accurately the effect of one variable on another. Highly
controlled settings are commonly used to conduct
experimental research. Sharma, Ryals, Gajewski, and
Wright (2010) conducted an experimental study
to examine the effects of a moderate-intensity
aerobic exercise program on painlike behavior and
neurotrophin-3 (NT-3) in female mice. The rationale
for conducting this study was that the literature and
clinical practice supported the use of aerobic exercise
in reducing pain and improving function in people
with chronic pain, but the molecular basis for these
positive actions was poorly understood. This study
was conducted in a laboratory using a selected type of
374 UNIT TWO  The Research Process
mouse, and the setting is described in the following
excerpt.
“All experiments were approved by the Institutional
Animal Care and Use Committee of the University of
Kansas Medical Center and adhered to the universi-
ty’s animal care guidelines. Forty CF-1 female mice
(weight = 25 g) were used to examine the effects of
moderately intense exercise on primary (muscular)
and secondary (cutaneous) hyperalgesia and NT-3
synthesis. Because women develop wide-spread pain
syndromes at a greater rate than age-matched men,
hyperalgesia was induced in female mice. The mice
were exposed to 12-hour light/dark cycle and had
access to food and water ad libitum.…
Initially, the mice were randomly assigned to either
the acidic saline injection (experimental) group or the
normal saline injection (placebo) group. Five days
after inducing hyperalgesia with acidic saline injection
into the right limb, the animals were further assigned
to either exercise or no-exercise group.… Two
6-lane, motorized treadmills were used for the exer-
cise training.” (Sharma et al., 2010, pp. 715-716)
Sharma et al. (2010) conducted their study in a
highly controlled laboratory setting in terms of the
housing and feeding of the mice, the light and tem-
perature of the environment, implementation of the
treatments, and the measurements of the dependent
variables. Only with animals could this type of setting
control be achieved in conducting this study. This type
of highly controlled setting removes the effect of
numerous extraneous variables, so the effects of the
independent variables on the dependent variables can
be clearly determined. However, because this research
was conducted on animals, the findings cannot be gen-
eralized to humans, and additional research is needed
to determine the molecular basis of the influence of
aerobic exercise on pain and functioning in humans.
Recruiting and Retaining
Research Participants
After a research team makes a decision about the size
of the sample, the next step is to develop a plan for
recruiting research participants, which involves
identifying, accessing, and communicating with
potential study participants who are representative of
the target population. Recruitment strategies differ,
depending on the type of study, population, and
setting. Special attention must focus on recruiting sub-
jects who tend to be underrepresented in studies, such
as minorities, women, children, and elderly adults
(Fulmer, 2001; Gul & Ali, 2010; Hendrickson, 2007;
Hines-Martin, Speck, Stetson, & Looney, 2009). The
sampling plan, initiated at the beginning of data col-
lection, is almost always more difficult than expected.
In addition to subject recruitment, retaining acquired
subjects is critical to achieve an acceptable sample
size and requires researchers to consider the effects of
the data collection strategies on subject attrition.
Retaining research participants involves the partici-
pants or subjects completing the required behaviors of
a study to its conclusion. The problems with retaining
participants increase as the data collection period
lengthens. Some researchers never obtain their planned
sample size because of the problems they encounter
as they try to recruit and retain subjects. These
researchers often are forced to complete their study
with a smaller sample size, which could decrease the
power of the study and potentially produce nonsignifi-
cant results (Aberson, 2010). With an increasing
number of studies being conducted in health care,
recruiting and retaining subjects have become more
complex issues for researchers to manage (Gul & Ali,
2010; McGregor, Parker, LeBlanc, & King, 2010).
Recruiting Research Participants
The effective recruitment of subjects is crucial to the
success of a study. A few studies examining the effec-
tiveness of various strategies of participant recruitment
and retention have appeared in the literature (David-
son, Cronk, Harrar, Catley, & Good, 2010; Hines-
Martin et al., 2009; Whitebird, Bliss, Savik, Lowry, &
Jung, 2010). However, most of the information avail-
able to guide researchers comes from the personal
experiences of skilled researchers, some of whom have
published their ideas (Gul & Ali, 2010; McGregor
et al., 2010). Some of the positive and negative factors
that influence a subject’s decision to participate in a
study are (1) the attitudes and ethics of the researchers,
(2) the subject’s need for a treatment, (3) the subject’s
interest in the study topic, (4) financial compensation,
(5) fear of the unknown, (6) time and travel con-
straints, (7) language barriers, and (8) the nature of the
informed consent (Gul & Ali, 2010; Hine-Martin et al.,
2009; Madsen et al., 2002; Papadopoulos & Lees,
2002; Sullivan-Bolyai et al., 2007).
The researcher’s initial communication with a
potential subject usually strongly affects the subject’s
decision about participating in the study. Therefore,
the approach must be pleasant, positive, informative,
culturally sensitive, and nonaggressive. The researcher

needs to explain the importance of the study and
clarify exactly what the subject will be asked to do,
how much of the subject’s time will be involved, and
what the duration of the study will be. Research par-
ticipants are valuable resources, and the researcher
must communicate this value to the potential subject.
High-pressure techniques, such as insisting that the
subject make an instant decision to participate in a
study, usually lead to resistance and a higher rate of
refusals. If the study involves minorities, researchers
must be culturally competent or knowledgeable and
skilled in relating to the particular ethnic group being
studied (Hines-Martin et al., 2009; Papadopoulos &
Lees, 2002). If the researcher is not of the same culture
as the potential subjects, he or she may employ a data
collector who is of the same culture. Hendrickson
(2007) used a video for recruiting Hispanic women for
her study, and she provided all the details related to
the study in the subjects’ own language in the video.
This approach greatly improved the subjects’ under-
standing of the study and their desire to participate.
If a potential subject refuses to participate in a
study, you must accept the refusal gracefully—in
terms of body language as well as words. Your actions
can influence the decision of other potential subjects
who observe or hear about the encounter. Studies in
which a high proportion of individuals refuse to par-
ticipate have a serious validity problem. The sample
is likely to be biased because often only a certain type
of individual has agreed to participate. You should
keep records of the numbers of persons who refuse
and, if possible, their reasons for refusal. With this
information, you can include the refusal rate in the
published research report with the reasons for refusal.
It would also be helpful if you could determine if the
potential subjects who refused to participate differed
from the individuals who agreed to participate in the
study. This information will help you to determine the
representativeness of your sample (Thompson, 2002).
Recruiting minority subjects for a study can be
particularly problematic. Minority individuals may
be difficult to locate and are often reluctant to partici-
pate in studies because of feelings of being “used”
while receiving no personal benefit from their involve-
ment or because of their distrust of the medical com-
munity. Effective strategies for recruiting minorities
include developing partnerships with target groups,
community leaders, and potential participants in the
community; using active face-to-face recruitment in
nonthreatening settings; and using appropriate lan-
guage to communicate clearly the purpose, benefits,
and risks of the study (Alvarez, Vasquez, Mayorga,
Feaster, & Mitrani, 2006). Hines-Martin et al. (2009)
CHAPTER 15  Sampling 375
studied the recruitment and retention process for inter-
vention research conducted with a sample of primarily
low-income African American women. Their complex,
multistage recruitment strategies are introduced in the
following excerpt.
“Phase 1 involved the development of a recruitment
team, composed of a co-investigator, in addition to
an African American nurse familiar with the target
population, and two women who were long-standing
community members.
Phase 1 activities began with periods of observa-
tion in the community setting and discussions with
community center personnel to improve the investi-
gators’ understanding of who used the community
center services and when. It became increasingly
clear that only two of the three communities felt a
connection with or used the community center rou-
tinely.… Therefore, the recruitment team, with the
assistance from nursing graduate students, walked
every block of the two relevant communities at dif-
ferent times of the day and different days of the week
to better understand when and where community
women could be found in their daily lives.… Com-
munity women were informed of new initiatives at
the center and were provided with recruitment flyers
including pictures of the research team. The recruit-
ment team then undertook usual recruitment activi-
ties, such as meeting with women’s groups in the
communities and recruitment at community fairs.”
(Hines-Martin et al., 2009, pp. 665-666)
If researchers use data collectors in their studies,
they need to verify the data collectors are following
the sampling plan, especially in random samples.
When the data collectors encounter difficult subjects
or are unable to make contact easily, they may simply
shift to the next person without informing the principal
investigator. This behavior could violate the rules of
random sampling and bias the sample. If data collec-
tors do not understand, or do not believe in, the impor-
tance of randomization, their decisions and actions can
undermine the intent of the sampling plan. Thus, data
collectors must be carefully selected and thoroughly
trained. A plan for the supervision and follow-up of
data collectors to increase their accountability should
be developed (Thompson, 2002).
If you conduct a survey study, you may never
have personal contact with the subjects. To recruit
such subjects, you must rely on the use of attention-
getting techniques, persuasively written material, and
376 UNIT TWO  The Research Process
strategies for following up on individuals who do not
respond to the initial written or email communication.
The strategies need to be appropriate to the potential
subjects; mailed surveys are probably still the best
way to obtain information from elderly adults. Because
of the serious problems of low response rates in survey
studies, using strategies to increase the response rate
is critical. For instance, we have received a teabag or
packet of instant coffee with a questionnaire, accom-
panied by a recommendation in the letter to have a cup
of tea or coffee “on” the researcher while we complete
the questionnaire. Creativity is required in the use of
such strategies because they tend to lose their effect
on groups who receive questionnaires frequently. In
some cases, small amounts of money ($1.00 to $5.00)
are enclosed with the letter, which may suggest that
the recipient buy a soft drink or that the money is a
small offering for completing the questionnaire. This
strategy imposes some sense of obligation on the
recipient to complete the questionnaire, but it is not
thought to be coercive. Also, you should plan emailing
or mailings to avoid holidays or times of the year
when activities are high for potential subjects, possi-
bly reducing the return rate.
Researchers frequently use the Internet to recruit
subjects and to collect survey data. This method makes
it easier for you to contact potential subjects and for
the subjects to provide the requested data. However,
an increased number of surveys are being sent by
the Internet, which can decrease the response rate of
potential subjects who are frequently surveyed. In
studies with surveys, the letter emailed to potential
subjects must be carefully composed. It may be your
only chance to persuade the subject to invest the time
needed to complete the questionnaire. You must sell
the reader on the importance of both your study and
his or her response. The tone of your letter will be the
potential subject’s only image of you as a person; yet,
for many subjects, their response to the perception of
you as a person most influences their decision about
completing the questionnaire. Seek examples of letters
sent by researchers who have had high response rates,
and save letters you received to which you responded
positively. You also might pilot-test your letter on
potential research participants who can give you feed-
back about their reactions to the letter’s tone.
The use of follow-up emails, letters, or cards has
been repeatedly shown to raise response rates to
surveys. The timing is important. If too long a period
has lapsed, the potential subject may have deleted the
questionnaire from his or her email box or discarded
the mailed copy. However, sending the follow-up too
soon could be offensive. Before the questionnaires are
emailed or mailed, precise plans need to be made for
monitoring the return of each questionnaire. A bar
graph could be developed to record the return of each
questionnaire as a means of suggesting when the
follow-up mailing or emailing should occur. The
cumulative number and percentage of responses would
be logged on the graph to reflect the overall data col-
lection process. The data from emailed questionnaires
can be immediately analyzed so that researchers
can easily keep track of the numbers of participants
responding. When the daily or weekly responses
decline, a follow-up email or first-class letter could
be sent encouraging individuals to complete the ques-
tionnaire. Study participants and questionnaires are
assigned the same code numbers, and nonrespondents
are identified by checking the list of code numbers of
unreturned questionnaires. A third follow-up question-
naire with a further modified cover letter could be
emailed or mailed to increase the return rate for the
questionnaires.
The factors involved in the decision of whether to
respond to a questionnaire are not well understood.
One factor is the time required to respond; this includes
the time needed to orient to the directions and the
emotional energy necessary to deal with the threats
and anxieties generated by the questions. There is also
a cognitive demand for thinking. Subjects seem to
make a judgment about the relevance of the research
topic and the potential for personal application of find-
ings. Previous experience with questionnaires is also
a deciding factor.
Traditionally, subjects for physiological nursing
studies have been sought in the hospital setting.
However, access to these subjects is becoming more
difficult—in part because of the larger numbers of
nurses and other healthcare professionals now conduct-
ing research. The largest involvement of research sub-
jects within a healthcare agency usually occurs with
medical research and mainly with clinical trials that
include large samples (Gul & Ali, 2010). Nurse
researchers are recruiting subjects from a variety of
clinical settings. Whitebird et al. (2010) identified three
successful recruitment methods to use in healthcare
agencies: (1) identifying potential participants using
administrative databases, (2) obtaining referrals of
potential participants through healthcare providers and
other sources, and (3) approaching directly a known
potential subject. An initial phase of recruitment may
involve obtaining community and institutional support
for the study. Support from other healthcare profession-
als, such as nurses and physicians, and clinical agency
staff is usually crucial to the successful recruitment of
research participants.

Recruitment of subjects for clinical trials requires
a different set of strategies because the recruitment
may be occurring simultaneously in several sites
(perhaps in different cities). Many of these studies
never achieve their planned sample size. The number
of subjects meeting the sampling criteria who are
available in the selected clinical sites may not be as
large as anticipated. Researchers must often screen
twice as many patients as are required for the study to
obtain a sufficient sample size. Screening logs must
be kept during the recruiting period to record data on
patients who met the criteria but were not entered into
the study. Researchers commonly underestimate the
amount of time required to recruit subjects for a clini-
cal trial. In addition to defining the number of subjects
and the time set aside for recruitment, it may be
helpful to develop short-term or interim recruitment
goals designed to maintain a constant rate of patient
entry (Gul & Ali, 2010). Hellard, Sinclair, Forbes,
and Fairley (2001) studied methods to improve the
recruitment and retention of subjects in clinical trials
and found that the four most important strategies were
to (1) use nonaggressive recruitment methods, (2)
maintain regular contact with the participants, (3)
ensure that the participants are kept well informed of
the progress of the study, and (4) provide constant
encouragement to subjects to continue participation.
Sullivan-Bolyai et al. (2007) identified the barriers
and strategies to improve the recruitment of study
participants from clinical settings. Table 15-3 identi-
fies these common barriers to research participant
recruitment and provides possible strategies to manage
them.
Studies may also benefit from the endorsement of
community leaders, such as city officials; key civic
leaders; and leaders of social, educational, religious, or
labor groups. In some cases, these groups may be
involved in planning the study, leading to a sense of
community ownership of the project. Community
groups may also help researchers to recruit subjects for
the study. Subjects who meet the sampling criteria
sometimes are found in the groups assisting with the
study. Endorsement may involve letters of support and,
in some cases, funding. These activities can add legiti-
macy to the study and make involvement in the study
more attractive to potential subjects (Alvarez et al.,
2006; Davidson et al., 2010; Hines-Martin et al., 2009).
Media support can be helpful in recruiting subjects.
Researchers can place advertisements in local news-
papers and church and neighborhood bulletins. Radio
stations can make public service announcements.
Members of the research team can speak to groups
relevant to the study population. Your team can place
CHAPTER 15  Sampling 377
posters in public places, such as supermarkets, drug-
stores, and public laundries. With permission, you can
set up tables in shopping malls with a member of the
research team present to recruit subjects. Plan for pos-
sible challenges in recruitment and include multiple
methods and locations in your application for human
subject approval for your study. Otherwise, you would
need to submit a modified protocol to the institutional
review board when you add a method or site for
recruitment.
Davidson et al. (2010) used multiple strategies to
recruit and retain college smokers in a cessation clini-
cal trial. Their four-phase recruitment process is pre-
sented in the following study excerpt.
“Participants in this study were members of Greek
fraternities and sororities enrolled at a large Midwest-
ern university, and data were collected from 2006 to
2009.… The clinical trial involved testing a four-
session, MI [motivational interviewing] counseling
intervention on smoking cessation. Participants were
recruited from college fraternity and sorority chap-
ters regardless of their interest in quitting smoking.
Recruitment involved four phases. First, out of 41
fraternity and sorority chapters from a large Midwest-
ern university, the 30 chapters with the larger mem-
berships were invited to participate. Second, within
these invited chapters, individuals were recruited to
participate in an initial, 5-minute, 8-item screening
survey (i.e., Screener).
Third, individual members of these 30 chapters
who met the inclusion criteria based on the Screener
and who were interested in participating in the study
were recruited to participate in a more extensive
(30-45 minute) computerized baseline assessment
approximately 1-4 days following the Screener.…
Fourthly, eligible individuals who completed the base-
line assessment were recruited for enrollment in the
clinical trial.” (Davidson et al., 2010, pp. 146-147)
The recruitment for this smoking cessation clinical
trial was accomplished by using the Greek chapters.
Davidson et al. (2010) developed relationships with
these Greek organizations by meeting with leaders
and members and attending special events. To accom-
plish phases two and three, the researchers met with
the participants at convenient times and in accessible
locations. The participants were also provided incen-
tives of food (cookies and pizza), small cash gifts,
and raffles for iPods. These creative strategies
increased the recruitment and retention of the study
participants.
378 UNIT TWO  The Research Process
TABLE 15-3  Barriers to Recruitment with Actions and Strategies for Engaging Health Care Providers
in the Referral Process
Barriers and Actions
HIPAA*
Create alternative recruitment methods
Work burden
Create compensations
Financial disincentives
Recognize that patient numbers or productivity
may be linked to the clinic’s livelihood
Provider competition
Create a partnership with healthcare
providers involved in recruitment so that
they are rewarded and acknowledged for
their participation in the research process
Provider concerns
Demystify research process
Develop a team atmosphere and a spirit of
“we’re all in this together”
Desire to protect patients
Work with healthcare providers to acknowledge
and respect patient decision-making abilities
Encourage healthy partnerships between
patients and healthcare providers
From Sullivan-Bolyai, S., Bova, C., Deatrick, J. A., Knafl, K., Grey, M., Leung, K., et al. (2007). Barriers and strategies for recruiting study participants in
clinical settings. Western Journal of Nursing Research, 29(4), 498–499.
*HIPAA, Health Insurance Portability and Accountability Act.
Strategies
Ask Clinicians to distribute letters to potential study participants
Obtain institutional review board waiver of authorization requirement for the use
or disclosure of personal health information
Work with clinics to secure a consent that meets HIPAA* regulations and allows
the staff to provide names and contact information of patients with specific
conditions that may be of interest to researchers
Recognize and acknowledge the burden that recruitment places on healthcare
providers
Provide Salary support
Provide educational incentives (e.g., purchase laptop, journals, books, pay for
conference attendance in the field under study) for healthcare providers who
do not normally have access to such opportunities as part of their job
Assess administrative or managerial perceptions of healthcare providers’
recruitment-related responsibilities, and if salary support is given, how that
money will be used
Discuss the designated recruitment tasks and responsibilities with the assigned
staff to determine their perceptions and expectations
Assess the clinic’s financial situation and determine if it is realistic, pragmatic, or
feasible to use that site, especially if its funding depends on patient numbers
Help keep participants linked to the clinical site while they are participating in
the study
Develop a research proposal that reflects the clinical site’s philosophical and
policy perspectives and priorities
Include healthcare providers in the development of a study
Hire and pay a clinical staff member to be responsible for introducing the study
to potential participants
Link recruitment activities to nursing clinical ladder or organization values
Maintain open communication between the clinical and research teams regarding
the workings of the study
Assess healthcare providers’ perceptions of research
Encourage healthcare providers to participate in developing the research proposal
Include healthcare providers in developing study-related manuscripts
Include healthcare providers in research team meetings at a mutually convenient
time
Express appreciation in an ongoing basis for healthcare providers’ involvement
in recruitment process
Share recruitment status information on a monthly basis with healthcare
providers
Share pilot or feasibility data with healthcare providers to support the study
rationale and choice of specific methods
Acknowledge responsibility of healthcare providers to protect patients from harm
Address concerns of healthcare providers by emphasizing the pilot data that
supports the protocol
Model respectful partnerships with study participants

Retaining Subjects
A serious problem in many studies is subject retention,
and sometimes participant attrition cannot be avoided.
Subjects move, die, or withdraw from a treatment. If
you must collect data at several points over time,
subject attrition can become a problem. Subjects who
move frequently and subjects without phones pose a
particular problem. Numerous strategies have been
found to be effective in maintaining the sample. It is
a good idea to obtain the names, email addresses, and
phone numbers (cell and home numbers if possible)
of at least two family members or friends when you
enroll the participant in the study. Ask if the partici-
pant would agree to give you access to unlisted phone
numbers in the event of changes in his or her number.
In some studies, subjects are reimbursed for time
and expenses related to participation. A bonus payment
may be included for completing a certain phase of the
study. Gifts can be used in place of money. Sending
greeting cards for birthdays and holidays helps main-
tain contact. Researchers found that money was more
effective than gifts in retaining subjects in longitudinal
studies. However, some people pointed out the moral
issues related to providing monetary payment to sub-
jects. This strategy can compromise the voluntariness
of participation in a study and particularly has the
potential of exploiting low-income persons.
Collecting data takes time. The participant’s time
is valuable and should be used frugally. During data
collection, it is easy to begin taking the participant for
granted. Taking time for social amenities with partici-
pants may also pay off. However, take care that these
interactions do not influence the data being collected.
Beyond that, nurturing subjects participating in the
study is critical. In some situations, providing refresh-
ments and pleasant surroundings is helpful. During the
data collection phase, you also may need to nurture
others who interact with the participants; these may
be volunteers, family, staff, students, or other profes-
sionals. It is important to maintain a pleasant climate
for the data collection process, which pays off in the
quality of data collected and the retention of subjects
(Davidson et al., 2010; Gul & Ali, 2010; Hines-Martin
et al., 2009; McGregor et al., 2010).
Qualitative studies and longitudinal studies require
extensive time commitment from subjects. They are
asked to participate in detailed interviews or to com-
plete numerous forms at various intervals during a
study (Marshall & Rossman, 2011; Munhall, 2012;
Patton, 2002). Sometimes data are collected with
diaries that require daily entries over a set period of
time. These studies face the greatest risk of participant
mortality. Chapters 4 and 12 provide more details on
CHAPTER 15  Sampling 379
the recruitment and retention of research participants
for qualitative studies.
Clinical trials can also require extensive time com-
mitments from subjects. Gul and Ali (2010) mentioned
the importance of overcoming participant barriers to
continuing in a study, such as time to complete data
collection forms, transportation problems, and con-
flicts with work and family commitments. There is no
formula for compensating study participants, but
many studies mention small monetary payments, gifts,
or free health or child care. It is important that the
incentives used to recruit and retain research partici-
pants be documented in the published study. Com-
munication is one of the most important facets to
retaining study participants. Davidson et al. (2010),
whose recruitment strategies were introduced earlier,
describe their success with retention in their smoking
cessation clinical trial in the following excerpt.
“A very high proportion of participants (89%) com-
pleted at least one session (90% treatment; 87%
comparison). The majority (73%) were retained,
completing three or more sessions (75% treatment;
70% comparison), and over half completed the
maximum of four sessions (63% treatment; 61%
comparison). At the follow-up assessment occurring
6 months after the baseline assessment, 79% of the
participants (n = 357) were retained (80% treatment;
78% comparison).” (Davidson et al., 2010, p. 150)
Research participants who have a personal invest-
ment in the study are more likely to complete the
study. This investment occurs through interactions
with and nurturing by the researcher. A combination
of the participant’s personal belief in the significance
of the study, the perceived altruistic motives of the
researcher in conducting the study, the ethical actions
of the researcher, and the nurturing support provided
by the researcher during data collection can greatly
diminish subject attrition (Hines-Martin et al., 2009;
Madsen et al., 2002; McGregor et al., 2010). The
recruitment and retention of subjects will continue to
be significant challenges for researchers, and creative
strategies are needed to manage these challenges.
KEY POINTS
• Sampling involves selecting a group of people,
events, behaviors, or other elements with which to
conduct a study. Sampling denotes the process of
380 UNIT TWO  The Research Process
making the selections; sample denotes the selected
group of elements.
• A sampling plan is developed to increase represen-
tativeness, decrease systematic bias, and decrease
the sampling error; there are two main types of
sampling plans—probability and nonprobability.
• Sampling error includes random variation and
systematic variation. Refusal and attrition rates
are important to calculate in a study to determine
potential systematic variation or bias.
• Four sampling designs have been developed to
achieve probability or random sampling: simple
random sampling, stratified random sampling,
cluster sampling, and systematic sampling.
• In nonprobability (nonrandom) sampling, not every
element of the population has an opportunity for
selection in the sample. The five nonprobability
sampling designs described in this textbook are (1)
convenience sampling, (2) quota sampling, (3)
purposive or purposeful sampling, (4) network or
snowball sampling, and (5) theoretical sampling.
• In quantitative studies, sample size is best deter-
mined by a power analysis, which is calculated
using the level of significance (usually α = 0.05),
standard power of 0.80 (80%), and effect size.
Factors important to sample size in quantitative
research include (1) type of study, (2) number of
variables studied, (3) measurement sensitivity, and
(4) data analysis techniques.
• The number of participants in a qualitative study is
adequate when saturation of information is achieved
in the study area, which occurs when additional
sampling provides no new information, only redun-
dancy of previously collected data. Important
factors that must be considered in determining
sample size to achieve saturation of data are (1)
scope of the study, (2) nature of the topic, (3)
quality of the data, and (4) study design.
• The three common settings for conducting nursing
research are natural, partially controlled, and highly
controlled. A natural setting, or field setting, is an
uncontrolled, real-life situation or environment. A
partially controlled setting is an environment that
the researcher has manipulated or modified in some
way. A highly controlled setting is an artificially
constructed environment, such as a laboratory or
research unit in a hospital, developed for the sole
purpose of conducting research.
• Recruiting and retaining research participants have
become significant challenges in research; some
strategies to assist researchers with these chal-
lenges are provided so that their samples might be
more representative of their target population.
REFERENCES
Aberson, C. L. (2010). Applied power analysis for the behavioral
sciences. New York, NY: Routledge Taylor & Francis Group.
Alvarez, R. A., Vasquez, E., Mayorga, C. C., Feaster, D. J., &
Mitrani, V. B. (2006). Increasing minority research participation
through community organization outreach. Western Journal of
Nursing Research, 28(5), 541–560.
Barhyte, D. Y., Redman, B. K., & Neill, K. M. (1990). Popula-
tion or sample: Design decision. Nursing Research, 39(5),
309–310.
Beaulieu, R., Kools, S. M., Kennedy, H. P., & Humphreys, J. (2011).
Young adult couples’ decision making regarding emergency con-
traceptive pills. Journal of Nursing Scholarship, 43(1), 41–48.
Botman, S. L., Moore, T. F., Moriarty, C. L. & Parsons, V. L. (2000).
Design and estimation for the National Health Interview Survey,
1995-2004. National Center for Health Statistics, Vital Health
Statistics, Series 2, No. 130, 1–32.
Charmaz, K. (2006). Constructing grounded theory: A practical
guide through qualitative analysis. Thousand Oaks, CA: Sage.
Clarke, A. (2005). Situational analysis grounded theory after the
postmodern turn. Thousand Oaks, CA: Sage.
Cohen, J. (1988). Statistical power analysis for the behavioral sci-
ences (2nd ed.). New York, NY: Academic Press.
Coyne, I. T. (1997). Sampling in qualitative research. Purposeful
and theoretical sampling; merging or clear boundaries. Journal of
Advanced Nursing, 26(3), 623–630.
Davidson, M. M., Cronk, N. J., Harrar, S., Catley, D., & Good, G.
E. (2010). Strategies to recruit and retain college smokers in ces-
sation trials. Research in Nursing & Health, 33(2), 144–155.
Degirmen, N., Ozerdogan, N., Sayiner, D., Kosgeroglu, N., &
Ayranci, U. (2010). Effectiveness of foot and hand massage in
post-cesarean pain control in a group of Turkish pregnant women.
Applied Nursing Research, 23(3), 153–158.
Denzin, N., & Lincoln, Y. (2000). Handbook of qualitative research
(2nd ed.). Thousand Oaks, CA: Sage.
Djukic, M., Kovner, C., Budin, W. C., & Norman, R. (2010). Physi-
cal work environment: Testing an expanded model of job satisfac-
tion in a sample of registered nurses. Nursing Research, 59(6),
441–451.
Fain, J. (2004). Reading, understanding, and applying nursing
research (2nd ed.). Philadelphia, PA: F. A. Davis Company.
Fawcett, J., & Garity, J. (2009). Evaluating research for evidence-
based nursing practice. Philadelphia, PA: F. A. Davis.
Floyd, J.A. (1993). Systematic sampling: Theory and clinical
methods. Nursing Research, 42(5), 290–293.
Foulabakhsh, J. M., & Stommel, M. (2010). Gender, symptom expe-
rience, and use of complementary and alternative medicine prac-
tices among cancer survivors in the U.S. cancer population.
Oncology Nursing Forum, 37(1), E7–E15. Retrieved from Oncol-
ogy Nursing Forum online.
Fulmer, T. (2001). Editorial: Recruiting older adults in our studies.
Applied Nursing Research, 14(2), 63.
Glaser, B. G., & Strauss, A. L. (1967). The discovery of grounded
theory: Strategies for qualitative research. Chicago: IL. Aldine.
Green, J., & Thorogood, N. (2004). Qualitative methods for health
research. Thousand Oaks, CA: Sage Publications.
Grove, S. K. (2007). Statistics for health care research: A practical
workbook. St. Louis, MO: Saunders.

Gul, R. B., & Ali, P. A. (2010). Clinical trials: The challenge of
recruitment and retention of participants. Journal of Clinical
Nursing, 19(1-2), 227–233.
Hellard, M.E., Sinclair, M.I., Forbes, A.B., & Fairley, C.K. (2001).
Methods used to maintain a high level of participant involvement
in a clinical trial. Journal of Epidemiology & Community Health,
55(5), 348–351.
Hendrickson, S. G. (2007). Video recruitment of non-English-
speaking participants. Western Journal of Nursing Research,
29(2), 232–242.
Hines-Martin, V., Speck, B. J., Stetson, B., & Looney, S. W. (2009).
Understanding systems and rhythms for minority recruitment in
intervention research. Research in Nursing & Health, 32(6),
657–670.
Jones, D., Duffy, M. E., & Flanagan, J. (2011). Randomized clinical
trial testing efficacy of a nurse-coached intervention in arthros-
copy patients. Nursing Research, 60(2), 92–99.
Kerlinger, F. N., & Lee, H. B. (2000). Foundations of behavioral
research (4th ed.). Fort Worth, TX: Harcourt College Publishers.
Kraemer, H. C., & Thiemann, S. (1987). How many subjects? Sta-
tistical power analysis in research. Newbury Park, CA: Sage.
Larson, E. (1994). Exclusion of certain groups from clinical research.
Image: Journal of Nursing Scholarship, 26(3), 185–190.
Leidy, N. K., & Weissfeld, L. A. (1991). Sample sizes and power
computation for clinical intervention trials. Western Journal of
Nursing Research, 13(1), 138–144.
Levy, P. S., & Lemsbow, S. (1980). Sampling for health profession-
als. Belmont, CA: Lifetime Learning.
Li, Y., & Mukamel, D. B. (2010). Racial disparities in receipt of
influenza and pneumococcus vaccinations among US nursing-
home residents. American Journal of Public Health, 100(S1),
S256–S262.
Madsen, S. M., Mirza, M. R., Holm, S., Hilsted, K. L., Kampmann,
K., & Riis, P. (2002). Attitudes towards clinical research amongst
participants and nonparticipants. Journal of Internal Medicine,
251(2), 156–168.
Marshall, C., & Rossman, G. B. (2011). Designing qualitative
research (5th ed.). Thousand Oaks, CA: Sage Publications.
McGregor, L., Parker, K., LeBlanc, P., & King, K. M. (2010). Using
social exchange theory to guide successful study recruitment and
retention. Nurse Researcher, 17(2), 74–82.
Melnyk, B. M., & Fineout-Overholt, E. (2011). Evidence-based
oasis-ebl|Rsalles|1476233281
practice in nursing & healthcare: A guide to best practice (2nd
ed.). Philadelphia: PA: Lippincott Williams & Wilkins.
Morse, J. M. (2000). Determining sample size. Qualitative Health
Research, 10(1), 3–5.
Munhall, P. L. (2007). Nursing research: A qualitative perspective
(4th ed.). Sudbury, MA: Jones & Bartlett.
Munhall, P. L. (2012). Nursing research: A qualitative perspective
(5th ed.). Sudbury, MA: Jones & Bartlett.
Neumark, D. E., Stommel, M., Given, C. W., & Given, B. A. (2001).
Brief report: Research design and subject characteristics predict-
ing nonparticipation in panel survey of older families with cancer.
Nursing Research, 50(6), 363–368.
CHAPTER 15  Sampling 381
O’Shea, E. R., Wallace, M., Griffin, M. Q., & Fitzpatrick, J. J.
(2011). The effect of an educational session on pediatric nurses’
perspectives toward providing spiritual care. Journal of Pediatric
Nursing, 26(1), 34–43.
Papadopoulos, I., & Lees, S. (2002). Developing culturally compe-
tent researchers. Journal of Advanced Nursing, 37(3), 258–264.
Patton, M. Q. (2002). Qualitative evaluation and research methods
(3rd ed.). Thousand Oaks, CA: Sage.
Pieper, B., Templin, T. N., Kirsner, R. S., & Birk, T. J. (2009).
Impact of injection drug use on distribution and severity of
chronic venous disorders. Wound Repair & Regeneration, 17(4),
485–491.
Pieper, B., Templin, T. N., Kirsner, R. S., & Birk, T. J. (2010).
The impact of vascular leg disorders on physical activity in
methadone-maintained adults. Research in Nursing & Health,
33(5), 426–440.
Sandelowski, M. (1995). Sample size in qualitative research.
Research in Nursing & Health, 18(2), 179–183.
Sandelowski, M. (2000). Combining qualitative and quantitative
sampling, data collection, and analysis techniques in mixed-
method studies. Research in Nursing & Health, 23(3),
246–255.
Sharma, N. K., Ryals, J. M., Gajewski, B. J., & Wrights, D. E.
(2010). Aerobic exercise alters analgesia and neurotrophin-3 syn-
thesis in an animal model of chronic widespread pain. Physical
Therapy, 90(5), 714–725.
Strauss, A., & Corbin, J. (1998). Basics of qualitative research,
techniques, and procedures for developing grounded theory (2nd
ed.). Thousand Oaks, CA: Sage.
Sternberg, R. M., & Barry, C. (2011). Transnational mothers cross-
ing the border and bringing their health care needs. Journal of
Nursing Scholarship, 43(1), 64–71.
Sullivan-Bolyai, S., Bova, C., Deatrick, J. A., Knafl, K., Grey, M.,
Leung, K., et al. (2007). Barriers and strategies for recruiting
study participants in clinical settings. Western Journal of Nursing
Research, 29(4), 486–500.
Thompson, S. K. (2002). Sampling (2nd ed.). New York, NY: John
Wiley & Sons.
Twiss, J. J., Waltman, N. L., Berg, K., Ott, C. D., Gross, G. J., &
Lindsey, A. M. (2009). An exercise intervention for breast cancer
survivors with bone loss. Journal of Nursing Scholarship, 41(1),
20–27.
Ulrich, C. M., Danis, M., Ratcliffe, S. J., Garrett-Mayer, E., Koziol,
D., Soeken, K. L., et al. (2006). Ethical conflict in nurse practi-
tioners and physician assistants in managed care. Nursing
Research, 55(6), 391–401.
Whitebird, R. R., Bliss, D. Z., Savik, K., Lowry, A., & Jung, H. G.
(2010). Comparing community and specialty provider-based
recruitment in a randomized clinical trial: Clinical trial in fecal
incontinence. Research in Nursing & Health, 33(6), 500–511.
Yarandi, H. N. (1991). Planning sample sizes: Comparison of factor
level means. Nursing Research, 40(1), 57–58.
Yates, F. (1981). Sampling methods for censuses and surveys. New
York, NY: Macmillan.

16
CHAPTER
Measurement Concepts
M
easurement is the process of assigning
numbers to objects, events, or situations in
accord with some rule (Kaplan, 1963). The
numbers assigned can indicate numerical values or
categories for the objects being measured for research
or practice. Instrumentation, a component of mea-
surement, is the application of specific rules to develop
a measurement device such as a scale or questionnaire.
Quality instruments are essential for obtaining trust-
worthy data when measuring outcomes for research
and practice (Doran, 2011; Melnyk & Fineout-
Overhold, 2011; Waltz, Strickland, & Lenz, 2010).
The rules of measurement were developed so that
the assigning of values or categories might be done
consistently from one subject (or event) to another and
eventually, if the measurement method is found to be
meaningful, from one study to another. The rules of
measurement established for research are similar to
the rules of measurement implemented in nursing
practice. For example, when nurses measure the urine
output from patients, they use an accurate measure-
ment device, observe the amount of urine in the device
or container in a consistent way, and precisely record
the urine output in the medical record. This practice
promotes accuracy and precision and reduces the
amount of error in measuring physiological variables
such as urine output.
When measuring a subjective concept such as pain
experienced by a child, researchers and nurses in prac-
tice need to use an instrument that captures the pain
the child is experiencing. A commonly used scale to
measure a child’s pain is the Wong-Baker FACES Pain
Rating Scale (Hockenberry & Wilson, 2009). By using
this valid and reliable rating scale to measure the
child’s pain, any change in the measured value can be
attributed to a change in the child’s pain rather than
measurement error. A copy of the Wong-Baker FACES
Pain Rating Scale is provided in Chapter 17. Selecting
accurate and precise physiological measurement
382
  http://evolve.elsevier.com/Grove/practice/
methods and valid and reliable scales and question-
naires is essential in measuring study variables and
outcomes in practice (Bannigan & Watson, 2009;
Bialocerkowski, Klupp, & Bragge, 2010; DeVon,
et al., 2007).
Researchers need to understand the logic within
measurement theory so that they can select and use
existing instruments or develop new quality measure-
ment methods for their studies. Measurement theory,
as with most theories, uses terms with meanings that
can be best understood within the context of the
theory. The following explanation of the logic of mea-
surement theory includes definitions of directness of
measurement, measurement error, levels of measure-
ment, and reference of measurement. The reliability
and validity of measurement methods, such as scales
and questionnaires, are detailed. The accuracy, preci-
sion, and error of physiological measures are described.
The chapter concludes with a discussion of sensitivity;
specificity; and likelihood ratios examined to deter-
mine the quality of diagnostic tests and instruments
used in healthcare research and practice.
Directness of Measurement
Measurement begins by clarifying the object, charac-
teristic, or element to be measured. Only then can one
identify or develop strategies or methods to measure
it. In some cases, identification of the measurement
object and measurement strategies can be objective,
specific, and straightforward, as when we are measur-
ing concrete factors, such as a person’s weight or waist
circumference; this is referred to as direct measure-
ment. Healthcare technology has made direct mea-
sures of objective elements—such as height, weight,
temperature, time, space, movement, heart rate, and
respiration—familiar to us. Technology is also avail-
able to measure many biological and chemical char-
acteristics, such as laboratory values, pulmonary

functions, and sleep patterns. Nurses are also experi-
enced in gathering direct measures of demographic
variables, such as age, gender, ethnicity, diagnosis,
marital status, income, and education.
However, in nursing, the characteristic we want to
measure often is an abstract idea or concept, such as
pain, stress, depression, anxiety, caring, or coping. If
the element to be measured is abstract, it is best clari-
fied through a conceptual definition (see Chapter 8).
The conceptual definition can be used to select or
develop appropriate means of measuring the concept.
The instrument or measurement strategy used in the
study must match the conceptual definition. An
abstract concept is not measured directly; instead,
indicators or attributes of the concept are used to rep-
resent the abstraction. This is referred to as indirect
measurement. For example, the complex concept of
coping might be defined by the frequency or accuracy
of identifying problems, the creativity in selecting
solutions, and the speed or effectiveness in resolving
the problem. A single measurement strategy rarely, if
ever, can completely measure all aspects of an abstract
concept. Multi-item scales have been developed to
measure abstract concepts, such as the Spielberger
State-Trait Anxiety Inventory developed to measure
individuals’ innate anxiety trait and their anxiety in a
specific situation (Spielberger, Gorsuch, & Lushene,
1970).
Measurement Error
There is no perfect measure. Error is inherent in any
measurement strategy. Measurement error is the dif-
ference between what exists in reality and what is
measured by an instrument. Measurement error exists
in both direct and indirect measures and can be random
or systematic. Direct measures, which are considered
to be highly accurate, are subject to error. For example,
the weight scale may not be accurate, laboratory
equipment may be precisely calibrated but may change
with use, or the tape measure may not be placed in the
same location or held at the same tension for each
measurement.
There is also error in indirect measures. Efforts to
measure concepts usually result in measuring only
part of the concept or measures that identify an aspect
of the concept but also contain other elements that are
not part of the concept. Figure 16-1 shows a Venn
diagram of the concept A measured by instrument A-1.
In this figure, A-1 does not measure all of concept A.
In addition, some of what A-1 measures is outside the
concept of A. Both of these situations are examples of
errors in measurement and are shaded in Figure 16-1.
CHAPTER 16  Measurement Concepts 383
A-1
A
Figure 16-1  Measurement error when measuring a concept.
Types of Measurement Errors
Two types of errors are of concern in measurement:
random error and systematic error. To understand
these types of errors, we must first understand the ele-
ments of a score on an instrument or an observation.
According to measurement theory, there are three
components to a measurement score: true score,
observed score, and error score. The true score (T) is
what we would obtain if there was no error in mea-
surement. Because there is always some measurement
error, the true score is never known. The observed
score (O) is the measure obtained for a subject using
a selected instrument during a study. The error score
(E) is the amount of random error in the measurement
process. The theoretical equation of these three mea-
sures is as follows:
Observed score = true score + random error
This equation is a means of conceptualizing random
error and not a basis for calculating it. Because the
true score is never known, the random error is never
known but only estimated. Theoretically, the smaller
the error score, the more closely the observed score
reflects the true score. Therefore, using instruments
that reduce error improves the accuracy of measure-
ment (Waltz et al., 2010).
Several factors can occur during the measurement
process that can increase random error. These factors
include (1) transient personal factors, such as fatigue,
hunger, attention span, health, mood, mental status,
and motivation; (2) situational factors, such as a hot
stuffy room, distractions, the presence of significant
384 UNIT TWO  The Research Process

others, rapport with the researcher, and the playfulness +2

or seriousness of the situation; (3) variations in the –2
administration of the measurement procedure, such as
0
interviews in which wording or sequence of questions
is varied, questions are added or deleted, or research- T1 02 01 T2 T3
ers code responses differently; and (4) processing of
data, such as errors in coding, accidentally marking Figure 16-2  Conceptualization of random error.
the wrong column, punching the wrong key when
entering data into the computer, or incorrectly totaling
instrument scores (Devon et al., 2007; Waltz et al.,
2010).
Random error causes individuals’ observed scores
to vary in no particular direction around their true A-1
score. For example, with random error, one subject’s
observed score may be higher than his or her true
score, whereas another subject’s observed score may
be lower than his or her true score. According to mea-
surement theory, the sum of random errors is expected
to be zero, and the random error score (E) is not
expected to correlate with the true score (T). Random
error does not influence the mean to be higher or lower
but rather increases the amount of unexplained vari-
ance around the mean. When this occurs, estimation A
of the true score is less precise.
If you were to measure a variable for three sub-
jects and diagram the random error, it might appear
as shown in Figure 16-2. The difference between the
true score of subject 1 (T1) and the observed score Figure 16-3  Conceptualization of systematic error.
(O1) is two positive measurement intervals. The dif-
ference between the true score (T2) and observed
score (O2) for subject 2 is two negative measurement
intervals. The difference between the true score (T3) (which is 0) yields the observed score, as shown by
and observed score (O3) for subject 3 is zero. The the following equations:
random error for these three subjects is zero (+2 − 2
T ( true score with systematic error ) +
+ 0 = 0). In viewing this example, one must remem-
E ( random error of 0) = O (observed score)
ber that this is only a means of conceptualizing
random error. or
Measurement error that is not random is referred
T+E =O
to as systematic error. A scale that weighs subjects 3
pounds more than their true weights is an example of Some systematic error is incurred in almost any
systematic error. All of the body weights would be measure; however, a close link between the abstract
higher, and, as a result, the mean would be higher theoretical concept and the development of the instru-
than it should be. Systematic error occurs because ment can greatly decrease systematic error. Because
something else is being measured in addition to the of the importance of this factor in a study, researchers
concept. A conceptualization of systematic error is spend considerable time and effort in selecting and
presented in Figure 16-3. Systematic error (repre- developing quality measurement methods to decrease
sented by the shaded area in the figure) is due to the systematic error.
part of A-1 that is outside of A. This part of A-1 mea- Another effective means of diminishing systematic
sures factors other than A and biases scores in a par- error is to use more than one measure of an attribute
ticular direction. or a concept and to compare the measures. To make
Systematic error is considered part of T (true score) this comparison, researchers use various data collec-
and reflects the true measure of A-1, not A. Adding the tion methods, such as scale, interview, and observa-
true score (with systematic error) to the random error tion. Campbell and Fiske (1959) developed a technique

A-2
A-1
A
A-4
Figure 16-4  Multiple measures of an abstract concept.
of using more than one method to measure a concept,
referred to as the multimethod-multitrait technique.
More recently, the technique has been described as a
version of mixed methodology, as discussed in
Chapter 10. These techniques allow researchers to
measure more dimensions of abstract concepts, and
the effect of the systematic error on the composite
observed score decreases. Figure 16-4 illustrates how
more dimensions of concept A are measured through
the use of four instruments, designated A-1, A-2, A-3,
and A-4.
For example, a researcher could decrease system-
atic error in measures of anxiety by (1) administering
oasis-ebl|Rsalles|1476233285
the Spielberger State-Trait Anxiety Inventory, (2)
recording blood pressure readings, (3) asking the
subject about anxious feelings, and (4) observing the
subject’s behavior. Multimethod measurement strate-
gies decrease systematic error by combining the values
in some way to give a single observed score of anxiety
for each subject. However, sometimes it may be dif-
ficult logically to justify combining scores from
various measures, and a mixed-methods approach
might be the most appropriate to use in the study.
Mixed-methods study uses a combination of quantita-
tive and qualitative approaches in their implementa-
tion (Creswell, 2009).
In some studies, researchers use instruments to
examine relationships. Consider a hypothesis that tests
CHAPTER 16  Measurement Concepts 385
Concept A Concept B
A-3
Area of relationship
Figure 16-5  True relationship of concepts A and B.
the relationship between concept A and concept B. In
Figure 16-5, the shaded area enclosed in the dark lines
represents the true relationship between concepts A
and B, such as the relationship between anxiety and
depression. For example, two instruments, A-1 (Spiel-
berger State Anxiety Scale) and B-1 (Center for Epi-
demiological Studies Depression Scale, Radloff,
1977), are used to examine the relationship between
concepts A and B. The part of the true relationship
actually reflected by A-1 and B-1 measurement
methods is represented by the colored area in Figure
16-6. Because two instruments provide a more accu-
rate measure of concepts A and B, more of the true
relationship between concepts A and B can be
measured.
If additional instruments (A-2 and B-2) are used to
measure concepts A and B, more of the true relation-
ship might be reflected. Figure 16-7 demonstrates with
different colors the parts of the true relationship
between concepts A and B that is measured when
concept A is measured with two instruments (A-1 and
A-2) and concept B is measured with two instruments
(B-1 and B-2).
Levels of Measurement
The traditional levels of measurement have been used
for so long that the categorization system has been
considered absolute and inviolate. In 1946, Stevens
organized the rules for assigning numbers to objects
so that a hierarchy in measurement was established
called the levels of measurement. The levels of mea-
surement, from lower to higher, are nominal, ordinal,
interval, and ratio.
386 UNIT TWO  The Research Process

Absolute
A-1 Zero
B-1
Equal Equal
Interval Interval
Categories Categories
Ranked Ranked Ranked
Categories Categories Categories
Exhaustive Exhaustive Exhaustive Exhaustive
Categories Categories Categories Categories
A
Exclusive Exclusive Exclusive Exclusive
B
Categories Categories Categories Categories
Nominal Ordinal Interval Ratio

Figure 16-8  Summary of the rules for levels of measurement.

Figure 16-6  Examining a relationship using one measure of each

concept. Nominal Level of Measurement
Nominal level of measurement is the lowest of the
four measurement levels or categories. It is used when
data can be organized into categories of a defined
property but the categories cannot be ordered. For
example, diagnoses of chronic diseases are nominal
A-1 data with categories such as hypertension, type 2 dia-
B-1 betes, and dyslipidemia. One cannot say that one cat-
egory is higher than another or that category A
(hypertension) is closer to category B (diabetes) than
to category C (dyslipidemia). The categories differ in
quality but not quantity. One cannot say that subject
A possesses more of the property being categorized
than does subject B. (Rule: The categories must be
unorderable.) Categories must be established so that
A each datum fits into only one of the categories. (Rule:
B The categories must be exclusive.) All the data must
fit into the established categories. (Rule: The catego-
ries must be exhaustive.)
Figure 16-8 provides a summary for the rules for
the four levels of measurement—nominal, ordinal,
interval, and ratio. Data such as ethnicity, gender,
marital status, religion, and diagnoses are examples of
A-2
nominal data. When data are coded for entry into the
B-2 computer, the categories are assigned numbers. For
example, gender may be classified as 1 = male and
2 = female. The numbers assigned to categories in
nominal measurement are used only as labels and
Figure 16-7  Examining a relationship using two measures of each cannot be used for mathematical calculations.
concept.
Ordinal Level of Measurement
Data that can be measured at the ordinal level can be
assigned to categories of an attribute that can be
ranked. There are rules for how one ranks data. As

with nominal-scale data, the categories must be exclu-
sive and exhaustive. With ordinal level data, the quan-
tity of the attribute possessed can be identified.
However, it cannot be shown that the intervals between
the ranked categories are equal (see Figure 16-8).
Ordinal data are considered to have unequal intervals.
Scales with unequal intervals are sometimes referred
to as ordered metric scales.
Many scales used in nursing research are ordinal
levels of measure. For example, one could rank inten-
sity of pain, degrees of coping, levels of mobility,
ability to provide self-care, or daily amount of exer-
cise on an ordinal scale. For daily exercise, the scale
could be 0 = no exercise; 1 = moderate exercise, no
sweating; 2 = exercise to the point of sweating; 3 =
strenuous exercise with sweating for at least 30
minutes per day; 4 = strenuous exercise with sweating
for at least 1 hour per day. This type of scale may be
referred to as a metric ordinal scale.
Interval Level of Measurement
In interval level of measurement, distances between
intervals of the scale are numerically equal. Such mea-
surements also follow the previously mentioned rules:
mutually exclusive categories, exhaustive categories,
and rank ordering. Interval scales are assumed to be a
continuum of values (see Figure 16-8). The researcher
can identify the magnitude of the attribute much more
precisely. However, it is impossible to provide the
absolute amount of the attribute because of the absence
of a zero point on the interval scale.
Fahrenheit and Celsius temperatures are commonly
used as examples of interval scales. A difference
between a temperature of 70° F and one of 80° F is
the same as the difference between a temperature of
30° F and one of 40° F. We can measure changes in
temperature precisely. However, it is impossible to say
that a temperature of 0° C or 0° F means the absence
oasis-ebl|Rsalles|1476233289
of temperature because these indicate very cold
temperatures.
Ratio Level of Measurement
Ratio level of measurement is the highest form of
measure and meets all the rules of the lower forms
of measures: mutually exclusive categories, exhaus-
tive categories, rank ordering, equal spacing between
intervals, and a continuum of values. In addition, ratio
level measures have absolute zero points (see Figure
16-8). Weight, length, and volume are common exam-
ples of ratio scales. Each has an absolute zero point,
at which a value of zero indicates the absence of
the property being measured: Zero weight means
the absence of weight. In addition, because of the
CHAPTER 16  Measurement Concepts 387
absolute zero point, one can justifiably say that object
A weighs twice as much as object B, or that container
A holds three times as much as container B. Labora-
tory values are also an example of ratio level of mea-
surement where the individual with a fasting blood
sugar (FBS) of 180 has an FBS twice that of an indi-
vidual with a normal FBS of 90. To help expand
understanding of levels of measurement (nominal,
ordinal, interval, and ratio) and to apply this knowl-
edge, Grove (2007) developed a statistical workbook
focused on examining the levels of measurement,
sampling methods, and statistical results in published
studies.
Importance of Level of Measurement for
Statistical Analyses
An important rule of measurement is that one should
use the highest level of measurement possible. For
example, you can collect data on age (measured) in a
variety of ways: (1) you can obtain the actual age of
each subject (ratio level of measurement); (2) you can
ask subjects to indicate their age by selecting from a
group of categories, such as 20 to 29, 30 to 39, and
so on (ordinal level of measurement); or (3) you can
sort subjects into two categories of younger than 65
years of age and 65 years of age and older (nominal
level of measurement). The highest level of measure-
ment in this case is the actual age of each subject,
which is the preferred way to collect these data. If you
need age categories for specific analyses in your
research, the computer can be instructed to create
age categories from the initial age data (Waltz et al.,
2010).
The level of measurement is associated with the
types of statistical analyses that can be performed on
the data. Mathematical operations are limited in the
lower levels of measurement. With nominal levels of
measurement, only summary statistics, such as fre-
quencies, percentages, and contingency correlation
procedures, can be used. However, if a variable such
as age is measured at the ratio level (actual age of the
subject), the data can be analyzed with more sophisti-
cated analysis techniques. Variables measured at the
interval or ratio level can be analyzed with the stron-
gest statistical techniques available, which are more
effective in identifying relationships among variables
or determining differences between groups (Corty,
2007; Grove, 2007).
Controversy over Measurement Levels
There is controversy over the system that is used to
categorize measurement levels, dividing researchers
into two factions: fundamentalists and pragmatists.
388 UNIT TWO  The Research Process
Pragmatists regard measurement as occurring on a
continuum rather than by discrete categories, whereas
fundamentalists adhere rigidly to the original system
of categorization (Nunnally & Bernstein, 1994;
Stevens, 1946).
The primary focus of the controversy relates to the
practice of classifying data into the categories ordinal
and interval. This controversy developed because,
according to the fundamentalists, many of the current
statistical analysis techniques can be used only with
interval and ratio data. Many pragmatists believe that
if researchers rigidly adhered to rules developed by
Stevens (1946), few if any measures in the social sci-
ences would meet the criteria to be considered interval-
level data. They also believe that violating Stevens’
criteria does not lead to serious consequences for the
outcomes of data analysis. Pragmatists often treat
ordinal data from multi-item scales as interval data,
using statistical methods (parametric analysis tech-
niques) to analyze them, such as Pearson’s product-
moment correlation coefficient, t-test, and analysis of
variance (ANOVA), which are traditionally reserved
for interval or ratio level data (Armstrong, 1981;
Knapp, 1990). Fundamentalists insist that the analysis
of ordinal data be limited to statistical procedures
designed for ordinal data, such as nonparametric pro-
cedures. Parametric statistical analysis techniques
were developed to analyze interval and ratio level
data, and nonparametric techniques were developed to
analyze nominal and ordinal data (see Chapter 21).
The Likert scale uses scale points such as “strongly
disagree,” “disagree,” “uncertain,” “agree,” and
“strongly agree.” Numerical values (e.g., 1, 2, 3, 4,
and 5) are assigned to these categories. Fundamental-
ists claim that equal intervals do not exist between
these categories. It is impossible to prove that there is
the same magnitude of feeling between “uncertain”
and “agree” as there is between “agree” and “strongly
agree.” Therefore, they hold this is ordinal level data,
and parametric analyses cannot be used. Pragmatists
believe that with many measures taken at the ordinal
level, such as scaling procedures, an underlying inter-
val continuum is present that justifies the use of para-
metric statistics (Knapp, 1990; Nunnally & Bernstein,
1994).
Our position agrees more with the pragmatists
than with the fundamentalists. Many nurse research-
ers analyze data from Likert scales and other rating
scales as though the data were interval level (Waltz
et al., 2010). However, some of the data in nursing
research are obtained through the use of crude mea-
surement methods that can be classified only into the
lower levels of measurement (ordinal or nominal).
Therefore, we have included the nonparametric statis-
tical procedures needed for their analysis in Chapters
22 to 25 on statistics.
Reference Testing
of Measurement
Referencing involves comparing a subject’s score
against a standard. Two types of testing involve
referencing: norm-referenced testing and criterion-
referenced testing. Norm-referenced testing ad-
dresses the question, “How does the average person
score on this test or instrument?” This testing in-
volves standardization of scores for an instrument
that is accomplished by data collection over several
years, with extensive reliability and validity informa-
tion available on the instrument. Standardization
involves collecting data from thousands of subjects
expected to have a broad range of scores on the
instrument. From these scores, population parameters
such as the mean and standard deviation (described
in Chapter 22) can be developed. Evidence of the
reliability and validity of the instrument can also be
evaluated through the use of the methods described
later in this chapter. The best-known norm-referenced
test is the Minnesota Multiphasic Personality Inven-
tory (MMPI), which is used commonly in psychology
and occasionally in nursing research and practice to
diagnosis personality disorders. The Graduate Record
Examination (GRE) is another norm-referenced test
commonly used as one of the admission criteria for
graduate study.
Criterion-referenced testing asks the question,
“What is desirable in the perfect subject?” It involves
comparing a subject’s score with a criterion of
achievement that includes the definition of target
behaviors. When the subject has mastered these
behaviors, he or she is considered proficient in
the behavior (DeVon et al., 2007; Sax, 1997). The
criterion might be a level of knowledge or desirable
patient outcomes. Criterion measures have been used
for years to evaluate outcomes in healthcare agencies
and to determine clinical expertise of students. For
example, a clinical evaluation form would include the
critical behaviors the nurse practitioner (NP) student
is expected to master in a pediatric course to be clini-
cally competent to care for pediatric patients at the
end of the course. Criterion-reference testing is also
used in nursing research. Criterion-referenced testing
might be used to measure the clinical expertise of a
nurse or the self-care of a cardiac patient after cardiac
rehabilitation.

Reliability
The reliability of an instrument denotes the consis-
tency of the measures obtained of an attribute, item,
or situation in a study or clinical practice. The greater
the reliability or consistency of the measures of a
particular instrument, the less random error in the
measurement method (Bannigan & Watson, 2009;
Bialocerkowski et al., 2010; DeVon et al., 2007). If
the same measurement scale is administered to the
same individuals at two different occasions, the mea-
surement is reliable if the individuals’ responses to the
items remain the same (assuming that nothing has
occurred to change their responses). For example, if
you use a scale to measure the anxiety levels of 10
individuals at two points in time 30 minutes apart, you
would expect the individuals’ anxiety levels to be rela-
tively unchanged from one measurement to the next
if the scale is reliable. If two data collectors observe
the same event and record their observations on a
carefully designed data collection instrument, the
measurement would be reliable if the recordings from
the two data collectors are comparable. The equiva-
lence of their results would indicate the reliability of
the measurement technique. If responses vary each
time a measure is performed, there is a chance that the
instrument is unreliable, meaning that it yields data
with a large random error.
Reliability plays an important role in the selection
of measurement methods for use in a study. Research-
ers need instruments that are reliable and provide
values with only a small amount of error. Reliable
instruments enhance the power of a study to detect
significant differences or relationships actually occur-
ring in the population under study. It is important to
examine the reliability of an instrument from previous
research before using it in a study. Estimates of instru-
ment reliability are specific to the population and
sample being studied. High reported reliability values
on an established instrument do not guarantee that its
reliability would be satisfactory in another sample
from a different population (Waltz et al., 2010).
Researchers need to perform reliability testing on each
instrument used in their study before performing other
statistical analyses. The reliability values must be
included in the published report of a study to docu-
ment that the instruments used were reliable for the
study sample (Bialocerkowski et al., 2010; DeVon
et al., 2007).
Reliability testing examines the amount of mea-
surement error in the instrument being used in a study.
Reliability is concerned with the dependability, con-
sistency, stability, precision, reproducibility, and
CHAPTER 16  Measurement Concepts 389
comparability of a measurement method (Bartlett &
Frost, 2008). The strongest measure of reliability is
obtained from heterogeneous samples versus homoge-
neous samples. Heterogeneous samples have more
between-participant variability, and this is a stronger
evaluation of reliability than homogeneous samples
with little between-participant variation. When criti-
cally appraising the reliability of an instrument in a
study, you need to examine the sample for heterogene-
ity by determining the variability of the scores among
study participants (Bartlett & Frost, 2008; Bialocer-
kowski et al., 2010).
All measurement techniques contain some random
error, and the errors might be due to the measurement
method used, the study participants, or the researchers
gathering the data. Reliability exists in degrees and is
usually expressed as a form of correlation coefficient,
with 1.00 indicating perfect reliability and 0.00 indi-
cating no reliability (Bialocerkowski et al., 2010). For
example, reliability coefficients of 0.80 or higher are
considered strong values for an established psychoso-
cial scale such as the State-Trait Anxiety Inventory by
Spielberger et al. (1970). With test-retest, the closer
that reliability coefficient is to 1.00, the more stable
the measurement method. The reliability coefficient
varies based on the type of reliability being examined.
The three most common types of reliability discussed
in healthcare studies are (1) stability reliability, (2)
equivalence reliability, and (3) internal consistency
(Bannigan & Watson, 2009; Bialocerkowski et al.,
2010; DeVon et al., 2007; Waltz et al., 2010).
Stability Reliability
Stability reliability is concerned with the consistency
of repeated measures of the same attribute with the
use of the same scale or instrument over time. It is
usually referred to as test-retest reliability. This
measure of reliability is generally used with physical
measures, technological measures, and paper-and-
pencil scales. The technique requires an assumption
that the factor to be measured remains the same at the
two testing times and that any change in the value or
score is a consequence of random error.
The optimal time period between test-retest mea-
surements depends on the variability of the variable
being measured, complexity of the measurement
process, and characteristics of the participants (Bialo-
cerkowski et al., 2010). Physical measures and equip-
ment can be tested and then immediately retested, or
the equipment can be used for a time and then retested
to determine the necessary frequency of recalibration.
For example, in measuring blood pressure (BP),
researchers often take two to three BP readings 5
390 UNIT TWO  The Research Process
minutes apart and average the readings to obtain a
reliable or precise measure of BP. Researchers can
follow the standards for recalibration of equipment or
be more conservative. The standard requirements
might be to recalibrate the BP equipment every 6
months, but researchers might choose to recalibrate
the equipment every month or even every week if
multiple BP readings are being taken each day for a
study. The test-retest of a measurement method might
have a longer period of time if the variable being
measured changes slowly. For example, the diagnosis
of osteoporosis is made by bone mineral density
(BMD) study of the hip, spine, and wrist. The BMD
score is determined with a dual-energy x-ray absorp-
tiometry (DEXA) scan. Because the BMD does not
change rapidly in people even with treatment, test-
retest over a 1- to 2-month time period could be used
to show reliable or consistent DEXA scan scores for
patients.
With paper-and-pencil educational tests, a period
of 2 to 4 weeks is recommended between the two
testing times, but the time period for retesting does
depend on what is measured and the instrument used
(Sax, 1997). After the same participants have been
retested with the same instrument, the investigators
perform a correlational analysis on the scores from the
two measurement times. This correlation is called the
coefficient of stability, and the closer the coefficient
is to 1.00, the more stable the instrument (Waltz et al.,
2010). For some scales, test-retest reliability has not
been as effective as originally anticipated. The proce-
dure presents numerous problems. Subjects may
remember their responses from the first testing time,
leading to overestimation of the reliability. Subjects
may be changed by the first testing and may respond
to the second test differently, leading to underestima-
tion of the reliability (Bialocerkowski et al., 2010).
Test-retest reliability requires the assumption that
the factor being measured has not changed between
the measurement points. Many of the phenomena
studied in nursing, such as hope, coping, pain, and
anxiety, do change over short intervals. Thus, the
assumption that if the instrument is reliable, values
will not change between the two measurement periods
may not be justifiable. If the factor being measured
does change, the test is not a measure of reliability. If
the measures stay the same even though the factor
being measured has changed, the instrument may lack
reliability. If researchers are going to examine the reli-
ability of an instrument with test-retest, they need to
determine the optimum time between administrations
of the instrument based on the variable being mea-
sured and the study participants (Devon et al., 2007).
Stability of a measurement method is important and
needs to be examined as part of instrument develop-
ment and discussed when the instrument is used in a
study. When describing test-retest results, researchers
need to discuss the process and the time period
between administering an instrument and the rationale
for this time frame. If a scale was administered twice
30 minutes apart or there was 1 month between test
and retesting, the consistency of the subjects’ scores
need to be discussed in terms of the timing for retest-
ing (Bannigan & Watson, 2009; Bialocerkowski et al.,
2010; DeVon et al., 2007).
Equivalence Reliability
Equivalence reliability compares two versions of the
same paper-and-pencil instrument or two observers
measuring the same event. Comparison of two observ-
ers is referred to as interrater reliability. Comparison
of two paper-and-pencil instruments is referred to as
alternate-forms reliability or parallel-forms reli-
ability. Alternative forms of instruments are of more
concern in the development of normative knowledge
testing. However, when repeated measures are part of
the design, alternative forms of measurement, although
not commonly used, would improve the design. Dem-
onstrating that one is actually testing the same content
in both tests is extremely complex, and the procedure
is rarely used in clinical research (Bialocerkowski
et al., 2010).
The procedure for developing parallel forms
involves using the same objectives and procedures to
develop two like instruments. These two instruments
when completed by the same group of study partici-
pants on the same occasion or two different occasions
should have approximately equal means and standard
deviations. In addition, these two instruments should
correlate equally with another variable. For example,
if two instruments were developed to measure pain,
the scores from these two scales should correlate
equally with perceived anxiety score. If both forms of
the instrument are administered during the same occa-
sion, a reliability coefficient can be calculated to deter-
mine equivalence. A coefficient of 0.80 or higher
indicates equivalence (Waltz et al., 2010).
Determining interrater reliability is a concern
when studies include observational measurement,
which is common in qualitative research. Interrater
reliability values need to be reported in any study in
which observational data are collected or judgments
are made by two or more data gatherers. Two tech-
niques determine interrater reliability. Both tech-
niques require that two or more raters independently
observe and record the same event using the protocol

developed for the study or that the same rater ob-
serves and records an event on two occasions. To
judge interrater reliability adequately, the raters need
to observe at least 10 subjects or events (DeVon et al.,
2007; Waltz et al., 2010). A digital recorder can be
used to record the raters to determine their consis-
tency in recording essential study information. Every
data collector used in the study must be tested for
interrater reliability and trained to a consistency in
data collection.
One procedure for calculating interrater reliability
requires a simple computation involving a comparison
of the agreements obtained between raters on the
coding form with the number of possible agreements.
This calculation is performed through the use of the
following equation:
Number of agreements ÷ number of possible
agreements = interrater reliability
This formula tends to overestimate reliability, a
particularly serious problem if the rating requires only
a dichotomous judgment, such as present or absent. In
this case, there is a 50% probability that the raters will
agree on a particular item through chance alone. If
more than two raters are involved, a statistical proce-
dure to calculate coefficient alpha (discussed later in
this chapter) may be used. ANOVA may also be used
to test for differences among raters. There is no abso-
lute value below which interrater reliability is unac-
ceptable. However, any value less than 0.80 (80%)
should generate serious concern about the reliability
of the data because there is 20% chance of error. The
more ideal interrater reliability value is 0.90, which
means 90% reliability and 10% error. The process for
determining interrater reliability and the value
achieved need to be included in the research report
(DeVon et al., 2007).
When raters know they are being watched, their
accuracy and consistency are considerably better than
when they believe they are not being watched. Inter-
rater reliability declines (sometimes dramatically)
when the raters are assessed covertly (Topf, 1988).
You can develop strategies to monitor and reduce the
decline in interrater reliability, but they may entail
considerable time and expense.
Internal Consistency
Tests of instrument internal consistency or homoge-
neity, used primarily with paper-and-pencil tests or
scales, address the correlation of various items within
the instrument. The original approach to determining
internal consistency was split-half reliability. This
strategy was a way of obtaining test-retest reliability
CHAPTER 16  Measurement Concepts 391
without administering the test twice. The instrument
items were split in odd-even or first-last halves, and a
correlational procedure was performed between the
two halves. In the past, researchers generally reported
the Spearman-Brown correlation coefficient in their
studies (Nunnally & Bernstein, 1994; Sax, 1997). One
of the problems with the procedure was that although
items were usually split into odd-even items, it was
possible to split them in a variety of ways. Each
approach to splitting the items would yield a different
reliability coefficient. The researcher could continue
to split the items in various ways until a satisfactorily
high coefficient was obtained.
More recently, testing the internal consistency of
all the items in the instrument has been seen as a better
approach to determining reliability. Although the
mathematics of the procedure are complex, the logic
is simple. One way to view it is as though one con-
ducted split-half reliabilities in all the ways possible
and then averaged the scores to obtain one reliability
score. Internal consistency testing examines the extent
to which all the items in the instrument consistently
measure a concept. Cronbach’s alpha coefficient is
the statistical procedure used for calculating internal
consistency for interval and ratio level data. This reli-
ability coefficient is essentially the mean of the inter-
item correlations and can be calculated using most
data analysis programs such as the Statistical Program
for the Social Sciences (SPSS). If the data are dichoto-
mous, such as a symptom list that has responses of
present or absent, the Kuder-Richardson formulas (KR
20 or KR 21) can be used to calculate the internal
consistency of the instrument (DeVon et al., 2007).
The KR 21 assumes that all the items on a scale or test
are equally difficult; the KR 20 is not based on this
assumption. Waltz et al. (2010) provided the formulas
for calculating both KR 20 and KR 21.
Cronbach’s alpha coefficients can range from 0.00,
indicating no internal consistency or reliability, to
1.00, indicating perfect internal reliability with no
measurement error. Alpha coefficients of 1.00 are not
obtained in study results because all instruments have
some measurement error. However, many respected
psychosocial scales used for 15 to 30 years to measure
study variables in a variety of populations have strong
0.8 or greater internal reliability coefficients. The
coefficient of 0.80 (or 80%) indicates the instrument
is 80% reliable with 20% random error (DeVon et al.,
2007; Fawcett & Garity, 2009; Grove, 2007). Scales
with 20 or more items usually have stronger internal
consistency coefficients than scales with 10 to 15
items or less. Often scales that measure complex con-
structs such as quality of life (QOL) have subscales
392 UNIT TWO  The Research Process
that measure different aspects of QOL, such as health,
physical functioning, and spirituality. Some of these
complex scales with distinct subscales, such as the
QOL scale, might have lower Cronbach’s alpha coef-
ficients since the scale is measuring different aspects
of QOL. The subscales have fewer items than the total
scale and usually lower Cronbach’s alpha coefficients
but do need to show internal consistency in measuring
a concept (Bialocerkowski et al., 2010; Waltz et al.,
2010).
Newer instruments, developed in the last 5 years,
might show moderate internal reliability (0.70 to 0.79)
when used in measuring variables in a variety of
samples. The subscales of these new instruments
might have internal reliability from 0.60 to 0.69. The
authors of these scales might continue to refine them
based on additional reliability and validity information
to improve the reliability of the total scale and sub-
scales. Reliability coefficients less than 0.60 are con-
sidered low and indicate limited instrument reliability
or consistency in measurement with high random
error. Higher levels of reliability or precision (0.90 to
0.99) are important for physiological measures that are
used to determine critical physiological functions such
as arterial pressure and oxygen saturation (Bialocer-
kowski et al., 2010; DeVon et al., 2007).
The quality of the instrument reliability needs to
be examined in terms of the type of study, measure-
ment method, and population (DeVon et al., 2007;
Kerlinger & Lee, 2000). In published studies, research-
ers need to identify the reliability coefficients of an
instrument from previous research and for their par-
ticular study. Because the reliability of an instrument
can vary from one population or sample to another, it
is important that the reliability of the scale and sub-
scales be determined and reported for the sample in
each study (Bialocerkowski, et al., 2010).
Dickerson, Kennedy, Wu, Underhill, and Othman
(2010) conducted a study of QOL and anxiety levels
of patients with implantable defibrillators. They pro-
vided the following discussion of the reliability of the
scales they used in their study.
“Anxiety. Anxiety was measured by the Spielberger
State-Trait Anxiety Inventory (STAI), which deter-
mines a subject’s current state of anxiety. This instru-
ment differentiates between the temporary condition
of ‘state anxiety’ and the longstanding quality of ‘trait
anxiety.’ The STAI is a 40-item instrument that gages
emotional reactions to the environment (e.g., ‘I am
tense,’ ‘I feel upset,’ or ‘I am worried’). Subjects rate
themselves on a 4-point scale of 1 (not at all) to 4
(very much so). The median α [alpha] reliability of the
inventory was reported to be 0.93 [from previous
studies].… Cronbach’s α scores for the present study
ranged from 0.90 to 0.96 for the 3 time periods. Only
the state anxiety score was used in this analysis.
“Quality-of-life measure. Quality-of-life was mea-
sured using the Ferrans and Powers Quality of Life
Index, Cardiac Version (QLI: CV) (Bliley & Ferrans,
1993). The QLI: CV measures the subject’s percep-
tion of QOL, according to a 72-item scale consisting
of 2 parts. The first part measures satisfaction with
various aspects of life and the second part measures
the importance of these aspects to a subject. In part
1, subjects respond to a 6-point scale, ranging from
‘very important’ (6 points) to ‘very unimportant’ (1
point). Scores are calculated by weighing the satisfac-
tion responses with the importance responses. They
reflect how satisfied subjects are with the aspects of
life that are important to them. The 4 subscales of
QOL scored include health and functioning, social and
economic, psychological/spiritual, and family. The reli-
ability of the total scale’s internal consistency was
supported by α coefficients ranging from 0.90 to 0.95.
Stability and reliability were supported by a test-retest
correlation of 0.87 at a 2-week interval, and 0.81 at
a 1-month interval.… Cronbach’s α scores for the
present study ranged from 0.95 to 0.96 for total QLI:
CV, and for the subscales, Cronbach’s α scores ranged
from 0.88 to 0.94.” (Dickerson et al., 2010, p. 468)
Dickerson et al. (2010) used two very reliable
scales to measure their study variables and docu-
mented this in their article. They measured anxiety
with the Spielberger STAI, which was developed more
than 40 years ago (Spielberger et al., 1970), has shown
strong internal consistency in previous research
(median alpha = 0.93), and was reliable in this study
(Cronbach’s alpha = 0.90 to 0.96). In previous studies,
the QLI: CV had strong internal consistency for the
total scale (alpha = 0.90-0.95) and stability reliability
with test-retest over 2 weeks and 1 month. In addition
to the strong stability reliability coefficients, the
researchers also provided the time frames for the test-
retests that were run on the scale. Another strength is
that the QLI: CV showed strong internal consistency
for the total scale (alpha = 0.95 to 0.96) and the four
subscales (alpha = 0.88 to 0.94) with the population
in this study.
Other approaches to testing internal consistency are
(1) Cohen’s kappa statistic, which determines the

percentage of agreement with the probability of chance
being taken out; (2) correlating each item with the
total score for the instrument; and (3) correlating each
item with each other item in the instrument. This pro-
cedure, often used in instrument development, allows
researchers to identify items that are not highly cor-
related and delete them from the instrument. Factor
analysis may also be used to develop instrument reli-
ability. The number of factors being measured influ-
ences the reliability of the instrument, and total
instrument scores may be more reliable than the scores
of the subscales. After performing the factor analysis,
the researcher can delete instrument items with low
factor weights. After these items have been deleted,
reliability scores on the instrument are higher. For
instruments with more than one factor, correlations
can be performed between items and factor scores (see
Chapter 23 for a discussion of factor analysis).
It is essential that an instrument be both reliable
and valid for measuring a study variable in a popula-
tion. If the instrument has low reliability values, it
cannot be valid because its measurement is inconsis-
tent and has high measurement error (DeVon et al.,
2007; Waltz et al., 2010). An instrument that is reliable
cannot be assumed to be valid for a particular study
or population. You need to determine the validity of
the instrument you are using for your study, which you
can accomplish in a variety of ways.
Validity
The validity of an instrument determines the extent to
which it actually reflects or is able to measure the
construct being examined. Several types of validity
are discussed in the literature, such as content validity,
predictive validity, criterion validity, and construct
validity. Within each of these types, subtypes have
been identified. These multiple types of validity are
very confusing, especially because the types are not
discrete but are interrelated (Bannigan & Watson,
2009; DeVon et al., 2007; Fawcett & Garity, 2009).
In this text, validity is considered a single broad
measurement evaluation that is referred to as con-
struct validity and includes various types, such as
content validity, validity from factor analysis, conver-
gent and divergent validity, validity from contrasting
groups, and validity from prediction of future and
current events (DeVon et al., 2007). All of the previ-
ously identified types of validity are now considered
evidence of construct validity. In 1999, in its Stan-
dards for Educational and Psychological Testing, the
American Psychological Association’s Committee to
Develop Standards published standards used to judge
CHAPTER 16  Measurement Concepts 393
the evidence of validity. This important work greatly
extends our understanding of what validity is and how
to achieve it. According to the American Psychologi-
cal Association’s Committee to Develop Standards
(1999), validity addresses the appropriateness, mean-
ingfulness, and usefulness of the specific inferences
made from instrument scores. It is the inferences made
from the scores, not the scores themselves, that are
important to validate (Devon et al., 2007; Goodwin &
Goodwin, 1991).
Validity, similar to reliability, is not an all-or-
nothing phenomenon but rather a matter of degree. No
instrument is completely valid. One determines the
degree of validity of a measure rather than whether or
not it has validity. Determining the validity of an
instrument often requires years of work. Many authors
equate the validity of the instrument with the rigorous-
ness of the researcher. The assumption is that because
the researcher develops the instrument, the researcher
also establishes the validity. However, this is an erro-
neous assumption because validity is not a commodity
that researchers can purchase with techniques. Validity
is an ideal state—to be pursued, but not to be attained.
As the roots of the word imply, validity includes truth,
strength, and value. Some authors might believe that
validity is a tangible “resource,” which can be acquired
by applying enough appropriate techniques. However,
we reject this view and believe measurement validity
is similar to integrity, character, or quality, to be
assessed relative to purposes and circumstances and
built over time by researchers conducting a variety of
studies (Brinberg & McGrath, 1985).
Figure 16-9 illustrates validity (the shaded area) by
the extent to which the instrument A-1 reflects concept
A-1 A
Systemic
error
Validity
Figure 16-9  Representation of instrument validity.
394 UNIT TWO  The Research Process
A. As measurement of the concept improves, validity
improves. The extent to which the instrument A-1
measures items other than the concept is referred to as
systematic error (identified as the unshaded area of
A-1 in Figure 16-9). As systematic error decreases,
validity increases.
Validity varies from one sample to another and
from one situation to another; therefore, validity
testing affirms the appropriateness of an instrument
for a specific group or purpose rather than the instru-
ment itself (DeVon et al., 2007; Waltz et al., 2010).
An instrument may be valid in one situation but not
valid in another. Instruments used in nursing studies
that were developed for use in other disciplines need
to be examined for validity in terms of nursing knowl-
edge. An instrument developed to measure cognitive
function in educational studies might not capture the
cognitive function level of elderly adults measured in
a nursing study. Researchers are encouraged to reex-
amine validity in each of their study situations. The
types of validity covered in this section include face
and content validity, readability of an instrument,
validity from factor analysis, validity from structural
analysis, validity from contrasting (or known) groups,
convergent and divergent validity, validity from dis-
criminant analysis, validity from prediction of future
and concurrent events, and successive verification
validity.
Face and Content Validity
In the 1960s and 1970s, the only type of validity that
most studies addressed was referred to as face valid-
ity, which verified basically that the instrument looked
like it was valid or gave the appearance of measuring
the construct it was supposed to measure. Face valid-
ity is a subjective assessment that might be made by
the researchers or potential subjects. Because this is a
subjective judgment with no clear guidelines for
making the judgment, this is considered the weakest
form of validity (DeVon et al., 2007). However, it is
still an important aspect of the usefulness of the instru-
ment because the willingness of subjects to complete
the instrument relates to their perception that the
instrument measures the construct they agreed to
provide (Thomas, 1992). Face validity is often consid-
ered a step before or an aspect of content validity.
Content validity examines the extent to which the
measurement method includes all the major elements
relevant to the construct being measured. This evi-
dence is obtained from the following three sources:
the literature, representatives of the relevant popula-
tions, and content experts (DeVon et al., 2007; Fawcett
& Garity, 2009; Waltz et al., 2010).
Documentation of content validity begins with
development of the instrument. The first step of instru-
ment development is to identify what is to be mea-
sured; this is referred to as the universe or domain of
the construct. You can determine your domain through
a concept analysis or an extensive literature search.
Qualitative methods can also be used for this purpose.
Johnson and Rogers (2006) developed the Medication-
Taking Questionnaire (MTQ) based on purposeful
action dimensions to determine the decision-making
process of individuals for adherence to medication
treatment for hypertension. They described their initial
instrument development process as follows.
“A total of 20 items (need, n = 8; effectiveness, n = 6;
and safety, n = 6) were initially developed to tap the
three underlying dimensions of purposeful action
based on the statements given by participants in a
qualitative study (Johnson, 2002; Johnson, Williams, &
Marshall, 1999). The method for item construction
was guided by the principles outlined in DeVellis
(1991) and Streiner and Norman (1995).… The
MTQ: Purposeful Action items were arranged in
a 7-point, Likert-type format describing responses
based on agreement (7 = always agree, 6 = very fre-
quently agree, 5 = usually agree, and 4 = occasionally
agree, 3 = rarely agree, 2 = almost never agree, 1 =
never agree). The 7-response option was used in an
attempt to obtain optimal variance while discouraging
a ceiling effect (Steiner & Norman, 1995). Higher
scores for the MTQ: Purposeful Action indicated
greater intent to take medications based on per-
ceived need, effectiveness, and safety.” (Johnson &
Rogers, 2006, p. 339)
Researchers need to describe the procedures used
to develop or select items for the instrument that rep-
resent the domain of the construct. One helpful strat-
egy commonly used is to develop a blueprint or matrix,
such as was used in developing test items for an exam-
ination that was done by Johnson (2002) in her dis-
sertation focused on development of the MTQ.
However, before developing such items, the blueprint
specifications must be submitted to an expert panel to
validate that they are appropriate, accurate, and repre-
sentative. At least five experts are recommended,
although a minimum of three experts is acceptable if
you cannot locate additional individuals with expertise
in the area. Researchers might seek out individuals
with expertise in various fields—for example, one

individual with knowledge of instrument develop-
ment, a second with clinical expertise in an appropri-
ate field of practice, and a third with expertise in
another discipline relevant to the content area.
The experts need specific guidelines for judging the
appropriateness, accuracy, and representativeness of
the specifications. Berk (1990) recommended that the
experts first make independent assessments and then
meet for a group discussion of the specifications. The
instrument specifications then can be revised and
resubmitted to the experts for a final independent
assessment. Davis (1992) recommended that the
researcher provide expert reviewers with theoretical
definitions of concepts and a list of which instrument
items are expected to measure each of the concepts.
The researcher asks the reviewers to judge how well
each of the concepts has been represented in the
instrument.
Researchers need to determine how to measure the
domain. The item format, item content, and proce-
dures for generating items must be carefully described.
Items are then constructed for each cell in the matrix,
or observational methods are designated to gather data
related to a specific cell. Researchers are expected to
describe the specifications used in constructing items
or selecting observations. Sources of content for items
must be documented. Then researchers can assemble,
refine, and arrange the items in a suitable order before
submitting them to the content experts for evaluation.
Specific instructions for evaluating each item and the
total instrument must be given to the experts.
In developing content validity for an instrument,
researchers can calculate a content validity ratio
(CVR) for each item on a scale by rating it 0 (not
necessary), 1 (useful), or 3 (essential). A method for
calculating the CVR was developed by Lawshe (1975)
and is presented in Table 16-1 (DeVon et al., 2007).
Minimum CVR scores for including items in an instru-
ment can be based on a one-tailed test with a 0.05 level
of significance.
The content validity score calculated for the com-
plete instrument is called the content validity index
(CVI). The CVI was developed to obtain a numerical
value that reflects the level of content-related validity
evidence for a measurement method (Waltz & Bausell,
1981). In calculating CVI, experts rate the content
relevance of each item in an instrument using a 4-point
rating scale. Lynn (1986, p. 384) recommended stan-
dardizing the options on this scale to read as follows:
“1 = not relevant; 2 = unable to assess relevance
without item revision or item is in need of such revi-
sion that it would no longer be relevant; 3 = relevant
but needs minor alteration; 4 = very relevant and
CHAPTER 16  Measurement Concepts 395
succinct.” In addition to evaluating existing items, the
experts were asked to identify important areas not
included in the instrument. The calculation for the
CVI is presented in Table 16-1 using the format devel-
oped by Lynn (1986). Complete agreement needs to
exist among the expert reviewers to retain an item with
seven or fewer reviewers. If few reviewers are used
and many of the experts support most of the items on
an instrument, this often results in an inflated CVI and
an inflation in the content validity of the instrument
(DeVon et al., 2007).
As presented earlier, Johnson and Rogers (2006)
developed the MTQ: Purposeful Action and described
their content validity testing process and outcomes as
follows.
“Content validity testing was undertaken to deter-
mine clarity and relevance of content. Participants and
experts were given verbal instructions and a packet
consisting of a consent form, written instructions,
clarity instrument, content validity instrument, and
demographic questionnaire. The clarity instrument
asked participants to rate items as clear or unclear
(Imle & Atwood, 1988). Participants were given a
definition of each subscale and asked to rate each
item’s relevancy using a 4-point scale from 1 (irrele-
vant) to 4 (extremely relevant; Lynn, 1986). Space
was provided to make comments after each rating
procedure. (p. 339)
Items met clarity criterion if 70% of participants
rated the item as clear and the content validity crite-
rion if 80% of participants rated the item as 3 or 4
(Imle & Atwood, 1988; Lynn, 1986). The comments
from the clarity and content validity criterion were
used to revise the MTQ: Purposeful Action items and
subscales.…
Of the 20 MTQ: Purpose Action items, 19 achieved
clarity and content validity agreement. The 1 item
that had an unacceptable clarity agreement was even-
tually eliminated from the questionnaire. Profession-
als expressed a concern about the lack of specificity
in the questions, but that was not an issue for the
hypertensive participants. For example, one profes-
sional indicated that the item, ‘Blood pressure pills
keep me from having problems,’ lacked specificity.
Because the purpose of this questionnaire was to
establish a general screening tool for individuals who
potentially may choose not to take their medications
rather than to create a diagnostic tool, the partici-
pants’ scores were given priority. Of the 20 items [see
396 UNIT TWO  The Research Process
TABLE 16-1  Two Methods of Calculating the Content Validity Ratio (CVR) and the Content Validity
Index (CVI)
Lawshe (1975)
Rating scale Scale used for rating items
0 1 3 1
Not necessary Useful Essential
Calculations To calculate CVR (a score for individual
scale items)
CVR = (ne − N/2)/(N/2)
Note: ne = The number of experts who rated
an item as “essential”
N = the total number of experts. Example: If
8 of 10 experts rated an item as essential,
CVR would be (8 − 5/5) = 0.60
Acceptable Depends on number of reviewers
range
From DeVon, H. A., Block, M. E., Moyle-Wright, P., Ernst, D. M., Hayden, S. J., Lazzara, D. J., Savoy, S. M., & Kostas-Polston, E. (2007). A psychometric
toolbox for testing validity and reliability. Journal of Nursing Scholarship, 39(2), 158.
Table 16-2 for the 20 items in the original question-
naire], 12 underwent minor grammatical revisions
guided by the comments of both the participants and
professionals. For example, items were made specific
to blood pressure and the term medication was
changed to pills. Several items were reworded, or the
tense of the verb was changed.” (Johnson & Rogers,
2006, pp. 341-342)
Before sending the instrument to experts for evalu-
ation, researchers need to decide how many experts
must agree on each item and on the total instrument
for the content to be considered valid. Items that do
not achieve minimum agreement by the expert panel
must be either eliminated from the instrument or
revised (DeVon et al., 2007; Lynn, 1986). Johnson and
Rogers (2006) described their panel of reviewers, who
were health professionals and patients prescribed anti-
hypertensive medications, for the MTQ in the follow-
ing excerpt.
“Content validity testing was conducted in a sample
of five hypertensive patients and five health care
Lynn (1986)
Scale used for rating items
2 3 4
Irrelevant Extremely Relevant
CVI for each scale item is the proportion of experts who rate the
item as a 3 or 4 on a 4-point scale. Example: If 4 of 6 content
experts rated an item as relevant (3 or 4), CVI would be: 4/6
= 0.67
This item would not meet the 0.83 level of endorsement required
to establish content validity using a panel of 6 experts at the
0.05 level of significance. Therefore, it would be dropped
CVI for the entire scale is the proportion of the total number of
items deemed content valid. Example: If 77 of 80 items were
deemed content valid, CVI would be: 77/80 = 0.96
Depends on number of reviewers
professionals who examined the MTQ for clarity and
content relevance (Imle & Atwood, 1988; Lynn,
1986). Professionals were invited to participate in the
study based on their known experience with antihy-
pertensive treatment and included two family physi-
cians, a cardiology nurse practitioner, a nurse working
with a statewide cardiovascular disease program, and
a nurse researcher who had published articles on
adherence. All professionals were Anglo American
and were nearly equally divided with regard to
gender.
Participants for the content validity phase who had
been prescribed antihypertensive medications and
lived in a situation in which they managed their own
medications were recruited through healthy aging
clinics, worksite wellness programs, hospital outpa-
tient clinics, and hospital emergency departments in
the intermountain west. The five hypertensive par-
ticipants were Anglo American, had at least a high
school education and ranged in age from 48 to 90
years (M = 62.0 ± 16.4).” (Johnson & Rogers, 2006,
p. 338)
Johnson and Rogers (2006) provided excellent
detail about the development of their instrument and
 CHAPTER 16  Measurement Concepts 397

TABLE 16-2  Medication-Taking Questionnaire: Purposeful Action—Initial 20 Items Statistics

Item-Total Mann-Whitney
M SD Correlation Adherence p Values*
Perceived Need
My blood pressure pills keep me from having a stroke. 5.8 1.5 0.58 0.08
I need to take my blood pressure pills. 6.4 1.4 0.77 0.01
I take my blood pressure pills for my health. 6.5 1.3 0.75 0.01
Blood pressure pills keep me from having health-related 5.7 1.5 0.63 0.17
problems.
I could have health problems if I do not take my blood 6.1 1.3 0.74 0.13
pressure pills.
It’s not a problem if I miss my blood pressure pills.† 5.1 2.0 0.30 0.02
I would rather treat my blood pressure without pills.† 4.1 2.3 0.37 0.26
I am OK if I do not take my blood pressure pills.† 5.6 1.8 0.64 0.012
Perceived Effectiveness
My blood pressure will come down enough without pills.† 5.4 1.8 0.40 0.10
I will have problems if I don’t take my blood pressure pills. 6.1 1.4 0.63 0.001
My blood pressure pills control my blood pressure. 6.0 1.4 0.66 0.46
Blood pressure pills benefit my health. 6.1 1.4 0.74 0.01
I feel better when I take my blood pressure pills. 5.4 1.8 0.56 0.01
I have problems finding pills that will control my blood 5.7 1.8 0.09 0.059
pressure.†
Perceived as Safe
The side effects from my blood pressure pills are a problem.† 5.2 1.9 0.40 0.10
The side effects from my blood pressure pills are harmful.† 5.6 1.8 0.63 0.27
My blood pressure pills are safe. 5.8 1.4 0.66 0.47
Taking my blood pressure pills is not a problem because they 6.0 1.4 0.74 0.02
benefit my health.
My blood pressure pills cause other health problems.† 5.4 1.8 0.56 0.35
I will become dependent on my blood pressure pills.† 3.9 2.3 −0.5 0.20

From Johnson, M. J., & Rogers, S. (2006). Development of the Purposeful Action Medication-Taking Questionnaire. Western Journal of Nursing Research,
28(3), 344.
*Difference between low (scored 1-3) versus high (scored 7-10) adherence.
†Reverse coded.
M, Mean; SD, standard deviation.

the process for determining content validity. They also
provided extensive information about the expert
review panel for conducting the content validity
testing. The strength of the review panel is that it
included both health professionals and patients taking
medications for hypertension. However, since all the
reviewers were Anglo American, there was no ethnic
diversity in the review process. The MTQ was a Likert
scale with 7-point response options (described earlier),
so it would be clearer if the researchers had called the
MTQ a Likert scale versus a questionnaire (Waltz
et al., 2010).
With some modifications, the content validity pro-
cedure previously described can be used with exist-
ing instruments, many of which have never been
evaluated for content-related validity. With the per-
mission of the author or researcher who developed
the instrument, you could revise the instrument to
improve its content-related validity (Lynn, 1986). In
addition, the panel of experts or reviewers evaluating
the items of the instrument for content validity might
also examine it for readability and language accept-
ability to possible subjects or data gatherers (Berk,
1990; DeVon et al., 2007).
Readability of an Instrument
Readability is an essential element of the validity and
reliability of an instrument. Assessing the level of
readability of an instrument is simple and takes
seconds with the use of a computer. There are more
398 UNIT TWO  The Research Process
than 30 readability formulas. These formulas count
language elements in the document and use this infor-
mation to estimate the degree of difficulty a reader
may have in comprehending the text. Readability for-
mulas are now a standard part of word-processing
software. Box 16-1 provides instructions for using the
Fog formula to determine the readability of a measure-
ment method.
Although readability has never been formally iden-
tified as a component of content validity, it is essential
that the items of an instrument be comprehended by
subjects. Miller and Bodie (1994) suggested that the
researcher should directly assess the reading compre-
hension level of the study population before using a
formula to calculate an instrument’s readability. They
indicated that it is a mistake to assume that someone’s
literacy is equivalent to the last grade level the indi-
vidual completed. Miller and Bodie (1994) recom-
mended that researchers use the Classroom Reading
Box 16-1 How to Find the Fog Index
(Fog Formula)
1. Pick a sample of writing 100 to 125 words
long. Count the average number of words
per sentence. In counting, treat
independent clauses as separate sentences.
“In school we studied; we learned; we
improved” is three sentences.
2. Count the words of three syllables or more.
Do not count: (a) capitalized words, (b)
combinations of short words such as
butterfly or manpower, or (c) verbs made
into three syllables by adding “-es” or “-ed”
such as trespasses or created. Divide the
count of long words by the number of
words in the passage to get the percentage.
3. Add the results from no. 1 (average
sentence length) and no. 2 (percentage of
long words). Multiply the sum by 0.4.
Ignore the numbers after the decimal point.
4. The result is the years of schooling needed
to understand the passage tested easily.
Few readers have more than 17 years of
schooling, so give any passage higher than
17 a Fog Index of 17-plus.
Adapted from Gunning, R., & Kallan, R. A. (1994).
How to take the fog out of business writing.
Chicago, IL: Dartnell. The Fog Index is a service
mark licensed exclusively to RK Communication
Consultants by D. and M. Mueller.
Inventory (CRI), which is based on the Flesch, Space,
Dale, and Fry reading comprehension scales (Flesch
1984; Silvaroli, 1986). This instrument determines the
level at which an individual can comprehend written
material without assistance. Johnson and Rogers
(2006) described the readability of their MTQ: Pur-
poseful Action as follows.
“Items were worded at approximately a sixth-grade
reading level, evaluated by using the Flesch-Kincaid
grade-level assessment program in Microsoft Word
(2000) (Rasin, 1997). Items ranged from a 1.0 to 6.2
grade level, with a 3.5 grade level readability score for
the overall questionnaire.” (Johnson & Rogers, 2006,
p. 339)
Validity from Factor Analysis
Factor analysis is a valuable approach for determining
evidence of an instrument’s construct validity. This
analysis technique is used to determine the various
dimensions or subcomponents of a phenomenon of
interest. To employ factor analysis, the instrument
must be administered to a large, representative sample
of participants at one time. Usually the data are ini-
tially analyzed with exploratory factor analysis (EFA)
to examine relationships among the various items of
the instrument. Items that are closely related are clus-
tered into a factor. The researcher needs to preset the
minimum loading for an item to be included in a
factor. The minimum loading is usually set at 0.30 but
might be as high as 0.50 (Waltz et al., 2010). The
factors identified are the subcomponents of the con-
struct the instrument was developed to measure.
Determining and naming the factors identified through
EFA require detailed work on the part of the researcher.
The researcher can validate the number of factors or
subcomponents in the instrument and measurement
equivalence among comparison groups through the
use of confirmatory factor analysis (CFA). Items that
do not fall into a factor (because they do not correlate
with other items) may be deleted (DeVon et al., 2007;
Munro, 2005; Stommel, Wang, Given, & Given, 1992;
Waltz et al., 2010). A more extensive discussion of
EFA and CFA is presented in Chapter 23.
Johnson and Rogers (2006) conducted an EFA to
determine the factor structure for their MTQ: Pur-
poseful Action scale. The EFA identifies the specific
factors or subscales for the scale and the items that fit
each of these subscales. The original scale had 20
items sorted into three subscales (labeled perceived
need, perceived effectiveness, and perceived as safe)
that are identified in Table 16-2. The EFA and the
results are presented in Table 16-3 and described as
follows.
 “Factor analysis is a grouping technique that allows for
evaluation of the dimensionality of scales (Munro, 2001;
Nunnally & Bernstein, 1994). A principal axis factoring
solution with an oblimen rotation, considered the best
analysis for achieving a theoretical solution uncontami-
nated by unique and random error variability was
undertaken.…
The EFA yielded two interpretable factors [see Table
16-3], which eliminated six additional items because of
factor loadings < 0.40. The first factor merged the need
and effectiveness items along with one item from the
Safe subscale. This factor was renamed treatment ben-
efits (benefits). The second factor, renamed medication
safety (safety), was reduced to three of the original safe
subscales items.
The Benefits subscale retained nine items that focused
on the actual perceived benefits of treatment, such as
preventing a stroke, controlling blood pressure, prevent-
ing further health problems, and feeling better when
taking medications, which indicated a desire to control
blood pressure to maintain and promote health and
well-being. The subscale had an eigenvalue of 5.5 and a
total item variance explained by the factor of 46%.…
The Safety subscale (three items) focused on side
effects of medications. This subscale had an eigenvalue
of 1.9 and a total item variance explained by the factor
of 16%.… Together, the two factor solution had a coef-
ficient alpha [Cronbach alpha] of 0.87 and an explained
variance of 62%.” (Johnson & Rogers, 2006, pp.
343-346)

TABLE 16-3 Principal Axis Factor Analysis with Oblimen Rotation Pattern (and Structure
in Parentheses) Coefficients for the Medication-Taking Questionnaire:  
Purposeful Action Two-Factor Solution
Factor Loadings
Eigen- % Variance Coefficient
1 2 h2 value Explained Alpha
Treatment benefits 5.5 45.9 0.90
I need to take my blood pressure 0.84 (0.85) (0.34) 0.73
pills.
Taking my blood pressure pills is 0.82 (0.84) (0.35) 0.72
not a problem because they
benefit my health.
I could have problems if I do not 0.81 (0.84) (0.21) 0.70
take my blood pressure pills.
Blood pressure pills keep me from 0.81 (0.79) (0.16) 0.63
having health-related problems.
My blood pressure pills keep me 0.75 (0.75) (0.23) 0.55
from having a stroke.
I feel better when I take my blood 0.74 (0.74) (0.21) 0.55
pressure pills.
My blood pressure pills control my 0.74 (0.74) (0.26) 0.55
blood pressure.
I am OK if I do not take my blood 0.72 (0.71) 0.52
pressure pills.*
My blood pressure will come 0.54 (0.48) 0.30
down enough without pills.*
Medication safety 1.9 15.6 0.80
The side effects from my blood (0.19) 0.87 (0.86) 0.74
pressure pills are harmful.*
The side effects from my blood (0.27) 0.84 (0.86) 0.71
pressure pills are a problem.*
My blood pressure pills cause (0.29) 0.82 (0.83) 0.70
other health problems.*
Total 7.4 61.5 0.88

From Johnson, M. J., & Rogers, S. (2006). Development of the Purposeful Action Medication-Taking Questionnaire. Western Journal of Nursing Research,
28(3), 345.
Note: n = 229.
*Item required reverse coding. Factor loadings in parentheses represent structure coefficients. If patterned or structure coefficient is not listed, the value was
<0.15.
400 UNIT TWO  The Research Process

Johnson and Rogers (2006) provided a clear,
concise rationale for the revisions that they made in
their MTQ: Purposeful Action scale. In this study, the
revised scale showed internal consistency (Cronbach
alpha = 0.87) and construct validity obtained through
content analysis and EFA. Johnson and Rogers (2006,
p. 348) also conducted CFA that “supported the
hypothesis that benefits and safety [factors or sub-
scales] underlie the cognitive component of medica-
tion taking in hypertensive medications.”
Validity from Structural Analysis
Structural analysis is used to examine the structure of
relationships among the various items of an instru-
ment. This approach provides insights beyond that
provided by factor analysis. Factor analysis deter-
mines what items group together. Structural analysis
determines how each item is related to other items.
Structural analysis goes a step beyond factor analysis.
The exact relationship of each item in a factor is exam-
ined through correlational analyses.
Convergent Validity
In some cases, instruments are available to measure a
construct, such as depression. However, for many pos-
sible reasons, the existing instruments may be unsat-
isfactory for a particular purpose or a particular
population, such as measuring major depression in
young children, and the researcher may choose to
develop a new instrument for a study. In examining
the validity of the new instrument, it is important to
determine how closely the existing instruments
measure the same construct as the newly developed
instrument (convergent validity). One can administer
all of the instruments (the new one and the existing
ones) to a sample concurrently and evaluate the results
using correlational analyses. If the measures are highly
positively correlated, the validity of each instrument
is strengthened.
Johnson and Rogers (2006) strengthened the valid-
ity of their 12-item MTQ: Purposeful Action scale
and its subscales (benefit and safety) by correlating
them with a variety of other instruments (Hamilton
Health Belief Model Hypertension [HBM] Scale with
the HBM subscales of Susceptibility, Severity, Bene-
fits, and Barriers; Lifestyle Busyness Questionnaire
with Busyness and Routine subscales; and Blood
Pressure Feedback Log). The results of these correla-
tions are presented in Table 16-4. The significant
positive correlations of 0.3 to 0.63 between the exist-
ing scales (Hamilton HBM Scale with Susceptibility
and Benefits subscales and the Blood Pressure Feed-
back Log for adherent group) and the MTQ and the
benefits subscale add to the construct validity of these
instruments. This is an example of examining conver-
gent validity for this scale, which was strong for the

TABLE 16-4 Validity Correlation Coefficients for the Medication-Taking Questionnaire: Purposeful
Action and Subscales
MTQ: Purposeful Action MTQ Benefit Subscale MTQ Safe Subscale
Hamilton HBM Scalea 0.30** 0.43** −0.12
HBM: Susceptibility subscale 0.36** 0.41** 0.01
HBM: Severity subscale 0.00 0.12 −0.27**
HBM: Benefits subscale 0.58** 0.63** 0.19
HBM: Barriers subscale −0.49** −0.42** −0.41**
Lifestyle Busyness Questionnaireb 0.08 0.11 −0.02
Busyness subscale 0.10 0.13 0.01
Routine subscale −0.07 −0.06 −0.06
Blood Pressure Feedback Logc
  Adherent 0.53** 0.54** 0.25*
  Nonadherent −0.60** −0.50** −0.53**

From Johnson, M. J., & Rogers, S. (2006). Development of the Purposeful Action Medication-Taking Questionnaire. Western Journal of Nursing Research,
28(3), 347.
HBM, Health Belief Model Hypertension Scale.
a
n = 107.
b
n = 104.
c
n = 102.
*p < 0.05, two-tailed.
**p < 0.01, two-tailed.

MTQ and the benefit subscale but not the safety
subscale.
Divergent Validity
Sometimes, instruments can be located that measure a
construct opposite to the construct measured by the
newly developed instrument (divergent validity). For
example, if the newly developed instrument measures
hope, you could search for an instrument that mea-
sures hopelessness or despair. If possible, you could
administer this instrument and the instruments used to
test convergent validity at the same time. This approach
of combining convergent and divergent validity testing
of instruments is called multitrait-multimethod
(MT-MM).
The MT-MM approach can be used when research­
ers are examining two or more constructs being mea-
sured by two or more measurement methods (DeVon
et al., 2007). Correlational procedures are conducted
with the different scales and subscales. If the conver-
gent measures positively correlate and the divergent
measures negatively correlate with other measures,
validity for each of the instruments is strengthened.
Johnson and Rogers (2006) used an MT-MM ap-
proach in examining convergent and divergent valid-
ity related to their MTQ: Purposeful Action. The
convergent validity findings were discussed in the
previous section. Table 16-4 shows that the MTQ and
the subscales benefits and safety were significantly,
negatively correlated with HBM Barriers subscale
and the Blood Pressure Feedback Log for nonadher-
ent hypertensive patients. These scales measure the
opposite construct from the MTQ and its subscales,
so these significant negative correlations indicated
that the construct validity was strengthened for these
instruments. The correlations with the Lifestyle
Busyness Questionnaire were too low (−0.07 to 0.13)
to add to the convergent or divergent validity of the
MTQ: Purposeful Action scale and subscales (see
Table 16-4).
Validity from Contrasting (or Known) Groups
To test the validity of an instrument, identify groups
that are expected (or known) to have contrasting
scores on the instrument. Generate hypotheses about
the expected response of each of these known groups
to the construct. Next, select samples from at least
two groups that are expected to have opposing
responses to the items in the instrument. Hagerty
and Patusky (1995) developed a measure called the
Sense of Belonging Instrument (SOBI). They tested
the instrument on the following three groups: com-
munity college students, clients diagnosed with major
CHAPTER 16  Measurement Concepts 401
depression, and retired Roman Catholic nuns, as
described in the following excerpt.
“The community college sample was chosen for its
heterogeneous mix of students and ease of access.
Depressed clients were included based on the litera-
ture and the researcher’s clinical experience that
interpersonal relationships and feeling ‘connected’
are difficult when one is depressed. It was hypothe-
sized that the depressed group would score signifi-
cantly lower on the SOBI than the student group. The
nuns were selected to examine the performance of
the SOBI with a group that, in accordance with the
theoretical basis of the instrument, should score sig-
nificantly higher than the depressed and student
groups.” (Hagerty & Patusky, 1995, p. 10)
The nuns had the highest sense of belonging, the
student groups followed, and the depressed group had
the lowest sense of belonging. This test increased the
validity of the instrument in that the scores of groups
were as anticipated.
Evidence of Validity from
Discriminant Analysis
Instruments sometimes have been developed to
measure constructs closely related to the construct
measured by a newly developed instrument. For
example, an instrument might exist to measure medi-
cation management in patients with diabetes that is
similar to the MTQ: Purposeful Action developed by
Johnson and Rogers (2006) for patients with hyperten-
sion. If such instruments can be located, you can
strengthen the validity of the MTQ instrument and the
other medication management instrument by testing
the extent to which the two instruments can finely
discriminate between these related concepts. Testing
of this discrimination involves administering the two
instruments simultaneously to a sample and perform-
ing a discriminant analysis (see Kerlinger & Lee,
2000, for a discussion of discriminant analysis).
Validity from Prediction of Future Events and
Concurrent Events
The ability to predict future performance or attitudes
on the basis of instrument scores adds to the validity of
an instrument. Nurse researchers often want to deter-
mine the ability of scales developed to measure selected
health behaviors to predict the future health status of
individuals. One approach might be to examine
reported stress levels of selected individuals in highly
stressful careers such as nursing and see if stress is
402 UNIT TWO  The Research Process
linked to the nurses’ future incidence of hypertension.
If study analysis links stress to future hypertension,
measuring a nurse’s stress could be used to predict his
or her future likelihood of becoming hypertensive. For
example, the validity of the Nursing Stress Scale (NSS)
could be tested in this manner. French, Lenton, Walters,
and Eyles (2000) did an expanded evaluation of the
reliability and validity of the NSS with a random
sample of 2280 nurses working in a wide range of
healthcare settings. They noted that the NSS included
nine subscales: death and dying, conflict with physi-
cians, inadequate preparation, problems with supervi-
sors, workload, problems with peers, uncertainty
concerning treatment, patients and their families, and
discrimination. Confirmatory factor analyses sup-
ported the factor structure. Cronbach alpha coefficients
of eight of the subscales were 0.70 or higher. The NNS
showed reliability and validity in measuring stress in
nurses and could be used in a study to determine the
link to hypertension. The accuracy of predictive valid-
ity is determined through regression analysis.
Validity can be tested by examining the ability to
predict the current value of one measure on the basis
of the value obtained on the measure of another
concept. For example, you might be able to predict the
self-esteem score of an individual who had a high
score on an instrument to measure coping. A person
who received a high score on coping might be expected
also to have a high self-esteem score. If these results
held true in a study in which both measures were
obtained concurrently, the two instruments would
have evidence of concurrent validity.
Successive Verification of Validity
After the initial development of an instrument, it is
hoped that other researchers would begin using the
instrument in additional studies. Each of these studies
could add to the validity and reliability information on
the instrument. There is a successive verification of
the validity of the instrument over time when used in
a variety of studies with different populations and set-
tings. For example, additional researchers are using
the MTQ: Purposeful Action in their studies, which
has the potential to add to the validity of this question-
naire (Lehane & McCarthy, 2007).
Accuracy, Precision, and Error of
Physiological Measures
Accuracy and precision of physiological and biochem-
ical measures tend not to be reported in published
studies. These routine physiological measures are
assumed to be accurate and precise, an assumption
that is not always correct. The most common physi-
ological measures used in nursing studies are blood
pressure, heart rate, weight, and temperature. These
measures are often obtained from the patient’s record
with no consideration given to their accuracy. It is
important to consider the possibility of differences
between the obtained value and the true value of phys-
iological measures. Thus, researchers using physio-
logical measures need to provide evidence of the
accuracy and precision of their measures (Ryan-
Wenger, 2010).
The evaluation of physiological measures may
require a slightly different perspective from that
applied to behavioral measures, in that standards for
most biophysical measures are defined by national and
international organizations such as the International
Organization of Standardization (IOS) (2011a) and the
Clinical Laboratory Standards Institute (CLSI) (2011).
CLSI develops standards for laboratory and other
healthcare-related biophysical measures. The IOS is
the world’s largest developer and publisher of interna-
tional standards and includes a network of 160 coun-
tries (see IOS website for details at http://www.iso.org/
iso/home.htm). The ISO standards were developed to
accomplish the following:
• Make the development, manufacturing, and supply
of products and services more efficient, safer, and
cleaner
• Facilitate trade between countries and make it
fairer
• Provide governments with a technical base for
health, safety, and environmental legislation and
conformity assessment
• Share technological advances and good manage-
ment practice
• Disseminate innovations
• Safeguard consumers and users in general of prod-
ucts and services
• Make life simpler by providing solutions to
common problems (ISO, 2011b)
You can locate the standards for different biophysi-
cal equipment, products, or services that you might
use in a study or in clinical practice. Within IOS, the
Joint Committee for Guides in Metrology (JCGM) has
two major areas of focus: (1) Guide to the Expression
of Uncertainty in Measurement (GUM) and (2) Inter-
national Vocabulary of Basic and General Terms in
Metrology (VIM) (JCGM, 2011). VIM is a document
that standardizes terminology related to biophysical
measurements, such as accuracy, precision, error, sen-
sitivity, specificity, and likelihood ratio that are
described in this section.

Accuracy
Accuracy involves determining the closeness of the
agreement between the measured value and the true
value of the quantity being measured (JCGM, 2011).
Accuracy is similar to validity, in which evidence of
content-related validity addresses the extent to which
the instrument measured the construct or domain
defined in the study. New measurement devices are
compared with existing standardized methods of mea-
suring a biophysical property or concept. For example,
measures of oxygen saturation with a pulse oximeter
were correlated with arterial blood gas measures of
oxygen saturation to determine the accuracy of the
pulse oximeter. Thus, there should be a very strong,
positive correlation (≥0.95) between pulse oximeter
and blood gas measures of oxygen saturation to
support the accuracy of the pulse oximeter (CLSI,
2011).
Accuracy of physiological measures depends on
the (1) quality of the measurement equipment or
device, (2) detail of the data collection plan, and (3)
expertise of the data collector (Ryan-Wenger, 2010).
The data collector or person conducting the biophysi-
cal measures must do the measurements in a standard-
ized way that is usually directed by a measurement
protocol. For example, BP readings in a study need to
be taken using a protocol: (1) place the subject in a
chair and allow 5 minutes of rest; (2) remove restric-
tive clothing from the subject’s arm; (3) measure the
subject’s upper arm and select the appropriate cuff
size; (4) instruct the subject to place his or her feet flat
on the floor; (5) support the subject’s arm when taking
the BP reading; and (6) take three BP readings each 5
minutes apart, average the readings, and enter the
averaged BP reading into a computer. Some measure-
ments, such as arterial pressure, can be obtained by
the biomedical device producing the reading and auto-
matically recorded in a computerized database. This
type of data collection greatly reduces the potential for
error and increases accuracy and precision.
The biomedical device or equipment used to
measure a study variable must be examined for accu-
racy. Researchers need to document the extent to
which the biophysical measure is an accurate mea-
surement of a study variable and the level of error
expected. Reviewing the ISO (2011b) and CLSI
(2011) standards could provide essential accuracy
information and information about the company that
developed the device or equipment.
Selectivity, an element of accuracy, is “the ability
to identify correctly the signal under study and to
distinguish it from other signals” (Gift & Soeken,
1988, p. 129). Because body systems interact, the
CHAPTER 16  Measurement Concepts 403
researcher must choose instruments that have selectiv-
ity for the dimension being studied. For example, elec-
trocardiographic readings allow one to differentiate
electrical signals coming from the myocardium from
similar signals coming from skeletal muscles.
To determine the accuracy of biochemical mea-
sures, review the standards set by CLSI (2011) and
determine if the laboratory where the measures are
going to be obtained is certified. Most laboratories are
certified, so researchers could contact experts in the
agency on the laboratory procedure and ask them to
describe the process for collection, analysis, and
values obtained for specimens. You might also ask
these experts to judge the appropriateness of the bio-
physical device for the construct being measured in
the study. Use contrasted groups’ techniques by select-
ing a group of subjects known to have high values on
the biochemical measures and comparing them with a
group of subjects known to have low values on the
same measure. In addition, to obtain concurrent valid-
ity, compare the results of the test with results from
the use of a known standard (CLSI, 2011), such as the
example of the comparison of pulse oximeter values
with blood gas values for oxygen saturation.
Precision
Precision is the degree of consistency or reproduc-
ibility of measurements made with physiological
instruments or devices. There should be close agree-
ment in the replicated measures of the same variable
or object under specified conditions (Ryan-Wenger,
2010). Precision is similar to reliability. The precision
of most physiological devices or equipment is deter-
mined by the manufacturer and is part of quality
control testing done in the agency using the device.
Similar to accuracy, precision depends on the collector
of the biophysical measures and the consistency of the
measurement equipment or device. The protocol for
collecting the biophysical measures improves preci-
sion and accuracy (see the previous example of pro-
tocol to measure BP).
The data collectors need to be trained to ensure
consistency, which is documented with intrarater
(within a single data collector) and interrater (among
data collectors) percentages of agreements. The kappa
coefficient of agreement is one of the most common
and simplest statistics to determine intrarater and
interrater accuracy and precision for nominal level
data (Cohen, 1960; Ryan-Wenger, 2010). The equip-
ment and devices used to measure physiological vari-
ables need to be maintained according to the standards
set by IOS and the manufacturers of the devices. Many
devices need to be recalibrated according to set criteria
404 UNIT TWO  The Research Process
to ensure consistency in measurements. Because of
fluctuations in some physiological measures, test-
retest reliability might be inappropriate.
Two procedures are commonly used to determine
the precision of biochemical measures. One is the
Levy-Jennings chart. For each analysis method, a
control sample is analyzed daily for 20 to 30 days. The
control sample contains a known amount of the sub-
stance being tested. The mean, the standard deviation,
and the known value of the sample are used to prepare
a graph of the daily test results. Only 1 value of 22 is
expected to be greater than or less than 2 standard
deviations from the mean. If two or more values are
more than 2 standard deviations from the mean, the
method is unreliable in that laboratory. Another
method of determining the precision of biochemical
measures is the duplicate measurement method. The
same technician performs duplicate measures on ran-
domly selected specimens for a specific number of
days. The results are essentially the same each day if
there is high precision. Results are plotted on a graph,
and the standard deviation is calculated on the basis
of difference scores. The use of correlation coeffi-
cients is not recommended (DeKeyser & Pugh, 1990).
Sensitivity
Sensitivity of physiological measures relates to “the
amount of change of a parameter that can be measured
precisely” (Gift & Soeken, 1988, p. 130). If changes
are expected to be small, the instrument must be very
sensitive to detect the changes. Thus, sensitivity is
associated with effect size (see Chapter 15). With
some instruments, sensitivity may vary at the ends of
the spectrum. This is referred to as the frequency
response. The stability of the instrument is also related
to sensitivity. This feature may be judged in terms of
the ability of the system to resume a steady state after
a disturbance in input. For electrical systems, this
feature is referred to as freedom from drift (Gift &
Soeken, 1988).
Error
Sources of error in physiological measures can be
grouped into the following five categories: environ-
ment, user, subject, machine, and interpretation. The
environment affects both the machine and the subject.
Environmental factors include temperature, baromet-
ric pressure, and static electricity. User errors are
caused by the person using the instrument and may be
associated with variations by the same user, different
users, changes in supplies, or procedures used to
operate the equipment. Subject errors occur when the
subject alters the machine or the machine alters the
subject. In some cases, the machine may not be used
to its full capacity. Machine error may be related to
calibration or to the stability of the machine. Signals
transmitted from the machine are also a source of error
and can cause misinterpretation (Ryan-Wenger, 2010).
Sources of error in biochemical measures are bio-
logical, preanalytical, analytical, and postanalytical.
Biological variability in biochemical measures is due
to factors such as age, gender, and body size. Vari-
ability in the same individual is due to factors such as
diurnal rhythms, seasonal cycles, and aging. Preana-
lytical variability is due to errors in collecting and
handling of specimens. These errors include sampling
the wrong patients; using an incorrect container, pre-
servative, or label; lysis of cells; and evaporation.
Preanalytical variability may also be due to patient
intake of food or drugs, exercise, or emotional stress.
Analytical variability is associated with the method
used for analysis and may be due to materials, equip-
ment, procedures, and personnel used. The major
source of postanalytical variability is transcription
error. This source of error can be greatly reduced by
entering data into the computer directly (DeKeyser &
Pugh, 1990).
When the scores obtained in a study are at the
interval or ratio level, a commonly used method of
evaluating precision and accuracy errors is the Bland-
Altman chart (Bland & Altman, 1986). This chart is a
scatter plot of the differences between observed scores
on the Y-axis and the combined mean of the two
methods on the X-axis. The distribution of the differ-
ence scores is examined in context of the limits of
agreement that are drawn as a horizontal line across
the chart or scatter plot (see Chapter 23). The limits
are set by the researchers and might include 1 or 2
standard deviations from the mean or might be the
clinical standards of the maximum amount of error
that is safe. The data points are examined for level of
agreement (congruence) and for level of bias (system-
atic error). Outliers are readily visible from the chart,
and each outlier case should be examined to identify
the cause of such a large discrepancy. Clinical labora-
tory standards indicate that “more than 3 outliers per
100 observations suggest there are major flaws in the
measurement system” (Ryan-Wenger, 2010, p. 381).
Schell et al. (2011) conducted a study to compare
upper arm and calf automatic noninvasive BPs in chil-
dren in a pediatric intensive care unit (PICU). The
researchers documented the accuracy of their BP mon-
itoring equipment, training of their data collectors, and
the procedures for taking the BPs in their study. The
errors in precision and accuracy are documented with
Bland-Altman charts for systolic BP, diastolic BP, and

mean arterial pressure readings. The chart of the sys-
tolic BP is included as an example in Figure 16-10.
This study was conducted to determine an alternative
method of obtaining BP when the injuries of the child
prevent BP readings using the upper arm.
“BP Monitor
“BP was obtained using a Spacelabs Ultraview SL
monitoring system (Spacelabs Healthcare, Issaquah,
WA), which consists of hemodynamic parameter
modules that can be inserted into stationary bedside
and portable monitor housings. All monitoring func-
tions were controlled through the modules. During
data collection, each set of arm and calf BP measure-
ments was obtained simultaneously using two identi-
cal parameter modules: one inserted into the subject’s
stationary bedside housing and the other inserted in
to a portable monitor housing brought to the sub-
ject’s bedside. Modules and housings are inspected
and tested annually by Biomedical Support Services
to ensure accurate functioning. The accuracy of these
monitors for arm BPs meets or exceeds SP10-1992
Association for the Advancement of Medical Instru-
mentation standards (mean error = ±4.5 mm Hg, SD
= ±7.3 mm Hg) for arm measurements (White et al.,
1993). Spacelabs Healthcare did not report data
regarding accuracy of calf BPs.
“Training of Data Collectors
“Data were collected by five pediatric intensive care
nurses who attended a data training session that
addressed location of arm and calf sites, measure-
ment of limb circumference, and use of the RASS
[Richmond Agitation Sedation Scale]. The nurses also
attended a BP monitor in-service offered by the
Spacelab representative when the monitors were
adopted in the PICU in January 2006.…
“Procedure
“Subjects were placed in a supine position with the
head of bed elevated 30° as determined by a handheld
protractor or the degree indicator incorporated into
the bed frame. Subjects remained in this position for
at least 5 minutes prior to data collection. Cuff sizes
were selected based on limb circumferences mea-
sured to the nearest 0.5 cm. Spacelabs cuff sizes were
as follows: neonate, 6-11 cm; infant, 8-11 cm; child,
12-19 cm; small adult, 17-26 cm; and adult, 24-32 cm.
Per manufacturer’s recommendations, if circumfer-
ence overlapped two categories of cuff size, the
CHAPTER 16  Measurement Concepts 405
larger cuff was selected. Using a paper tape measure,
arm circumference was obtained at the point halfway
between the elbow and the shoulder. Calf circumfer-
ence was measured at the point midway between the
ankle and the knee. The BP cuffs were applied to the
arm and calf on the same side. Subjects’ extremities
were positioned at the side of their bodies, resting on
the bed, for all measurements.… Systolic, diastolic,
and mean BP values for the arm and calf as well as a
simultaneous heart rate were documented. Data col-
lectors notified the child’s nurse or physician if an
abnormal arm reading was obtained.” (Schell et al.,
2011, pp. 6-7)
“To promote best practice, clinicians should base
treatment choices on individual patient data, not
group data. Therefore, Bland-Altman analyses were
used to determine agreement between arm and calf
oscillometric BPs for individual subjects. Perfect
agreement occurs when all data points lie on the line
of equality of the X-axis. The bias (mean difference
between arm and calf pressures) systolic BP was
8.0 mm Hg with the limits of agreement −18.9 and
34.9 mm Hg. Limits of agreement indicated that 95%
of the sample falls between these values [see Figure
16-10]. The limits of agreement for diastolic BP were
−22.7 and 25.0 mm Hg with a bias of 1.1 mm Hg.”
(Schell et al., 2011, p. 9)
Schell et al. (2011) provided evidence of the accu-
racy, precision, and error of the BP monitoring equip-
ment used in their study. They also provided a detailed
discussion of the procedures for data collection that
followed a rigorous protocol to ensure accurate and
precise BP readings were obtained for all ages of chil-
dren based on their measured arm and calf sizes. The
data collectors were trained in BP monitoring by the
Spacelab representative, which would increase their
expertise in the use of the equipment. However, the
study would have been strengthened by a discussion
of the intrarater and interrater percentage of agreement
for the data collectors. The use of the Bland-Altman
plot to identify the error in precision and accuracy for
systolic BPs, diastolic BPs, and mean arterial pres-
sures added to the credibility of the findings. The
researchers found that the arm and calf BPs were not
interchangeable for many of the children 1 to 8 years
old. “Clinical BP differences were the greatest in chil-
dren between ages 2 and less than 5 years. Calf BPs
are not recommended for this population. If the calf is
unavoidable due to medical reasons, trending of BP
 406 UNIT TWO  The Research Process

from this site should remain consistent during the & Haynes, 2000). When you order a diagnostic test,
child’s stay” (Schell et al., 2011, p. 10). how can you be sure that the results are valid or accu-
rate? This question is best answered by current, quality
research to determine the sensitivity and specificity of
60 the test.
Calf Systolic BP - Arm Systolic BP

40 +1.96 SD Sensitivity and Specificity

20
0 8.0
–20
–40
60 80 100 120
AVERAGE of Calf Systolic BP and Arm Systolic BP
Figure 16-10  Bland-Altman plot of systolic BP. (From Schell, K.,
Briening, E., Lebet, R., Pruden, K., Rawheiser, S., & Jackson, B. [2011].
Comparison of arm and calf automatic noninvasive blood pressures in
pediatric intensive care patients. Journal of Pediatric Nursing, 26[1], 9.)
Sensitivity, Specificity, and
Likelihood Ratios
An important part of building evidence-based practice
(EBP) is the development, refinement, and use of
quality diagnostic tests and measures in research and
practice. Researchers want to use the most accurate
and precise measure or test in their study to promote
quality outcomes. If a quality diagnostic test does not
exist, some nurses have participated in the develop-
ment and refinement of new biophysical tests. Clini-
cians want to know what diagnostic test, such as
laboratory or imaging study, to order to help screen
for and accurately determine the absence or presence
of an illness (Sackett, Straus, Richardson, Rosenberg,
34.9 The accuracy of a screening test or a test used to
confirm a diagnosis is evaluated in terms of its ability
Mean to assess correctly the presence or absence of a disease
or condition as compared with a gold standard. The
gold standard is the most accurate means of currently
–1.96 SD diagnosing a particular disease and serves as a basis
–18.9 for comparison with newly developed diagnostic or
screening tests (Campo, Shiyko, & Lichtman, 2010).
If the test is positive, what is the probability that the
140 160
disease is present? If the test is negative, what is the
probability that the disease is not present? When you
talk to the patient about the results of their tests, how
sure are you that they do or do not have the disease?
Sensitivity and specificity are the terms used to
describe the accuracy of a screening or diagnostic test
(Table 16-5). There are four possible outcomes of a
screening test for a disease: (1) true positive, which
accurately identifies the presence of a disease; (2) false
positive, which indicates a disease is present when it
is not; (3) true negative, which indicates accurately
that a disease is not present; or (4) false negative,
which indicates that a disease is not present when it is
(Campo et al., 2010; Grove, 2007). The 2 × 2 contin-
gency table shown in Table 16-5 should help you to
visualize sensitivity and specificity and these four out-
comes (Craig & Smyth, 2012; Sackett et al., 2000).
Sensitivity and specificity can be calculated based
on research findings and clinical practice outcomes to
determine the most accurate diagnostic or screening
tool to use in identifying the presence or absence of a
disease for a population of patients. The calculations
for sensitivity and specificity are provided as follows:

TABLE 16-5 Results of Sensitivity and Specificity of Screening Tests

Diagnostic Test Result Disease Present Disease Not Present or Absent Total
Positive test a (true positive) b (false positive) a+b
Negative test c (false negative) d (true negative) c+d
Total a+c b+d a+b+c+d

From Grove, S. K. (2007). Statistics for health care research: A practical workbook. Philadelphia, PA: Saunders, p. 335.
a = The number of people who have the disease and the test is positive (true positive).
b = The number of people who do not have the disease and the test is positive (false positive).
c = The number of people who have the disease and the test is negative (false negative).
d = The number of people who do not have the disease and the test is negative (true negative).
 CHAPTER 16  Measurement Concepts 407

Sensitivity calculation = probability of disease
= a/(a + c)
= truee positive rate
Specificity calculation = probability of disease
= d/(b + d)
= truee negative rate
Sensitivity is the proportion of patients with the
disease who have a positive test result or true positive.
The ways the researcher or clinician might refer to the
test sensitivity include the following:
• Highly sensitive test is very good at identifying the
patient with a disease.
• If a test is highly sensitive, it has a low percentage
of false negatives.
• Low sensitivity test is limited in identifying the
patient with a disease.
• If a test has low sensitivity, it has a high percentage
of false negatives.
• If a sensitive test has negative results, the patient
is less likely to have the disease.
• Use the acronym SnNout: High sensitivity (Sn), test
is negative (N), rules the disease out (out) (Campo
et al., 2010; Grove, 2007).
Specificity of a screening or diagnostic test is the
proportion of patients without the disease who have a
negative test result or true negative. The ways the
researcher or clinician might refer to the test specific-
ity include the following:
• Highly specific test is very good at identifying
patients without a disease.
• If a test is very specific, it has a low percentage of
false positives.
• Low specificity test is limited in identifying patients
without a disease.
• If a test has low specificity, it has a high percentage
of false positives.
• If a specific test has positive results, the patient is
more likely to have the disease.
• Use the acronym SpPin: High specificity (Sp), test
is positive (P), rules the disease in (in).
Sarikaya, Aktas, Ay, Cetin, and Celikmen (2010)
conducted a study to determine the sensitivity and
specificity of rapid antigen diagnostic testing (RADT)
for diagnosing pharyngitis in patients in the emer-
gency department. Acute pharyngitis is primarily a
viral infection, but in 10% of the cases it is caused by
bacteria. Most cases of bacterial pharyngitis are caused
by group A beta-hemolytic streptococci (GABHS).
One laboratory method for diagnosing GABHS is
RADT, which has become more popular than a throat
culture because it can be processed rapidly during an
emergency department and primary care visit.
“We conducted a study to define the sensitivity and
specificity of RADT, using throat culture results as the
gold standard, in 100 emergency department patients
who presented with symptoms consistent with strep-
tococal pharyngitis. We found that RADT had a sen-
sitivity of 68.2% (15 of 22), a specificity of 89.7% (70
of 78), a positive predictive value of 65.2% (15 of 23),
and a negative predictive value of 90.9% (70 of 77).
We conclude that RADT is useful in the emergency
department when the clinical suspicion is GABHS, but
results should be confirmed with a throat culture in
patients whose RADT results are negative.” (Sarikaya
et al., 2010, p. 180)
The results of the study by Sarikaya et al. (2010)
were put into Table 16-6 so that you might see how
the sensitivity and specificity were calculated in this
study.
Sensitivity calculation = probability of disease
= a/ (a + c) = true positive rate
Sensitivity = probability of GABHS pharyngitis
= 15/ (15 + 7) = 15 / 22 = 68.18% = 68.2%

TABLE 16-6 Results of Sensitivity and Specificity of Rapid Antigen Diagnostic Testing (RADT)
RADT Result GABHS Disease Present GABHS Disease Absent Total
Positive test a (true positive) = 15 b (false positive) = 8 a + b = 15 + 8 = 23
Negative test c (false negative) = 7 d (true negative) = 70 c + d = 7 + 70 = 77
Total a + c = 15 + 7 = 22 b + d = 8 + 70 = 78 a + b + c + d = 100

GABHS, Group A beta-hemolytic streptococci.

a = The number of people who have GABHS pharyngitis disease and the test is positive (true positive).
b = The number of people who do not have GABHS pharyngitis disease and the test is positive (false positive).
c = The number of people who have GABHS pharyngitis disease and the test is negative (false negative).
d = The number of people who do not have GABHS pharyngitis disease and the test is negative (true negative).
408 UNIT TWO  The Research Process
Specificity calculation = probability of disease
= d/ ( b + d) = true negative rate
Specificity = probability no GABHS pharyngitis
= 70 / (8 + 70) = 70 / 78 = 89.74% = 89.7%
The sensitivity of 68.2% indicates the percentage
of patients with a positive RADT who had GABHS
pharyngitis (true positive rate). The specificity of
89.7% indicates the percentage of patients with a
negative RADT who did not have GABHS pharyngitis
(true negative rate). In developing a diagnostic or
screening test, researchers need to achieve the highest
sensitivity and specificity possible. In selecting screen-
ing tests to diagnose illnesses, clinicians need to deter-
mine the most sensitive and specific screening test but
also need to examine cost and ease of access to these
tests in making their final decision (Craig & Smyth,
2012; Grove, 2007; Sackett et al., 2000).
Likelihood Ratios
Likelihood ratios (LRs) are additional calculations
that can help researchers to determine the accuracy of
diagnostic or screening tests, which are based on the
sensitivity and specificity results. LRs are calculated
to determine the likelihood that a positive test result
is a true positive and a negative test result is a true
negative. The ratio of the true positive results to false
positive results is known as the positive LR (Campo
et al., 2010). The positive LR is calculated as follows
using the data from the study by Sarikaya et al. (2010):
Positive LR = sensitivity ÷ 100% - specificity
Positive LR for GABHS pharyngitis
= 68.2% ÷ 100% − 89.7% = 68.2% ÷ 10.3% = 6.62
The negative LR is the ratio of true negative
results to false negative results, and it is calculated
as follows:
Negative LR = 100% − sensitivity ÷ specificity
Negative LR for GABHS pharyngitis
= 100% − 68.2% ÷ 89.7% = 31.8% ÷ 89.7% = 0.35
The very high LRs (or LRs that are >10) rule in the
disease or indicate that the patient has the disease. The
very low LRs (or LRs that are <0.1) virtually rule out
the chance that the patient has the disease (Campo
et al., 2010; Craig & Smyth, 2012; Melnyk & Fineout-
Overholt, 2011; Sackett et al., 2000). Understanding
sensitivity, specificity, and LR increases your ability
to read clinical studies and to determine the most
accurate diagnostic test to use in research and clinical
practice.
KEY POINTS
• Measurement is the process of assigning numbers
to objects, events, or situations in accord with some
rule.
• Instrumentation is the application of specific rules
to develop a measurement device or instrument.
• Measurement theory and the rules within this
theory have been developed to direct the measure-
ment of abstract and concrete concepts.
• There is direct measurement and indirect measure-
ment.
• Healthcare technology has made researchers famil-
iar with direct measures of concrete elements, such
as height, weight, heart rate, temperature, and
blood pressure.
• Indirect measurement is used with abstract con-
cepts, when the concepts are not measured directly,
but when the indicators or attributes of the concepts
are used to represent the abstraction. Common
abstract concepts measured in nursing include
anxiety, stress, coping, quality of life, and pain.
• Measurement error is the difference between what
exists in reality and what is measured by a research
instrument.
• The levels of measurement, from lower to higher,
are nominal, ordinal, interval, and ratio.
• Reliability refers to how consistently the measure-
ment technique measures the concept of interest
and includes stability reliability, equivalence reli-
ability, and internal consistency.
• The validity of an instrument is determined by
the extent to which the instrument actually
reflects the abstract construct being examined and
includes such types as face and content validity,
validity from factor analysis, validity from struc-
tural analysis, convergent validity, divergent valid-
ity, validity from contrasting groups, validity from
discriminant analysis, validity from prediction of
future and concurrent events, and successive veri-
fication validity.
• Evaluation of physiological measures requires a
different perspective from that of behavioral mea-
sures and requires evaluation for accuracy, preci-
sion, and error.
• The accuracy of screening or diagnostic tests is
determined by calculating the sensitivity, specific-
ity, and likelihood ratios for the test.

REFERENCES
American Psychological Association’s Committee to Develop Stan-
dards. (1999). Standards for educational and psychological
testing. Washington, DC: American Psychological Association.
Armstrong, G. D. (1981). Parametric statistics and ordinal data: A
pervasive misconception. Nursing Research, 30(1), 60–62.
Bannigan, K., & Watson, R. (2009). Reliability and validity in a
nutshell. Journal of Clinical Nursing, 18(23), 3237–3243.
Bartlett, J. W., & Frost, C. (2008). Reliability, repeatability and
reproducibility: Analysis of measurement errors in continuous
variables. Ultrasound Obstetric Gynecology, 31(4), 466–475.
Berk, R. A. (1990). Importance of expert judgment in content-
related validity evidence. Western Journal of Nursing Research,
12(5), 659–671.
Bialocerkowski, A., Klupp, N., & Bragge, P. (2010). Research
methodology series: How to read and critically appraise a reli-
ability article. International Journal of Therapy & Rehabilitation,
17(3), 114–120.
Bland, J. M., & Altman, D. G. (1986). Statistical methods for assess-
ing agreement between two methods of clinical measurement.
Lancet, 1(8476), 307–310.
Bliley, A. V., & Ferrans, C. E. (1993). Quality of life after coronary
angioplasty. Heart & Lung, 22(3), 193–199.
Brinberg, D., & McGrath, J. E. (1985). Validity and the research
process. Beverly Hills, CA: Sage.
Campbell, D. T, & Fiske, D. W. (1959). Convergent and discrimi-
nant validation by the multitrait-multimethod matrix. Psychologi-
cal Bulletin, 56(2), 81–105.
Campo, M., Shiyko, M. P., & Lichtman, S. W. (2010). Sensitivity
and specificity: A review of related statistics and controversies in
the context of physical therapist education. Journal of Physical
Therapy Education, 24(3), 69–78.
Clinical and Laboratory Standards Institute (CLSI). (2011). Harmo-
nized terminology database. Retrieved from http://www.clsi.org/
Content/NavigationMenu/Resources/HarmonizedTerminology
Database/Harmonized_Terminolo.htm.
Cohen, J. A. (1960). A coefficient of agreement for nominal scales.
Education & Psychological Measurement, 20(1), 37–46.
Corty, E. W. (2007). Using and interpreting statistics: A practical
text for the health, behavioral, and social sciences. St. Louis, MO:
Mosby.
Craig, J. V., & Smyth, R. L. (2012). The evidence-base practice
manual for nurses (3rd ed.). Edinburgh, Scotland: Churchill
Livingstone.
Creswell, J. W. (2009). Research design: Qualitative, quantitative,
and mixed methods approaches (3rd ed.). Los Angeles, CA: Sage.
Davis, L. L. (1992). Instrument review: Getting the most from a
panel of experts. Applied Nursing Research, 5(4), 194–197.
DeKeyser, F. G, & Pugh, L. C. (1990). Assessment of the reliability
and validity of biochemical measures. Nursing Research, 39(5),
314–317.
DeVellis, R. F. (1991). Scale development: Theory and applications.
Newbury Park, CA: Sage.
DeVon, H. A., Block, M. E., Moyle-Wright, P., Ernst, D. M.,
Hayden, S. J., Lazzara, D. J., et al. (2007). A psychometric toolbox
for testing validity and reliability. Journal of Nursing Scholarship,
39(2), 155–164.
CHAPTER 16  Measurement Concepts 409
Dickerson, S. S., Kennedy, M., Wu, Y. B., Underhill, M., & Othman,
A. (2010). Factors related to quality-of-life pattern changes in
recipients of implantable defibrillators. Heart & Lung, 39(6),
466–476.
Doran, D. M. (2011). Nursing outcomes: The state of the science
(2nd ed.). Sudbury, MA: Jones & Bartlett Learning.
Fawcett, J., & Garity, J. (2009). Evaluating research for evidence-
based nursing practice. Philadelphia: F.A. Davis.
Flesch, R. (1984). A new readability yardstick. Journal of Applied
Psychology, 32, 221–233.
French, S. E., Lenton, R., Walters, V., & Eyles, J. (2000). An empiri-
cal evaluation of an expanded nursing stress scale. Journal of
Nursing Measurement, 8(2), 161–178.
Gift, A. G., & Soeken, K. L. (1988). Assessment of physiologic
instruments. Heart & Lung, 17(2), 128–133.
Goodwin, L. D., & Goodwin, W. L. (1991). Estimating construct
validity. Research in Nursing & Health, 14(3), 235–243.
Grove, S. K. (2007). Statistics for health care research: A practical
workbook. Philadelphia, PA: Saunders.
Gunning, R., & Kallan, R. A. (1994). How to take the fog out of
business writing. Chicago, IL: Dartnell.
Hagerty, B. M. K., & Patusky, K. (1995). Developing a measure of
sense of belonging. Nursing Research, 44(1), 9–13.
Hockenberry, M. J., & Wilson, D. (2009). Wong’s essentials of
pediatric nursing (8th ed.). St. Louis, MO: Mosby.
Imle, M. A., & Atwood, J. R. (1988). Retaining qualitative validity
while gaining reliability and validity: Development of the Transi-
tion to Parenthood Concerns Scale. Advanced Nursing Science,
11(1), 61–75.
International Organization for Standardization (ISO). (2011a). Stan-
dards development. Retrieved from http://www.iso.org/iso/
standards_development.htm.
International Organization for Standardization (ISO). (2011b). Dis-
cover IOS: What standards do. Retrieved from http://www.iso.org/
iso/about/discover-iso_what-standards-do.htm.
Johnson, M. J. (2002). The development and testing of three
medication-taking questionnaires for the medication adherence
model constructs for hypertensive patients. Unpublished doctoral
dissertation. University of Utah, Salt Lake City, UT.
Johnson, M. J., & Rogers, S. (2006). Development of the Purposeful
Action Medication-Taking Questionnaire. Western Journal of
Nursing Research, 28(3), 335–351.
Johnson, M. J., Williams, M., & Marshall, E. S. (1999). Adherent
and nonadherent medication-taking elderly hypertensive patients.
Clinical Nursing Research, 8(4), 318–335.
Joint Committee for Guides in Metrology. (2011). JCGM: Joint
Committee for Guides in Metrology. Retrieved from http://
www.iso.org/sites/JCGM/GUM-introduction.htm.
Kaplan, A. (1963). The conduct of inquiry: Methodology for behav-
ioral science. New York, NY: Harper & Row.
Kerlinger, F. N., & Lee, H. B. (2000). Foundations of
behavioral research (4th ed.) Fort Worth, TX: Harcourt College
Publishers.
Knapp, T. R. (1990). Treating ordinal scales as interval scales: An
attempt to resolve the controversy. Nursing Research, 39(2),
121–123.
Lawshe, C. H. (1975). A quantitative approach to content validity.
Personnel Psychology, 28, 563–575.
410 UNIT TWO  The Research Process
Lehane, E., & McCarthy, G. (2007). An examination of intentional
and unintentional aspects of medication non-adherence in patients
diagnosed with hypertension. Journal of Clinical Nursing, 16(4),
698–706.
Lynn, M. R. (1986). Determination and quantification of content
validity. Nursing Research, 35(6), 382–385.
Melnyk, B. M., & Fineout-Overholt, E. (2011). Evidence-based
practice in nursing & healthcare: A guide to best practice (2nd
ed.). Philadelphia, PA: Lippincott Williams & Wilkins.
Miller, B., & Bodie, M. (1994). Determination of reading compre-
hension level for effective patient health-education materials.
Nursing Research, 43(2), 118–119.
Munro, B. H. (2001). Statistical methods for health care research
(4th ed.). Philadelphia, PA: Lippincott Williams & Wilkins.
Munro, B. H. (2005). Statistical methods for health care research
(5th ed.). Philadelphia, PA: Lippincott Williams & Wilkins.
Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric theory (3rd
ed.). New York, NY: McGraw-Hill.
Radloff, L. S. (1977). The CES-D scale: A self-report depression
scale for research in the general population. Applied Psychologi-
cal Measures, 1, 385–394.
Rasin, J. H. (1997). Measurement issues with the elderly. In
M. Frank-Stromberg & S. J. Olsen (Eds.), Instruments for clinical
health-care research (2nd ed., pp. 44–53). Boston, MA: Jones &
Bartlett.
Ryan-Wenger, N. A. (2010). Evaluation of measurement precision,
accuracy, and error in biophysical data for clinical research and
practice. In C. F. Waltz, O. L. Strickland, & E. R. Lenz (Eds.),
Measurement in nursing and health research (4th ed.) (pp. 371–
383). New York, NY: Springer Publishing Company.
Sackett, D. L., Straus, S. E., Richardson, W. S., Rosenberg, W., &
Haynes, R. B. (2000). Evidence-based medicine: How to practice
and teach EBM (2nd ed.). Edinburgh: Churchill Livingstone.
Sarikaya, S., Aktas, C., Ay, D., Cetin, A., & Celikmen, F. (2010).
Sensitivity and specificity of rapid antigen detection testing for
diagnosing pharyngitis in emergency department. Ear Nose &
Throat Journal, 89(4), 180–182.
Sax, G. (1997). Principles of educational and psychological mea-
surement and evaluation (4th ed.). Belmont, CA: Wadsworth
Publishing Company.
Schell, K., Briening, E., Lebet, R., Pruden, K., Rawheiser, S., &
Jackson, B. (2011). Comparison of arm and calf automatic non-
invasive blood pressures in pediatric intensive care patients. Journal
of Pediatric Nursing, 26(1), 3–12.
Silvaroli, N. J. (1986). Classroom reading inventory (5th ed.).
Dubuque, IA: William C. Brown.
Spielberger, C. D., Gorsuch, R. L., & Lushene, P. R. (1970). Manual
for the State-Trait Anxiety Inventory (Form Y). Palo Alto, CA:
Consulting Psychologists Press.
Stevens, S. S. (1946). On the theory of scales of measurement.
Science, 103, 677–680.
Stommel, M., Wang, S., Given, C. W., & Given, B. (1992). Confir-
matory factor analysis (CFA) as a method to assess measurement
equivalence. Research in Nursing & Health, 15(5), 399–405.
Streiner, D. L., & Norman, G. R. (1995). Health measurement
scales: A practical guide to their development and use (2nd ed.).
Oxford, UK: Oxford University Press.
Thomas, S. (1992). Face validity. Western Journal of Nursing
Research, 14(1), 109–112.
Topf, M. (1988). Interrater reliability decline under covert assess-
ment. Nursing Research, 37(1), 47–49.
Waltz, C. W., & Bausell, R. B. (1981). Nursing research: Design,
statistics and computer analysis. Philadelphia: F. A. Davis.
Waltz, C. F., Strickland, O. L., & Lenz, E. R. (2010). Measurement
in nursing and health research (4th ed.). New York, NY: Springer
Publishing Company.
White, W. B., Berson, A. S., Robbins, C., Jamieson, M. J., Prisant,
L. M., Roccella, E., et al. (1993). National standard for measure-
ment of resting and ambulatory blood pressures with automated
sphygmomanometers. Hypertension, 21(4), 504–509.
   http://evolve.elsevier.com/Grove/practice/
17
Measurement Methods Used in
Developing Evidence-Based Practice
N
ursing research examines a wide variety of
phenomena, requiring an extensive array of
measurement methods. However, nurse re-
searchers have sometimes found limited instruments
available to measure phenomena central to the studies
essential for generating evidence-based practice
(EBP). Thus, for the last 30 years, nurse researchers
have made it a priority to develop valid and reliable
instruments to measure phenomena of concern to
nursing. As a result, the number and quality of mea-
surement methods have greatly increased (DeVon
et al., 2007; Waltz, Strickland, & Lenz, 2010).
Knowledge of measurement methods is important
to all aspects of nursing. To perform a critical
appraisal of a study, nurses need knowledge of mea-
surement theory and an understanding of the state of
the art for developing instruments to examine the
phenomena under study. For example, when evaluat-
ing someone else’s research, you might want to know
whether the researcher was using an older tool that
has been surpassed by more precise and accurate
physiological measures. It might help you to know
that measuring a particular phenomenon has been a
problem with which nurse researchers have struggled
for many years. Your understanding of the successes
and struggles in measuring nursing phenomena may
stimulate your creative thinking and lead you to con-
tribute your own research to the development of mea-
surement approaches. Some nursing phenomena have
not been adequately examined because reliable and
valid instruments are not available to measure them,
which makes it difficult for nurse researchers to gen-
erate the essential evidence needed for practice
(Brown, 2009; Craig & Smyth, 2012; Melnyk &
Fineout-Overholt, 2011).
This chapter describes the common measurement
approaches used in nursing research, including physi-
ological measures, observations, interviews, question-
naires, and scales. Other methods of measurement
CHAPTER
discussed include Q-sort methodology, the Delphi
technique, diaries, and use of existing databases. The
chapter also describes the process for locating existing
instruments, determining their reliability and validity,
and assessing their readability. Directions are pro-
vided for describing an instrument in a written report.
The chapter concludes with a description of the
process of scale construction and issues related to
translating an instrument into another language.
Physiological Measurement
Much of nursing practice is oriented toward physio-
logical dimensions of health. Therefore, many of our
questions require us to be able to measure these di-
mensions. Of particular importance are studies linking
physiological, psychological, and social variables.
The need for physiological research reached national
attention in 1993 when the National Institute of
Nursing Research (NINR) recommended an increase
in physiologically-based nursing studies because 85%
of NINR-funded studies involved nonphysiological
variables. According to NINR staff, a review of physi-
ological studies funded by the NINR found that “the
biological measurements used in the funded grants
often were not state-of-the-science, and the biological
theory underlying the measurements often was unde-
rutilized” (Cowan, Heinrich, Lucas, Sigmon, & Hin­
shaw, 1993, p. 4). Cowan et al. proposed a 10-year
plan to enhance the education of nurse researchers in
physiological measurement, expand the number and
quality of physiological studies conducted, and in-
crease the funding for physiological research. Rudy
and Grady (2005) noted in their small study of funded
researchers (N = 31) that nursing is building a group
of nurse scientists who are committing their research
careers to studying various biological and pathological
phenomena. Over the last 15 years, nurse researchers
have expanded their use and development of precise
411
412 UNIT TWO  The Research Process
and accurate physiological measures. An example is
the current research taking place in genetics, which
was encouraged by Grady, Director of the NINR, with
the implementation of the Summer Genetics Institute
(SGI) to expand the conduct of genomic research
(NINR, 2012).
The 2011 Strategic Plan for NINR emphasized the
conduct of biological research to provide a foundation
for understanding and managing diseases and to test
preventive care and self-management strategies. NINR
(2011) proposed to invest in research to “[i]mprove
quality of life by managing symptoms of chronic
illness,” which will require expansion in the number of
biologically based studies and the quality of physiolog-
ical measurements used in these studies (see the NINR
most current Mission and Strategic Plan document at
http://www.ninr.nih.gov/AboutNINR/NINRMission
andStrategicPlan/). The increased number of biological
researchers and the expanded funding for biological
research have increased the quality and quantity of
physiological measures used in nursing studies.
Physiological measures include two categories,
biophysical and biochemical. Biophysical measures
might include the use of the stethoscope and sphyg-
momanometer to measure blood pressure, and a bio-
chemical measure might include the laboratory value
for total cholesterol. Physiological measures can be
acquired in a variety of ways from instruments within
the body (in vivo), such as a reading from an arterial
line, or from application of an instrument on the outside
of a subject (in vitro), such as a blood pressure cuff
(Stone & Frazier, 2010). The following sections
describe how to obtain physiological measures by self-
report, observation, direct or indirect measurement,
laboratory tests, electronic monitoring, and the creative
development of new instruments. The measurement of
physiological variables across time is also addressed.
This section concludes with a discussion of how to
select physiological measures for a particular study.
Obtaining Physiological Measures
by Self-Report
Self-report has been used effectively to obtain physi-
ological information and may be particularly useful
when the subjects are not in closely monitored settings
such as hospitals, clinics, or research facilities. Phe-
nomena that have been or could be measured by self-
report include hours of sleep, patterns of daily
activities, eating patterns, dieting patterns, stool fre-
quency and consistency, patterns of joint stiffness,
variations in degree of mobility, and exercise patterns.
For some variables, self-report may be the only means
of obtaining the information. Such may be the case
when study participants experience a physiological
phenomenon that cannot be observed or measured
by others. Nonobservable physiological phenomena
include pain, nausea, dizziness, indigestion, patterns
of hunger or thirst, hot flashes, tinnitus, pruritus,
fatigue, malaise, and dyspnea (DeVon et al., 2007;
Waltz et al., 2010).
Bhengu et al. (2011) studied the physiological
symptoms experienced by individuals infected with
human immunodeficiency virus (HIV) and receiving
antiretroviral therapy. These physiological symptoms
were measured using a self-report checklist completed
by the HIV-infected patients. The measurement
method was the revised Sign and Symptom Checklist
for HIV patients (SSC-HIVrev) developed by Holze-
mer et al. (1999) and revised by Holzemer, Hudson,
Kirksey, Hamilton, and Bakken (2001). The SSC-
HIVrev instrument is described in the following
excerpt.
“[The SSC-HIVrev has been widely used and] found
to be valid and reliable for measuring HIV-related
symptoms.… Respondents report the presence and
intensity of the symptoms based on the following:
‘Below is a list of potential problems related to HIV
that you may be experiencing today. If you have the
problems, please rate the degree of intensity of the
problem. If you do not have the problem, do not
check a box.’
Items are scored using the following scale: 0 = not
checked, 1 = mild, 2 = moderate, and 3 = severe.
Total symptom score is a count of the number of
symptoms checked as present on the day of complet-
ing the questionnaire, ranging from 0 to 72. A total
symptom intensity score is a weighting of symptoms
based on the 1-to-3 scale of mild, moderate, or
severe. Prior studies indicate that the six factors have
strong reliability estimates and stable factor structure
that supports the construct validity of the 26-item
instrument. Additional evidence supports the concur-
rent validity of the scale as well as its sensitivity to
change over time. The final version of the SSC-HIVrev
(Parts I and II; Holzemer et al., 2001) used in this
study is a 26-item scale available for use by clinicians
and researchers to measure the patient’s self-report
of HIV-related signs and symptoms. In this study,
alpha reliabilities ranged from .77 to .91 (malaise/
fatigue = .91, confusion/distress/pain = .89, fever/
chills = .83, gastrointestinal discomfort = .85, short-
ness of breath = .80, and nausea/vomiting = .77).”
(Bhengu et al., 2011, pp. 4-5)
 CHAPTER 17  Measurement Methods Used in Developing Evidence-Based Practice 413
The self-report SSC-HIVrev scale was refined by
Holzemer et al. (1999, 2001) over the years and has
documented reliability and validity. The SSC-HIVrev
has demonstrated construct validity through factor
analysis, with six factors being identified, and concur-
rent validity. The scale has strong reliability in previ-
ous studies and in this study for the total scale and the
six subscales (alpha reliabilities ranging from 0.77 to
0.91). Using self-report measures may enable nurses
to study research questions that were not previously
considered, which could be an important means to
build knowledge in areas not yet explored. The insight
gained could alter the way nurses manage patient situ-
ations that are now considered problematic and
improve patient outcomes (Doran, 2011). However,
self-report is a subjective way to measure physiologi-
cal variables, and studies are strengthened by having
both subjective and objective measurements of physi-
ological variables.
Obtaining Physiological Measures
by Observation
Researchers sometimes obtain data on physiological
parameters by using observational data collection
measures. These measures provide criteria for quan-
tifying various levels or states of physiological func-
tioning. In addition to collecting clinical data, this
method provides a means to gather data from the
observations of caregivers. This source of data has
been particularly useful in studies involving criti-
cally ill patients in intensive care units (ICUs) and
patients with Alzheimer’s disease, advanced cancer,
and severe mental illness. Observation is also an
effective way to gather data on frail elderly adults,
infants, and young children. Studies involving home
health agencies and hospices often use observation
tools to record physiological dimensions of patient
status. These data are sometimes stored electroni-
cally and are available to researchers for large data-
base analysis. Measuring physiological variables
using observation requires a quality tool for data col-
lection and consistent use of this tool by data collec-
tors. If the observations in a study are being
conducted using multiple data collectors, it is essen-
tial that the consistency or interrater reliability of the
data collectors be determined (see Chapter 16)
(Bialocerkowski, Klupp, & Bragge, 2010; DeVon
et al., 2007; Waltz et al., 2010).
Klein, Dumpe, Katz, and Bena (2010) developed a
Nonverbal Pain Assessment Tool (NPAT) to measure
the pain experience by nonverbal adult patients in the
ICU. Testing of the tool occurred in three phases that
focused on the internal reliability, content validity, and
criterion validity of the tool and the interrater reli-
ability of the data collectors. The following excerpt
describes development of the NPAT and its demon-
strated reliability and validity.
“Content validity examines the extent of the tool’s
ability to measure the construct under consideration
(in this study, pain). Construction of the scale began
with an in-depth review of the literature to determine
commonly accepted signs and behaviors of pain.
Three nurse experts, including 2 clinical nurse special-
ists and a nurse from the Pain Management Service,
reviewed the tool and selected behaviors.
Criterion-related validity compares the new tool to
a ‘gold standard.’… We hypothesized that a significant
correlation would be found between the NPAT score
and the patient’s self-report of pain, the ‘gold stan-
dard’ for pain assessment.” (Klein et al., 2010, p. 523)
“The internal reliability for the entire scale was .82
(Cronbach’s alpha).… Subscale internal reliability
scores comprised: emotion, .77; movement, .78;
verbal, .79; facial, .77; and position, .78.… To deter-
mine the interrater reliability of the revised NPAT, a
convenience sampling included all patients more than
16 years old and admitted to any of the 4 ICUs during
the data collection period. The same teams of nurses
were used. Data were collected for 50 patients,
although data from only 39 patients were useable.
The concordance correlation coefficient was .72
(95% confidence interval), demonstrating strong
interrater reliability.… The criterion validity of the
revised NPAT was again tested.… The concordance
correlation coefficient was .66 (95% confidence
interval), indicating moderate to strong validity.”
(Klein et al., 2010, pp. 525-526)
Klein et al. (2010) found the NPAT had strong
internal reliability for both the total scale (Cron-
bach’s alpha = 0.82) and the subscales (Cronbach’s
alpha ranging from 0.77 to 0.79). Because the NPAT
is a new tool, these researchers described the content
and criterion validity of the tool and recognized the
need for additional research to determine the reliabil-
ity and validity of the tool with different samples.
The final copy of this tool is presented later in this
chapter.
Obtaining Physiological Measures Directly
or Indirectly
Physiological variables can be measured either directly
or indirectly. Direct measures are more accurate
414 UNIT TWO  The Research Process
because there is an objective measurement of the study
variable. For example, patients might be asked to
report any irregular heartbeats during waking hours
over a 24-hour period, which is an indirect measure-
ment of heart rhythm, and each patient’s heart could
be monitored with a Holter monitor over the same
24-hour time frame (direct measure of heart rhythm).
Whenever possible, researchers usually select direct
measures of study variables because of the accuracy
and precision of these measurement methods.
However, if a direct measurement method does not
exist, an indirect measurement method could be used
in the initial investigation of a physiological variable.
Sometimes researchers use both direct and indirect
measurement methods to expand the understanding of
a physiological variable. Dubbert, White, Grothe,
O’Jile, and Kirchner (2006) studied the physical activ-
ity of patients who are severely mentally ill. These
researchers measured the variable physical activity
with indirect and direct measurement methods that are
described in the following excerpt.
“Self-Reported Physical Activity
[Indirect Measurement]
“The 42-item Community Health Activities Model
Program for Seniors (CHAMPS) … was used to
assess frequency and duration of a variety of physical
activities for the previous 4 weeks. The CHAMPS
yields estimates of kilocalorie (kcal) energy expendi-
ture per unit time and physical activity frequency. In
nonclinical samples, CHAMPS scores have test-retest
reliability intraclass correlations (ICCS) R = 0.76 for
moderate and R = 0.66 for total estimated kcal expen-
diture over 6 months and validity correlations in the
R = 0.20 to 0.30 range with performance on a
6-minute walk test.…
“Objectively Measured Physical Activity
[Direct Measurement]
“Participants wore RT3 (Stayhealthy, Inc., Monrovia,
CA) accelerometers to obtain objective estimates of
daily physical activity. The RT3 instrument, about the
size of a pager, measures acceleration of movement
along three axes, which was averaged into a compos-
ite score (i.e., vector movement [VM]), representing
the overall magnitude of activity for each minute. RT3
software transformed VM into an estimate of energy
expenditure (i.e., kcal per day and per hour), using
participant’s height, weight, age, and gender.”
(Dubbert et al., 2006, p. 206)
Obtaining Physiological Measures from
Laboratory Tests
Laboratory tests are usually very precise and accurate
and provide direct measures of many physiological
variables. Biochemical measures, such as total choles-
terol, triglycerides, hemoglobin, and hematocrit, must
be obtained through invasive procedures. Sometimes
these invasive procedures are part of routine patient
care, and researchers, with institutional review board
(IRB) approval, can obtain the results from the
patient’s record. Although nurses are now performing
some biochemical measures in the nursing unit, these
measures often require laboratory analysis. When
invasive procedures are not part of routine care but are
instead performed specifically for a study, great care
must be taken to protect the subjects and to follow
guidelines for informed consent and IRB approval.
Neither the patients nor their insurers can be billed for
invasive procedures that are not part of routine care;
thus, the researcher must seek external funding or the
institution in which the patient is receiving care must
agree to forego billing for the procedure.
Researchers need to ensure the accuracy and preci-
sion of laboratory measures and the methods of col-
lecting specimens for their studies. The laboratory
performing the analyses needs to be certified and in
compliance with national standards developed by the
Clinical and Laboratory Standards Institute (CLSI,
2011). The data collectors need to be trained to ensure
that intrarater reliability and interrater reliability are
maintained during the data collection process (see
Chapter 16) (Bialocerkowski et al., 2010; Waltz et al.,
2010). Smith, Annesi, Walsh, Lennon, and Bell (2010)
examined the effects of a behavioral treatment on vol-
untary physical activity, self-efficacy, and risk factors
for type 2 diabetes in obese preadolescents 10 to 14
years old. The risk factors for diabetes measured in
this study included lipid values and glucose/insulin
ratio that were determined before and after the 12-week
behavioral treatment. The following excerpt describes
the blood analyses that were conducted in this study.
“Blood Analyses
“Total cholesterol, low-density lipoprotein (LDL)
cholesterol, and high-density lipoprotein (HDL) cho-
lesterol are measures of lipids in the blood. Each is
expressed as milligram per deciliter. For ages less than
18 years, normal ranges are 125 to 170 mg/dl, less
than 110 mg/dl, and 38 to 76 mg/dl, respectively.
Generally, lower values on total cholesterol and LDL
 CHAPTER 17  Measurement Methods Used in Developing Evidence-Based Practice 415
cholesterol and a higher value on HDL cholesterol are
preferable. Glucose/insulin ratio is a measure of meta-
bolic functioning and expressed as a ratio of glucose
(mg/dl) to insulin (mcIU/ml). The target value is 7.0
or above (Silfen et al., 2001). Generally, a higher score
is preferable. Blood was drawn while participants
were in a fasting state and analyzed by Quest Diag-
nostics (Madison, NJ) at no cost to participants’ fami-
lies.” (Smith et al., 2010, p. 395)
These researchers clearly described the blood anal-
yses performed in their study and the normal values
for preadolescents. To promote precision and accuracy
in the lipid values and the glucose/insulin ratios
obtained, the participants were instructed to fast and
the blood was analyzed in a certified laboratory (Quest
Diagnostics). The blood was drawn in a physician’s
office and transferred to the laboratory for analysis.
The study report would have been strengthened by a
discussion of the data collection process in the physi-
cian’s office to ensure that the blood specimens were
consistently collected and managed in the delivery to
the laboratory.
Obtaining Physiological Measures through
Electronic Monitoring
The availability of electronic monitoring equipment
has greatly increased the possibilities of physiological
measurement in nursing studies, particularly in critical
care environments. Understanding the processes
of electronic monitoring can make the procedure
less formidable to individuals critically appraising
published studies and individuals considering using
the method for measurement.
To use electronic monitoring, usually sensors are
placed on or within study participants. The sensors
Signal
Sensing conditioning
equipment equipment
Transducer Amplifier
Stimulus Subject
Signal
Electrodes
processor
Recording, data processing,
and transmission equipment
measure changes in body functions such as electrical
energy. Figure 17-1 shows the process of electronic
measurement. Many sensors need an external stimulus
to trigger the measurement process. Transducers
convert the electrical signal to numerical data. Electri-
cal signals often include interference signals as well as
the desired signal, so you may choose to use an ampli-
fier to decrease interference and amplify the desired
signal. The electrical signal is digitized (converted to
numerical digits or values) and stored in a computer.
In addition, it is immediately displayed on a monitor.
The display equipment may be visual or auditory or
both. One type of display equipment is an oscilloscope
that displays the data as a waveform; it may provide
information such as time, phase, voltage, or frequency
of the target event or behavior. The final phase is
the recording, data processing, and transmission that
might be done through computer, camera, graphic
recorder, or magnetic tape recorder (Stone & Frazier,
2010). A graphic recorder provides a printed version
of the data. Some electronic equipment simultane-
ously records multiple physiological measures that are
displayed on a monitor. The equipment is often linked
to a computer or might be wireless, which allows the
researcher to review the data. The computer often con-
tains complex software for detailed analysis of the data
and provides a printed report of the analysis results
(Pugh & DeKeyser, 1995; Stone & Frazier, 2010).
The advantages of using electronic monitoring
equipment are the collection of accurate and precise
data, recording of data accurately within a computer-
ized system, potential for collection of large amounts
of data frequently over time, and transmission of data
electronically for analysis. One disadvantage of using
certain sensors to measure physiological variables is
that the presence of a transducer within the body can
alter the reading. For example, the presence of a flow
Figure 17-1  Process of electronic measurement.
Display
equipment
416 UNIT TWO  The Research Process
A reading and the RFID number approximately every
B on the computer was used to configure the computer
Figure 17-2  A, OR wireless thermometer. The disc has an elliptical
cross section, and the sensing element consists of a metal strip located
at the center of the skin-contact side. B, ThermoSENSOR. The device
has been placed over the first piece of hypoallergenic adhesive film
dressing on the lower abdomen and is about to be secured to the lower
abdomen by a second piece of the same dressing. (From Ng, K., Wong,
S., Lim, S., & Goh, Z. [2010]. Evaluation of the Cadi ThermoSENSOR
wireless skin-contact thermometer against ear and axillary temperatures
in children. Journal of Pediatric Nursing, 25[3], 177.)
transducer in a blood vessel can partially block the
vessel and alter blood flow resulting in an inaccurate
reflection of the flow (Ryan-Wenger, 2010).
Ng, Wong, Lim, and Goh (2010) compared the
Cadi ThermoSENSOR wireless skin-contact ther-
mometer (Figure 17-2) readings with the ear and axil-
lary temperatures in children. The ThermoSENSOR
thermometer provides a continuous measurement of
body temperature and transmits the readings wire-
lessly to a central server. The measurement with the
ThermoSENSOR thermometer is described in the fol-
lowing excerpt.
“Developed by Cadi Scientific in Singapore as part of
an integrated wireless system for temperature moni-
toring and location tracking, this system uses a reus-
able skin-contact thermometer or sensor called the
ThermoSENSOR. This thermometer takes the form
of a small disc that can be easily adhered to the
patient’s skin, and each disc is assigned a unique radio
frequency identification (RFID) number [see Figure
17-2]. The thermometer measures body tempera-
ture continuously and transmits a temperature
30 seconds to a computer or server through one or
more signal receivers (nodes) installed in the vicinity
of the patient [Figure 17-3].” (Ng et al., 2010, pp.
176-177)
“Before the study, a ThermoSENSOR wireless
temperature monitoring system was installed in the
ward. A wireless signal receiver (node) was installed
in the ceiling of each of the five-bedded rooms.…
These receivers were connected to the hospital’s
local area network (LAN). The system worked in
such a way that temperature readings and RFID
numbers transmitted by a sensor were received by
one or more wireless receivers in the vicinity of the
sensor and transferred through the LAN to a personal
computer.… Web-based application software
designed for use with the wireless system and installed
to receive, store, and display the temperature and
RFID data. A total of 32 sensors were used for the
study.
The ThermoSENSOR uses a thermistor as the
sensing element. When in use, the sensor is attached
to the patient using a two-layer dressing system that
prevents the sensor from coming in direct contact
with the skin [see Figure 17-2]. The sensor is water
resistant and can be cleaned by immersing it in a
cleaning and disinfectant solution. The manufacturer
provided the following specifications for the sensor:
operating ambient temperature range, 10° C to 50°
C; thermistor accuracy, ± 0.2° C for temperature
range of 32.0° C to 42.0° C; data transmission rate,
every 30 seconds on average; radio frequency,
868.4 MHz; typical transmission range, 10 m
(unblocked); power source, internal 3-V lithium coin-
cell battery; battery life, 12 months (continuous oper-
ation); dimensions, diameter of 36 mm, height of
11.6 mm; weight, 10 g without battery; applicable
radio equipment standards, ETSI 300 220, ETSI EN
301 489.” (Ng et al., 2010, pp 177-178)
 CHAPTER 17  Measurement Methods Used in Developing Evidence-Based Practice 417

Wireless signal receiver

Ward 1

Ethernet hub
Temperature sensor

Central server

Wireless signal receiver

Ward 2

Monitoring and e-mail Monitoring and e-mail

alert via Tablet PC Monitoring SMS alert via alert via desktop computer
connected to LAN via PDA mobile phone connected to LAN

Figure 17-3  Setup of the ThermoSENSOR wireless temperature monitoring system. Each sensor transmits data wirelessly to a signal receiver (node)
that is within the prescribed transmission range. The signal receiver uploads the data to a central server through the local area network (LAN), through
which the data can be accessed from computers and other devices that are connected, wirelessly or by wired means, to the LAN. The server can be
configured to send out e-mail and short message service (SMS) alerts. (From Ng, K., Wong, S., Lim, S., & Goh, Z. [2010]. Evaluation of the Cadi
ThermoSENSOR wireless skin-contact thermometer against ear and axillary temperatures in children. Journal of Pediatric Nursing, 25[3], 177.)

Ng et al. (2010) provided detailed descriptions and
pictures of both the ThermoSENSOR thermometer
and the wireless setup. The thermometer was consis-
tently applied to the abdomen of each child and was
cleaned in a precise way. The manufacturer specifica-
tions of the thermometer documented that it was an
accurate device to measure temperature. The wireless
system was described in detail with documentation of
its precision and accuracy in obtaining and transfer-
ring the children’s temperatures to a computer for
recording, display, and analysis of the data. The find-
ings of the study indicated that the ThermoSENSOR
wireless skin-contact thermometer readings were
comparable to the ear and axillary temperature
readings.
Genetic Advancements in Measuring
Nucleic Acids
The Human Genome Project has greatly expanded the
understanding of deoxyribonucleic acid (DNA) that
contains the code for controlling human development.
The U.S. Human Genome Project was begun in 1990
by the Department of Energy and the National Insti-
tutes of Health and was completed in 2003. The
genome is the entire DNA sequence in an organism,
including its genes. The genes carry information for
making all the proteins required by the organism that
are used to determine how the body looks, functions,
and behaves. The DNA is a double-stranded helix and
serves as the code for the production of the single-
stranded messenger ribonucleic acid (RNA) (Stone &
Frazier, 2010).
“The project goals were to:
• Identify all the approximately 20,000-25,000 genes
in human DNA,
• Determine the sequences of the 3 billion chemical
base pairs that make up human DNA,
• Store this information in databases,
• Improve tools for data analysis,
• Transfer related technologies to the private
sector, and
• Address the ethical, legal, and social issues (ELSI)
that may arise from the project” (Department of
Education Genomic Science, 2011)
The advancements in genetics have facilitated the
development of new technologies that have permitted
the analysis of normal and abnormal genes for the
detection and diagnosis of genetic diseases. Through
the use of molecular cloning, sufficient quantities of
DNA and RNA have been produced to permit analysis
in research. The Southern blotting technique is the
standard way for analyzing the structure of DNA. The
Northern blotting technique is used for RNA analysis.
Analyses of both normal and mutant genes are of
interest, and the Western blotting technique is used to
examine the mutant proteins in cells obtained from
418 UNIT TWO  The Research Process
patients with diseases. In addition, polymerase chain
reaction can selectively amplify DNA and RNA mol-
ecules for study (Stone & Frazier, 2010).
It is important that nurses be aware of the advances
in technologies to measure nucleic acids and use them
in their programs of research. Nurses are becoming
more aware of the conduct of genetic research through
doctoral and postdoctoral programs specialized in this
area. In addition, the NINR provides the SGI to expand
researchers’ expertise in conducting genetic research.
You can access information on the SGI at the follow-
ing NINR (2012) website: http://www.ninr.nih.gov/
Training/TrainingOpportunitiesIntramural/Summer
GeneticsInstitute/. These educational opportunities
have expanded genetic research in nursing and
increased the number of studies focused on the mea-
surement of nucleic acids by nurses.
Jones, Munro, Grap, Kitten, and Edmond (2010)
conducted a study to determine the impact of oral care
on the bacteremia risk in mechanically ventilated
adults. The researchers wanted to determine if tooth
brushing induced transient bacteremia in this group of
patients. They used DNA typing to identify organisms
for the blood and oral cultures collected before and 1
minute and 30 minutes after tooth brushing. Their
measurement of oral microbial organisms is presented
in the following excerpt.
“Oral microbial culture. A swab of the oral cavity
for microbial culture was performed immediately pre-
ceding the first tooth brushing intervention. The oral
cavity was swabbed in the following order using a
single swab: upper and lower buccal and lingual gingival
margin (obtaining organisms from the gum line and
tooth surface), and palate. The oral microbial cultures
were performed using BBL Culture Swab Plus collec-
tion and transport media (Becton, Dickinson and Co,
Sparks, MD) and were analyzed using standard opera-
tions for the clinical microbiology laboratory. Cultures
were analyzed and quantified for the following poten-
tially pathogenic organisms: viridians group Strepto-
cocci, Staphylococcus aureus, Pseudomonas aeruginosa,
Enterococcus spp., Klebsiella pneumoniae, and Candida
spp. These organisms are most commonly cited as
causes of bloodstream infections in mechanically ven-
tilated patients. Positive cultures were frozen and
stored for comparison with blood culture organisms
by DNA typing. We prospectively planned the micro-
bial analysis using multi-locus sequence typing to iden-
tify species at the strain level. Multi-locus sequence
typing is a relatively new and powerful technique that
involves molecular comparison of collections of essen-
tial genes also referred to as the ‘housekeeping’
genes.… Comparison of DNA sequences from iso-
lates found in blood cultures and oral cultures would
enhance the determination of whether the isolates
were identical or different, and differentiate transient
bacteremia from intravenous line or sample contami-
nation from bacteremia of oral origin. DNA typing
would reduce the likelihood that confounding vari-
ables in the ICU (e.g., the presence of invasive lines,
frequent and invasive procedures, intubation, comor-
bidities, and immunosuppression) would adversely
affect the analysis.” (Jones et al., 2010, p. S59)
Through the use of DNA typing, Jones et al. (2010)
were able to measure very precisely bacteremia risk
associated with oral care in mechanically ventilated
patients. By comparison of DNA sequences in the
blood and oral cultures, other confounding potential
causes of bacteremia could be eliminated. The
researchers found that tooth brushing did not induce
transient bacteremia in the patient population. This
research provides a basis for future research focused
on standardizing effective and safe oral care in this
population.
Developing New Physiological Measures
Some studies require imaginative approaches to mea-
suring phenomena that are traditionally observed in
clinical practice but are not measured. The first step in
this process is to recognize that the phenomenon being
observed by the nurse can be measured. Once that idea
has emerged, one can begin envisioning various means
of measuring the phenomenon. As new physiological
measurements are developed, they must be compared
with previous methods to determine the best strategy
for measuring each physiological outcome based on the
patient’s condition. Gelinas et al. (2010) developed a
new approach for detecting pain in adults by measuring
the cerebral regional oxygen saturation (rSO2) using
near-infrared spectroscopy (NIRS). The NIRS tech-
nique (IN-VOS-4100 system; Somanetics, Troy, MI)
was used to measure pain during the nociceptive pro-
cedures, such as intravenous and arterial line inser-
tions, sternal bone incision, and thorax opening, during
cardiac surgery. The rSO2 measurements were com-
pared with the scores from the Critical-Care Pain
Observation Tool (CPOT) and the faces pain thermom-
eter (FPT) to determine discriminant and criterion
validity. These physiological measures used in this
study are presented in the following excerpt.
 CHAPTER 17  Measurement Methods Used in Developing Evidence-Based Practice 419

“Measurement Instruments .001, using the 5-point descriptive pain scale) and
discriminant (t = −5.10, p ≤ .001, comparing patients’
“Near-infrared spectroscopy. Near-infrared spec-
pain intensity at rest and during turning) validity in
troscopy technology is based on the property of near-
cardiac-surgery patients, showing an association with
infrared light to diffuse throughout biological tissue.
a higher pain intensity score during turning.” (Gelinas
At near-infrared wavelengths, hemoglobin and cyto-
et al., 2010, p. 488)
chrome c-oxydase, also known as the enzyme cyto-
chrome aa3, are the main chromophores (i.e.,
substances absorbing light at a given wave-length).
The light source of the oximeter provides two con-
These researchers detailed the use of the NIRS
tinuous wavelengths of near-infrared light (730 and
technique to measure rSO2, and this measurement of
810 nm) in the frontal region.… The ratio of oxygen-
pain was comparable to the CPOT and the FPT. Com-
ated hemoglobin and total hemoglobin is measured,
parison of the rSO2 with the CPOT and FPT added to
and a subtraction of the superficial signal from the
the criterion, convergent, and discriminant validity of
deeper signal is performed by the monitor to obtain
the measurement methods. Gelinas et al. (2010, p.
the regional hemoglobin oxygen saturation (rSO2) in
485) concluded, “Although further research is needed
the frontal cortex, i.e., the main variable in this
in critically ill adult patients undergoing more painful
study.… Brain-activity imaging was associated with
procedures, the NIRS may become a promising tech-
changes in cerebral oxygenation indicators as mea-
nique for assessing pain.”
sured with the NIRS, supporting the validity of this Obtaining Physiological Measures
noninvasive technique. The NIRS system used in the
present study was the INVOS-4100. It continuously
across Time
monitors rSO2, and was used according to the manu-
Many nursing studies using physiological measure-
facturer’s instructions.
ment methods focused on a single point in time. Thus,
“Critical-Care Pain Observation Tool. The
there is insufficient information on normal variations
CPOT was a key pain measure in this study, because
in physiological measures across time and much less
it was used to examine criterion validity of rSO2. This
information on changes in physiological measures
behavioral pain scale includes 4 behaviors (facial
across time in individuals with abnormal physiologi-
expression, body movements, muscle tension, and
cal states. In some cases, physiological states exhibit
vocalization), with a possible total score ranging from
cyclic activity and are associated with circadian
0 to 8. It was developed and validated in different ICU
rhythms and day-night patterns. An important ques-
groups.… The CPOT showed good interrater reli-
tion to ask is “How labile is the measure?” Some
ability (ICC = .80 to .93), discriminant validity (signifi-
measures vary within the individual from time to time,
cantly higher CPOT scores during procedural pain),
even when conditions are similar. Circadian rhythms,
and criterion validity. During procedural pain (turning),
activities, emotions, dietary intake, or posture can also
correlation coefficients of .59 and .71 (p ≤ .05) were
affect physiological measures. Researchers need to
obtained between patients’ self-reports of pain inten-
determine to what extent these factors would affect the
sity and their CPOT scores, whereas correlation
ability to interpret measurement outcomes. When a
coefficients of .49 and .40 (p ≤ .05) were evident
clinician observes variation in a physiological value,
when patients were at rest before and after turning.
it is important to know whether the variation is within
Finally, a sensitivity of 86% and a specificity of 78%
the normal range or signals a change in the patient’s
were obtained for a cutoff score >2 in the presence
condition. Thus, additional studies need to be con-
of pain on the CPOT during nociceptive exposure of
ducted to describe patterns of physiological function
postoperative ICU adults.
over time.
“Faces pain thermometer. The faces pain ther-
In the previously discussed study using NIRS,
mometer (FPT) consists of a thermometer graded
Gelinas et al. (2010) conducted a repeated-measures
from 0 to 10, including 6 faces.… The FPT was
design that involved obtaining measurements from 40
another key pain measure in this study, because it
subjects at two test periods. The first test period
provided a standardized indicator of the patient’s self-
occurred while patients were awake, and the rSO2,
report of pain intensity, which was also used to deter-
CPOT, and FPT measurements were recorded for each
mine the criterion validity of rSO2. This scale
subject. The second test period took place after the
demonstrated good convergent (r = .80 to .86, p ≤
induction of anesthesia, and it was possible to record
only the rSO2. These simultaneous measurements of
420 UNIT TWO  The Research Process
pain using various instruments as well as the repeated
measures of the rSO2 when patients were anesthetized
allowed the examination of reliability and validity of
the NIRS technique to measure rSO2 as a pain indica-
tor over time.
Selecting a Physiological Measure
Researchers designing a physiological study have less
assistance in selecting methods of measurement than
researchers conducting studies using psychosocial
variables. Multiple books and electronic sources are
available that discuss various methods for measuring
psychosocial variables. In addition, numerous articles
in nursing journals describe the development of psy-
chosocial variables or discuss various means of mea-
suring a particular psychosocial variable. However,
literature guiding the selection of physiological vari-
ables is still sparse. You might consider the following
factors when selecting a physiological measure for a
study:
1. What physiological variables are relevant to the
study?
2. Do the variables need to be measured continu-
ously or at a particular point in time?
3. Are repeated measures needed?
4. Do certain characteristics of the population under
study place limits on the measurement approaches
that can be used?
5. How has the variable been measured in previous
research?
6. Is more than one measurement method available
to measure the physiological variable being
studied (Stone & Frazier, 2010)?
7. Which measurement method is the most accurate
and precise for the population you are studying
(Fawcett & Garity, 2009; Ryan-Wenger, 2010)?
8. Could the study be designed to include more than
one measurement method for the variable being
studied (DeVon et al., 2007; Fawcett & Garity,
2009)?
9. Where can the measurement device or devices be
obtained to measure the physiological variable
being studied?
10. Can the measurement device be obtained from the
manufacturer for use in the study, or must it be
purchased?
11. What are the standards for the measurement
device or equipment that has been designated
nationally and internationally (International Orga-
nization for Standardization [IOS], 2011)?
It is more difficult to identify previous research on
physiological measures than it is to find research on
psychosocial measures. The sources most commonly
used to identify physiological measurement methods
are previous studies that have measured a particular
physiological variable. Literature reviews or meta-
analyses can provide reference lists of relevant
studies. Because the measure might have been used
in studies unrelated to the current research topic, it is
usually important to examine the research literature
broadly.
Physiological measures must be linked conceptu-
ally with the framework of the study. The logic of
operationalizing the concept in a particular way
must be well thought out and expressed clearly (see
Chapter 7). It is often a good idea to use diverse
physiological measures of a single concept, which
reduces the impact of extraneous variables that might
affect measurement. The operationalization of a
physiological variable in a study should clearly indi-
cate the physiological measure to be used. The link
of the physiological variable to the concept in the
framework must be made explicit in the published
report of your study.
You also need to evaluate the accuracy and preci-
sion of physiological measures. Until more recently,
researchers commonly used information from the
equipment manufacturer to describe the accuracy of
measurement. This information is useful, but it is
insufficient to evaluate accuracy and precision. The
accuracy and precision of physiological measures are
discussed in Chapter 16 (CLSI, 2011; IOS, 2011;
Ryan-Wenger, 2010).
You need to consider problems you might encoun-
ter when using various approaches to physiological
measurement. One factor of concern is the sensitivity
of the measure. Will the measure detect differences
finely enough to avoid a type II error—known as a
false negative—that occurs when the investigator
claims there is no difference between groups or rela-
tionships among variables when one really exists (see
Chapter 21)? Physiological measures are usually norm
referenced. Data obtained from a study participant are
compared with a norm as well as with other partici-
pants. You need to determine whether the norm used
for comparison is relevant for the population you are
studying. Laboratories are certified by ensuring that
the analyses conducted in the laboratory meet a
national standard (CLSI, 2011). New physiological
measures are compared with the “gold standard” or
the current best measurement method for a physiologi-
cal variable. For example, Klein et al. (2010) com-
pared their NPAT with the “gold standard” of the
patients’ self-reports of pain using the CPOT.
Many measurement strategies require the use of
specialized equipment. In many cases, the equipment
 CHAPTER 17  Measurement Methods Used in Developing Evidence-Based Practice 421
is available in the patient care area and is part
of routine patient care in that unit. Otherwise, the
researcher may need to purchase, rent, or borrow the
equipment specifically for the study. You need to be
skilled in operating the equipment or obtain the assis-
tance of someone who has these skills. You need to
ensure that the equipment is operated in an optimal
fashion and is used in a consistent manner. Sometimes
equipment must be recalibrated, or reset, regularly to
ensure consistent readings. For example, weight
scales are recalibrated periodically to ensure that the
weight indicated is accurate and precise. According
to federal guidelines, recalibration must be performed
as follows:
• In accordance with the manufacturers’ instructions
• In accordance with national and international stan-
dards (IOS, 2011)
• In accordance with criteria set up by the laboratory
(CLSI, 2011)
• At least every 6 months
• After major preventive maintenance or replace-
ment of a critical part
• When quality control indicates a need for recalibra-
tion
Reporting Physiological Measures
in Studies
When publishing the results of a physiological study,
researchers must describe the measurement technique
in considerable detail to allow an adequate critical
appraisal of the study, enable others to replicate the
study, and promote clinical application of the results.
At the present time, only a few physiological replica-
tion studies have been reported in the nursing litera-
ture. A detailed description of physiological measures
in a research report includes the following:
1. Description of the equipment or device used in
performing the measurement
2. Identification of the name of the equipment
manufacturer
3. Account of the accuracy and precision of the equip-
ment or device based on previous research, the
manufacturers’ specifications, and national and
international standards
4. Explanation of the exact procedure followed to
measure the physiological variable
5. Overview of the process the device used to record,
retrieve, and store data
The examples discussed in this section can be used
as models for describing the methods for obtaining
and implementing physiological measures to obtain
accurate and precise measures of physiological vari-
ables to ensure quality study outcomes.
Observational Measurement
Observational measurement is the use of unstruc-
tured and structured inspection to gauge a study vari-
able. This section focuses on structured observational
measurement; unstructured observational measure-
ment is described in Chapter 12. Although data col-
lection by observation is most common in qualitative
research, it is used to some extent in all types of
studies (Marshall & Rossman, 2011; Munhall, 2012).
First, you must decide what you want to observe, and
then you need to determine how to ensure that every
variable is observed in a similar manner in each
instance. Much attention must be given to training
data collectors, especially when the observations are
complex and examined over time (Waltz et al., 2010).
You must create opportunities for the observational
technique to be pilot-tested and to generate data on
interrater reliability. Observational measurement tends
to be more subjective than other types of measurement
and is often seen as less credible. However, in many
cases, observation is the only possible way to obtain
important evidence for practice.
Structured Observations
The first step in a structured observation is to define
carefully what specific behaviors or events are to be
inspected or observed in a study. From that point,
researchers determine how the observations are to be
made, recorded, and coded. In most cases, the research
team develops an observational checklist or category
system to direct collecting, organizing, and sorting of
the specific behaviors or events being observed. The
extent to which these categories are exhaustive varies
with the study.
Category Systems
The observational categories should be mutually
exclusive. If the categories overlap, the observer will
be faced with making judgments regarding which cat-
egory should contain each observed behavior, and data
collection and recording may be inconsistent. In some
category systems, only the behavior that is of interest
is recorded. Most category systems require the
observer to make some inference from the observed
event to the category. The greater the degree of infer-
ence required, the more difficult the category system
is to use. Some systems are applicable in a wide
variety of studies, whereas others are specific to the
study for which they were designed. The number of
categories used varies considerably with the study. An
optimal number for ease of use and therefore effec-
tiveness of observation is 15 to 20 categories.
422 UNIT TWO  The Research Process
Klein et al. (2010) developed the NPAT that was
introduced earlier in this chapter. The NPAT included
categories of behaviors that were to be observed to
determine the pain level for nonverbal adults in the
ICU (see Figure 17-4). The interrater reliability of the
tool in this study was ensured when “Two RNs, trained
in the use and scoring of the NPAT, simultaneously
observed a patient unable to verbalize his or her pain”
(Klein et al., 2010, p. 523).
Another type of category system used to direct the
collection of observational data is a checklist. Obser-
vational checklists are techniques used to establish
whether a behavior occurred. The observer places a
tally mark on a data collection form each time he or
she witnesses the behavior. Behavior other than that
on the checklist is ignored. In some studies, the
observer may place multiple tally marks in various
categories while witnessing a particular event.
However, in other studies, the observer is required
to select a single category in which to place the
tally mark.
Rating Scales
Rating scales (discussed in detail later in this chapter)
can be used for observation and for self-reporting. A
rating scale allows the observer to rate the behavior or
event on a scale. This method provides more informa-
tion for analysis than the use of dichotomous data,
which indicate only that the behavior either occurred
or did not occur. The NPAT also included a rating scale
in which each observational category was scored on a
scale or 0 to 2 or 0 to 3 (see Figure 17-4). The tool
resulted in a total score between 0 and 10, with 0
indicating no pain and 10 indicating the worst pain
ever experienced by the patient (Klein et al., 2010).
Interviews
Interviews involve verbal communication during
which the subject provides information to the
researcher. Although this measurement strategy is
most common in qualitative and descriptive studies, it
also can be used in other types of studies. The various
approaches to conducting interviews range from very
unstructured interviews in which study participants
control content (see Chapter 12) to interviews in
which the participants respond to a questionnaire that
the researcher has carefully designed (Waltz et al.,
2010). Although most interviews are conducted face
to face, telephone interviews are also commonly used.
Using the interview method for measurement
requires careful detailed work with a scientific
approach. Excellent books are available on the
techniques of developing interview questions (Briggs,
1986; Dillman, Smyth, & Christian, 2009; Dillon,
1990; Foddy, 1993; Fowler, 1990; Gorden, 1987,
1998; McLaughlin, 1990; Mishler, 1986). If you plan
to use this strategy, consult a text on interview meth-
odology before designing your instrument. Because
nurses frequently use interview techniques in nursing
assessment, the dynamics of interviewing are familiar;
however, using this technique for measurement in
research requires greater sophistication.
Structured Interviews
Structured interviews are verbal interactions with
subjects that allow the researcher to exercise increas-
ing amounts of control over the content of the inter-
view to obtain essential data for a study. The researcher
designs the questions before data collection begins,
and the order of the questions is specified. In some
cases, the interviewer is allowed to explain the
meaning of the question further or modify the way in
which the question is asked so that the subject can
understand it better. In more structured interviews, the
interviewer is required to ask the question precisely as
it has been designed. If the subject does not under-
stand the question, the interviewer can repeat it only.
The subject may be limited to a range of responses
previously developed by the researcher, similar to
those in a questionnaire. If the possible responses are
lengthy or complex, they may be printed on a card so
that study participants can review them visually before
selecting a response.
Designing Interview Questions
The process for developing and sequencing interview
questions is similar to the process used to design ques-
tionnaires and is explained in the section on question-
naires. Briefly, questions progress from broad and
general to narrow and specific. Questions are grouped
by topic, with fairly “safe” topics being addressed first
and sensitive topics reserved until late in the interview
process. Other data such as age, educational level,
income, and other demographic information are
usually collected last. These data are best obtained
from other sources, such as patient records, to allow
more time for the primary interview questions. The
wording of questions in an interview depends on the
educational level of the study participants. Different
participants may interpret the wording of certain ques-
tions in a variety of ways, and researchers need to
anticipate this possibility. After the interview protocol
has been developed, it is wise to seek feedback from
an expert on interview technique and from a content
expert.
 CHAPTER 17  Measurement Methods Used in Developing Evidence-Based Practice 423

Is patient able to make vocalizations or sound cues?

Score under the yes or no column; add scores for total score (range 0-10)

YES NO
SCORE SCORE
EMOTION An affective EMOTION
response to a situation
0 0
Smiling; calm; relaxed or none due to coma state or analgesia
Anxious; irritable; withdrawn; closes eyes;
1 1
does not engage with physical environment

2 Tearful/crying or uncooperative 2

MOVEMENT Change in placement and MOVEMENT

positioning of the body and
extremities when not
engaged in any care activities
0 0
None; sleeping comfortable; no unusual movements;
or none due to coma state or analgesia
Restless or slow, decreased movement;
1 2
reluctant to move; muscle tenseness
Rigidity; increasing motion; stiffening; tossing; turning;
2 3
flapping of arms; stiffening
VERBAL CUES Sound cues or vocalizations other than speech
0 No vocalization
n/a
1 Whimpering; moaning; sighing
2 Screaming; crying out
FACIAL CUES Expressions on face FACIAL CUES
0 Relaxed, calm expression or none due to coma state or analgesia 0
1 Drawn around the mouth and eyes; narrowed eyes 1

2 Wincing; grimacing; clenched teeth; furrowed brows; tightened lips 2

POSITIONING/GUARDING POSITIONING/GUARDING

Body responses that imply a protection of

the body from contact with external touch

0 Relaxed body or none due to coma state or analgesia 0

1 Guarding/tense 2

Jumpy when touched; clutching of siderails;

2 3
withdraws when touched

TOTAL
Choose only one behavior per category

Figure 17-4  Nonverbal Pain Assessment Tool—Final. (From Klein, D. G., Dumpe, M., Katz, E., & Bena, J. [2010]. Pain assessment in the intensive care
unit: Development and psychometric testing of the nonverbal pain assessment tool. Heart & Lung, 39[6], 527.)
424 UNIT TWO  The Research Process
Pilot-Testing the Interview Protocol
After the research team has satisfactorily developed
the interview protocol, team members need to pretest
or pilot-test it on subjects similar to the individuals
who will be included in their study. Pilot-testing
allows the research team to identify problems in the
design of questions, sequencing of questions, or pro-
cedure for recording responses. It also provides an
opportunity to assess the reliability and validity of the
interview instrument (Waltz et al., 2010).
Training Interviewers
Skilled interviewing requires practice, and interview-
ers must be familiar with the content of the interview.
They need to anticipate situations that might occur
during the interview and develop strategies for dealing
with them. One of the most effective methods of
developing a polished approach is role-playing.
Playing the role of the subject can give the interviewer
insight into the experience and facilitate an effective
response to unscripted situations.
The interviewer must establish a permissive atmo-
sphere in which the subject is encouraged to respond
to sensitive topics. He or she also needs to develop an
unbiased verbal and nonverbal manner. The wording
of a question, the tone of voice, a raised eyebrow, and
a shifting body position all can communicate a posi-
tive or negative reaction to the subject’s responses—
either of which can alter the data.
Preparing for an Interview
If you are serving as the interviewer and expect the
meeting to be lengthy, you need to make an appoint-
ment. Dress nicely for the meeting, but do not over-
dress, and be prompt. Choose a site for the interview
that is quiet and private and provides a pleasant envi-
ronment. Before the appointment, carefully plan the
instructions you will give to the subject. For example,
you might say, “I am going to ask you a series of
questions about …. Before you answer each question
you need to …. Select your answer from the following
…, and then you may elaborate on your response. I
will record your answer and then, if it is not clear, I
may ask you to explain some aspect further.”
Probing
Interviewers use probing to obtain more information
in a specific area of the interview. In some cases, you
may have to repeat a question. If your subject answers,
“I don’t know,” you may have to press for a response.
In other situations, you may have to explain the ques-
tion further or ask the subject to explain statements
that he or she has made. At a deeper level, you may
pick up on a comment the participant made and begin
asking questions to understand better what the subject
meant. Probes should be neutral to avoid biasing
participants’ responses. Probing for additional infor-
mation needs to be done within reasonable guidelines
so that participants do not feel they are being
cross-examined.
Recording Interview Data
Data obtained from interviews are recorded, either
during the interview or immediately afterward. The
recording may be in the form of handwritten notes,
video recordings, or audio recordings. If you hand-
record your notes, you must have the skill to identify
key ideas (or capture essential data) in an interview
and concisely record this information. Data must be
recorded without distracting the interviewee. Some
interviewees have difficulty responding if it is obvious
that the interviewer is taking notes or recording the
conversation. In such a case, the interviewer may need
to record data after completing the interview. If you
wish to record the interview, you first must obtain IRB
approval and then obtain the participant’s permission.
Plan to prepare verbatim transcriptions of the record-
ings before data analysis. In some studies, researchers
use content analysis to capture the meaning within the
data (see Chapter 12).
Advantages and Disadvantages
of Interviews
Interviewing is a flexible technique that can allow
researchers to explore greater depth of meaning than
they can obtain with other techniques. Use your inter-
personal skills to encourage your subject’s coopera-
tion and elicit more information. The response rate to
interviews is higher than the response rate to question-
naires; thus interviews can offer a more representative
sample. Interviews allow researchers to collect data
from participants who are unable or unlikely to com-
plete questionnaires, such as very ill subjects or sub-
jects whose reading, writing, and ability to express
themselves are marginal. Interviews are a form of self-
report, and the researcher must assume that the infor-
mation provided is accurate. Interviewing requires
much more time than questionnaires and scales and is
more costly. Because of time and cost, sample size is
usually limited. Subject bias is always a threat to
the validity of the findings, as is inconsistency in
data collection from one subject to another (Dillman
et al., 2009).
Interviewing children requires a special under-
standing of the art of asking children questions. The
interviewer must use words that children tend to use
 CHAPTER 17  Measurement Methods Used in Developing Evidence-Based Practice 425
to define situations and events. Interviewers also must
be familiar with the language skills that exist at dif-
ferent stages of development. Children view topics
differently than adults do. Their perception of time,
past and present, is also different. Holaday and Turner-
Henson (1989) provided detailed suggestions for
developing an interview guide or questionnaire appro-
priate for children.
Some researchers use a combination of structured
interview questions and open-ended or unstructured
interview questions to gather the data needed for a
study. Harralson (2007, p. 96) used structured and
unstructured questions to examine the “factors associ-
ated with delay in seeking emergency medical atten-
tion for acute ischemic symptoms in a sample of
predominantly African American women.” Harralson
interviewed female patients who presented with symp-
toms of acute myocardial infarction in a large, urban
teaching hospital in the United States. The following
excerpt describes the interview process used in this
study.
“Structured Interview
“The study used a structured interview to explore the
variables of interest. The 45-minute interview
included questions pertaining to sociodemographics,
social support, general physical health, medical
comorbidities, perceived and practical barriers to
seeking health care, and CHD [coronary heart
disease] symptoms and severity. Open-ended ques-
tions addressed the patients’ experiences from
symptom onset until a decision was made to seek
medical attention. Open-ended questions included
questions about patients’ physical and emotional feel-
ings during the experience, decision-making pro-
cesses (i.e., who they told and who they sought
advice from), and beliefs about what was happening
to them at the onset of the symptoms of AMI [acute
myocardial infarction].
The structured interview was developed specifi-
cally for this study on the basis of a systematic review
of the literature that examined concepts and factors
associated with delay in seeking medical treatment.
In addition, several nurses, cardiologists, and social
scientists reviewed the interview.
To reduce recall bias in this study, interviews were
conducted within 5 days of the acute ischemic event.
This recall time frame is similar to time periods used
in other studies reviewed in the background section.”
(Harralson, 2007, p. 98)
Questionnaires
A questionnaire is a printed self-report form designed
to elicit information that can be obtained from a sub-
ject’s written responses. The information derived
through questionnaires is similar to information
obtained by interview, but the questions tend to have
less depth. The subject is unable to elaborate on
responses or ask for questions to be clarified, and the
data collector cannot use probe strategies. However,
questions are presented in a consistent manner, and
there is less opportunity for bias than in an
interview.
Questionnaires can be designed to determine facts
about the study participants or persons known by the
participants; facts about events or situations known by
the participants; or beliefs, attitudes, opinions, levels
of knowledge, or intentions of the participants. Ques-
tionnaires can be distributed to large samples directly
or indirectly through email or mail. The design, devel-
opment, and administration of questionnaires have
been addressed in many excellent books that focus on
survey techniques (Berdie, Anderson, & Niebuhr,
1992; Converse & Presser, 1986; Saris & Gallhofer,
2007; Thomas, 2004). Two nursing methodology texts
(Shelley, 1984; Waltz et al., 2010) provide detailed
explanations of the procedure for questionnaire
development.
Although questions on a questionnaire appear easy
to design, a well-designed item requires considerable
effort. Similar to interviews, questionnaires can have
varying degrees of structure. Some questionnaires ask
open-ended questions that require written responses.
Others ask closed-ended questions with options that
the researcher has selected. Data from open-ended
questions are often difficult to interpret, and content
analysis may be used to extract meaning. Open-ended
questionnaire items are not advised if your data are
being obtained from large samples.
Researchers frequently use computers to gather
questionnaire data (Saris, 1991; Thomas, 2004). Com­
puters are sometimes set up at the data collection
site, such as a clinic or hospital; the questionnaire
is presented on screen; and subjects respond by using
the keyboard or mouse. Data are stored in a computer
file and are immediately available for analysis. Data
entry errors are greatly reduced. Most researchers
email subjects and direct them to a website where they
can complete the questionnaire online, allowing the
data to be stored and analyzed immediately. Thus,
researchers can keep track of the number of subjects
completing their questionnaire and the evolving
results.
 426 UNIT TWO  The Research Process
Development of Questionnaires
The first step in either selecting or developing a ques-
tionnaire is to identify the information desired. For
this purpose, the research team develops a blueprint
or table of specifications. The blueprint identifies the
essential content to be covered by the questionnaire;
the content must be at the educational level of the
potential subjects. It is difficult to stick to the blueprint
when designing the questionnaire because it is tempt-
ing to add “just one more question” that seems to be
a “neat idea” or a question that someone insists “really
should be included.” However, as a questionnaire
lengthens, fewer subjects are willing to respond, and
more questions are left blank.
The second step is to search the literature for ques-
tionnaires or items in questionnaires that match the
blueprint criteria. Sometimes published studies include
questionnaires, but frequently you must contact the
authors of a study to request a copy of their question-
naire. Researchers are encouraged to use questions in
exactly the same form as questionnaires in previous
studies to determine the validity of the questionnaire
for new samples. However, questions that are poorly
written need to be modified, even if rewriting makes
it more difficult to compare the validity results of the
questionnaire directly with questionnaires from previ-
ous studies.
In some cases, you may find a questionnaire in the
literature that matches the questionnaire blueprint that
you have developed for your study. However, you may
have to add items to or delete items from an existing
questionnaire to accommodate your blueprint. In some
situations, items from several questionnaires are com-
bined to develop an appropriate questionnaire.
An item on a questionnaire has two parts: a ques-
tion (or stem) and a response set. Each question must
be carefully designed and clearly expressed. Problems
include ambiguous or vague language, leading ques-
oasis-ebl|Rsalles|1476233357
tions that influence the response, questions that assume
a preexisting state of affairs, and double questions.
In some cases, respondents interpret terms used in
the question in one way when the researcher intended
a different meaning. For example, the researcher
might ask how heavy the traffic is in the neighborhood
in which the family lives. The researcher might be
asking about automobile traffic, but the respondent
interprets the question in relation to drug traffic. The
researcher might define neighborhood as a region
composed of a three-block area, whereas the respon-
dent considers a neighborhood to be a much larger
area. Family could be defined as people living in one
house or as all close blood relations. If a question
includes a term that is unfamiliar to the respondent or
for which several meanings are possible, the term
must be defined (Saris & Gallhofer, 2007; Waltz et al.,
2010).
Leading questions suggest to the respondent the
answer the researcher desires. These types of ques-
tions often include value-laden words and indicate the
researcher’s bias. For example, a researcher might ask,
“Do you believe physicians should be coddled on the
nursing unit?” or “All hospitals are stressful places to
work, aren’t they?” These examples are extreme, and
leading questions are usually constructed more subtly.
The degree of formality with which the question is
expressed and the permissive tone of the questions are,
in many cases, important for obtaining a true measure.
A permissive tone suggests that any of the possible
responses would be acceptable.
Questions implying a preexisting state of affairs
often lead respondents to admit to a previous behavior
regardless of how they answer. Examples are “How
long has it been since you used drugs?” or, to an ado-
lescent, “Do you use a condom when you have sex?”
Double questions ask for more than one bit of
information: “Do you like critical care nursing and
working closely with physicians?” It would be possi-
ble for the respondent to like working in critical care
settings but dislike working closely with physicians.
In this case, the question would be impossible to
answer accurately. A similar question is, “Was the
in-service program educational and interesting?”
Questions with double negatives are often difficult
for study participants to interpret. For example, one
might ask, “Do you believe nurses should not question
doctors’ orders? Yes or No.” In this case, the wording
of this question can be easily misinterpreted and the
word “not” possibly overlooked. This situation can
lead participants to respond in a way contrary to how
they actually think or feel.
Each item in a questionnaire has a response set that
provides the parameters within which the respondent
can answer. This response set can be open and flexible,
as it is with open-ended questions, or it can be narrow
and directive, as it is with closed-ended questions. For
example, an open-ended question might have a
response set of three blank lines. With closed-ended
questions, the response set includes a specific list of
alternatives from which to select.
Response sets can be constructed in various ways.
The cardinal rule is that every possible answer must
have a response category. If the sample includes
respondents who might not have an answer, a response
category of “don’t know” or “uncertain” should be
included. If the information sought is factual, include
“other” as one of the possible responses. However,
 CHAPTER 17  Measurement Methods Used in Developing Evidence-Based Practice 427
recognize that the item “other” is essentially lost data.
Even if the response is followed by a statement such
as “Please explain,” it is rarely possible to analyze the
data meaningfully. If a large number of study partici-
pants (>10%) select the alternative “other,” the
alternatives included in the response set might not
be appropriate for the population studied (Saris &
Gallhofer, 2007).
The simplest response set is the dichotomous yes/
no option. Arranging responses vertically preceded by
a blank reduces errors. For example,
____ Yes
____ No
is better than
____ Yes ____ No
because in the latter example, the respondent might
not be sure whether to indicate yes by placing a
response before or after the “Yes.”
Response sets must be mutually exclusive, which
might not be the case in the following response set
because a respondent might legitimately need to select
two responses:
____ Working full-time
____ Full-time graduate student
____ Working part-time
____ Part-time graduate student
Cazzell (2010) developed the Self-Report College
Student Risk Behavior Questionnaire, an eight-item
questionnaire with response set of yes and no possible
answers. This questionnaire was developed and refined
as part of her dissertation at The University of Texas
at Arlington. Cazzell’s questionnaire was developed
based on the 87 risk behaviors identified in a national
survey conducted by the U.S. Centers for Disease
Control and Prevention on the Youth Risk Behavior
Surveillance System (Brener et al., 2004). Cazzell’s
(2010) Self-Report College Student Risk Behavior
Questionnaire was developed based on the eight most
commonly identified adolescent risk behaviors from
the Centers for Disease Control and Prevention survey.
Content validity of this eight-item questionnaire was
developed by having two addiction experts, a doctor-
ally prepared social worker and a pediatric clinical
nurse specialist evaluate the items. The content valid-
ity index calculated for the questionnaire was 0.88,
supporting the inclusion of these eight items in the
questionnaire. Cazzell (2011) continued her develop-
ment and refinement of this questionnaire and
expanded question #2 on use of alcohol to target binge
drinking (see Figure 17-5).
The questionnaire instructions should be pilot-
tested on naive subjects who are willing and able to
express their reactions to the instructions. Each
question should clearly instruct the subject how to
respond (i.e., choose one, mark all that apply), or
instructions should be included at the beginning of the
questionnaire. The subject must know whether to
circle, underline, or fill in a circle as he or she responds
to items. Clear instructions are difficult to construct
and usually require several attempts. Cazzell (2011)
provided clear directions and an example of how to
complete her questionnaire and directed the students
to report their participation in these risk behaviors
over the past 30 days (see Figure 17-5).
After the questionnaire items have been developed,
you need to plan carefully how they will be ordered.
Questions related to a specific topic must be grouped
together. General items are included first with progres-
sion to more specific items. More important items
might be included first, with subsequent progression
to items of lesser importance. Questions of a sensitive
nature or questions that might be threatening should
appear last on the questionnaire. In some cases, the
response to one item may influence the response to
another. If so, the order of such items must be care-
fully considered. Open-ended questions should be
presented last because their responses require more
time than needed for closed-ended questions. The
general trend is to ask for demographic data about the
subject at the end of the questionnaire.
An introductory page in the computer or a cover
letter for a mailed questionnaire is needed to
explain the purpose of the study and identify the
researchers, the approximate amount of time required
to complete the form, and organizations or institutions
supporting the study. Often researchers indicate that
completion of the questionnaire implies informed
consent. A questionnaire on the computer when com-
pleted can be easily submitted, and the data can be
analyzed immediately. Returning mailed question-
naires is much more complex. The instructions need
to include an address to which the questionnaire can
be returned. This address must be at the end of the
questionnaire and on the cover letter and envelope.
Respondents often discard both the envelope and the
cover letter and do not know where to send the ques-
tionnaire after completing it. It is also wise to provide
a stamped, addressed envelope for the subject to return
the questionnaire. If possible, the best way to provide
questionnaires to potential subjects is by emailing a
web address so participants can easily complete the
questionnaire in their own time, and their responses
are automatically submitted at the end of the question-
naire. Sending questionnaires by email has many
advantages, but one disadvantage is being able to
access only individuals with email. Researchers need
428 UNIT TWO  The Research Process

Unique ID:

Self-Report College Student Risk Behavior Measure

Shade Circles Like This -->

Not Like This --> 

Answer YES or NO based on your participation in these behaviors

over the past 30 days.

1. I smoked a cigarette (even a puff). Yes No

2. I drank alcohol (even one drink). Yes No

If you answered YES to Question #2:

a. If you are a female, did you have 4 or Yes No

more drinks on one occasion?
b. If you are a male, did you have 5 or Yes No
more drinks on one occasion?

3. I used an illegal drug (even once). Yes No

4. I had sexual intercourse without a condom. Yes No

5. I rode in a car without wearing my seatbelt Yes No

(even once).

6. I drove a car without wearing my seatbelt Yes No

(even once).

7. I rode in a car with a person driving under Yes No

the influence (even once).

8. I drove a car while under the influence Yes No

(even once).

Figure 17-5  Self-Report College Student Risk Behavior Questionnaire. (From Cazzell, M. [personal communication, March 20, 2011]. Self-report college
student risk behavior questionnaire. The University of Texas at Arlington, Arlington, TX.)

to determine if the population they are studying has

email access (Thomas, 2004).
Your questionnaire must be pilot-tested to deter-
mine the clarity of questions, effectiveness of instruc-
tions, completeness of response sets, time required to
complete the questionnaire, and success of data col-
lection techniques. As with any pilot test, the subjects
and techniques must be as similar as possible to those
planned for the main study. In some cases, the open-
ended questions are included in a pilot test to obtain
information for the development of closed-ended
response sets for the main study.
Questionnaire Validity
One of the greatest risks in developing response sets
is leaving out an important alternative or response. For
example, if the questionnaire item addressed the job
position of nurses working in a hospital and the sample
included nursing students, a category must be added
to represent the student role. When seeking opinions,
there is a risk of obtaining a response from an indi-
vidual who actually has no opinion on the research
topic. When an item requests knowledge that the
respondent does not possess, the subject’s guessing
 CHAPTER 17  Measurement Methods Used in Developing Evidence-Based Practice 429
interferes with obtaining a true measure of the study
variables.
The response rate to questionnaires is generally
lower than that with other forms of self-reporting,
particularly if the questionnaires are mailed out. If the
response rate is less than 50%, the representativeness
of the sample is seriously in question. The response
rate for mailed questionnaires is usually small (25%
to 35%), so researchers are frequently unable to obtain
a representative sample, even with randomization.
There seems to be a stronger response rate for ques-
tionnaires that are sent by email, but the response is
still usually less than 50%. Strategies that can increase
the response rate for an emailed or mailed question-
naire are discussed in Chapter 20.
Study participants commonly fail to respond to
all the questions on a questionnaire. This problem,
especially with long questionnaires, can threaten the
validity of the instrument. In some cases, study par-
ticipants may write in an answer if they do not agree
with the available choices, or they might write com-
ments in the margin. Generally, these responses cannot
be included in the analysis; however, you should keep
a record of such responses. These responses might be
used later to refine the questionnaire questions and
responses.
Consistency in the way the questionnaire is admin-
istered is important to validity. Variability that could
confound the interpretation of the data reported by the
study participants is introduced by administering some
questionnaires in a group setting, mailing some ques-
tionnaires, and emailing some questionnaires. There
should not be a mix of mailing or emailing to business
addresses and to home addresses. If questionnaires are
administered in person, the administration needs to be
consistent. Several problems in consistency can occur:
(1) Some subjects may ask to take the form home to
complete it and return it later, whereas others will
complete it in the presence of the data collector; (2)
some subjects may complete the form themselves,
whereas others may ask a family member to write the
responses that the respondent dictates; and (3) in some
cases, a secretary or colleague may complete the form,
rather than the individual whose response you are
seeking. These situations may lead to biases in
responses that are unknown to the researcher and can
alter the true measure of the variables.
Analysis of Questionnaire Data
Data from questionnaires are often at the nominal or
ordinal level of measurement, which limit analyses for
the most part to descriptive statistics, such as frequen-
cies and percentages, and nonparametric inferential
statistics, such as chi square, Spearman rank-order
correlation, and Mann-Whitney U (see Chapters 22
through 25). However, in certain cases, ordinal data
from questionnaires are treated as interval data, and
t-tests and analysis of variance are used to test for
differences between responses of various subsets of
the sample (Grove, 2007). Discriminant analysis may
be used to determine the ability to predict membership
in various groups from responses to particular
questions.
Scales
Scales, a form of self-report, are a more precise means
of measuring phenomena than questionnaires. Most
scales have been developed to measure psychosocial
variables. However, self-reports can be obtained on
physiological variables such as pain, nausea, or func-
tional capacity by using scaling techniques as dis-
cussed earlier in this chapter. Scaling is based on
mathematical theory, and there is a branch of science
whose primary concern is the development of mea-
surement scales. From the point of view of scaling
theory, considerable measurement error, both random
and systematic error, is expected in a single item.
Therefore, in most scales, the various items on the
scale are summed to obtain a single score, and these
scales are referred to as summated scales. Less
random and systematic error exists when using the
total score of a scale in conducting data analyses,
although subscale comparisons are usually of interest
and are conducted. Using several items in a scale to
measure a concept is comparable to using several
instruments to measure a concept (see Figure 16-4 in
Chapter 16). The various items in a scale increase the
dimensions of the concept that are reflected in the
instrument. The types of scales commonly used in
nursing studies include rating scale, Likert scale,
semantic differential scale, and visual analogue scale
(VAS).
Rating Scale
A rating scale lists an ordered series of categories of
a variable that are assumed to be based on an underly-
ing continuum. A numerical value is assigned to each
category, and the fineness of the distinctions between
categories varies with the scale, making this one of the
crudest forms of scaling technique. The general public
commonly uses rating scales. In conversations, one
can hear statements such as “On a scale of 1 to 10, I
would rank that ….” Rating scales are easy to develop;
however, one must be careful to avoid end statements
that are so extreme that no subject would select them.
430 UNIT TWO  The Research Process
0 1
No hurt Hurts
little bit
Figure 17-6  Wong-Baker FACES Pain Rating Scale. (From Hockenberry, M. J., & Wilson, D. [2009]. Wong’s essentials of pediatric nursing [8th ed., p.
1203]. St. Louis, MO: Mosby.).
Numeric Rating Scale (NRS)
No
Pain
0 1 2 3
Figure 17-7  Numeric Rating Scale (NRS).
A rating scale could be used to rate the degree of
cooperativeness of the patient or the value placed by
the subject on nurse-patient interactions. This type of
scale is often used in observational measurement to
guide data collection. The Wong-Baker FACES Pain
Rating Scale is commonly used to assess the pain
of children in clinical practice and has been shown
to be valid and reliable over the years (Figure 17-6)
(Hockenberry & Wilson, 2009). Pain in adults is
often assessed with a numeric rating scale such as
the one presented in Figure 17-7. Klein et al. (2010)
developed the NPAT rating scale, which was intro-
duced earlier in this chapter to determine the pain level
for nonverbal adults in the ICU (see Figure 17-4).
Likert Scale
The Likert scale determines the opinion or attitude of
a subject and contains a number of declarative state-
ments with a scale after each statement. The Likert
scale is the most commonly used of the scaling tech-
niques in nursing and healthcare studies. The original
version of the scale included five response categories.
Each response category was assigned a value, with
a value of 1 given to the most negative response and
a value of 5 given to the most positive response
(Nunnally & Bernstein, 1994).
Response choices in a Likert scale most commonly
address agreement, evaluation, or frequency. Agree-
ment options may include statements such as strongly
2 3 4 5
Hurts Hurts Hurts Hurts
little more even more whole lot worst
Moderate Unbearable
Pain Pain
4 5 6 7 8 9 10
agree, agree, uncertain, disagree, and strongly dis-
agree. Evaluation responses ask the respondent for an
evaluative rating along a good/bad continuum, such as
positive to negative or excellent to poor. Frequency
responses may include statements such as never,
rarely, sometimes, frequently, and all the time. The
terms used are versatile and must be selected for their
appropriateness to the stem (Spector, 1992). Some-
times seven options are given, and sometimes only
four options are given.
Use of the uncertain or neutral category is contro-
versial because it allows the subject to avoid making
a clear choice of positive or negative statements.
Thus, sometimes only four or six options are offered,
with the uncertain category omitted. This type of
scale is referred to as a forced choice version. Some-
times respondents become annoyed at forced choice
items and refuse to complete them. Researchers who
use the forced choice version consider an item that is
left blank as a response of “uncertain.” However,
responses of “uncertain” are difficult to interpret, and
if a large number of respondents select that option
or leave the question blank, the data may be of little
value.
How the researcher phrases item stems depends on
the type of judgment that the respondent is being
asked to make. Agreement items are declarative state-
ments such as “Nurses should be held accountable for
managing a patient’s pain.” Frequency items can be
 CHAPTER 17  Measurement Methods Used in Developing Evidence-Based Practice 431
behaviors, events, or circumstances to which the
respondent can indicate how often they occur. A fre-
quency stem might be “You read research articles in
nursing journals.” An evaluation stem could be “The
effectiveness of ‘X’ drug for relief of nausea after
chemotherapy.” Items must be clear, concise, and con-
crete (Spector, 1992).
An instrument using a Likert scale usually consists
of 15 to 30 items, each addressing an element of the
concept being measured. Half the statements should
be expressed positively and half should be expressed
negatively, termed counterbalancing, to avoid insert-
ing response-set bias into the participants’ responses.
Response-set bias tends to occur when participants
anticipate that either the positive or the negative
(agree or disagree) response is consistently provided
either in the right or left hand columns of the scale.
Participants might note a pattern that agreeing with
scale items consistently falls to the right and dis-
agreeing to the left. Thus, they might fail to read all
questions carefully and just mark the right or left
column based on whether they agree or disagree with
scale items. Response-set bias can be avoided by
wording some scale items positively and other items
negatively. Participants would need to mark some
items in the right column (agree) and others in the
left column (disagree) of the scale based on their
sentiments.
Scale values of negatively worded items must be
reversed before analysis so that the participants’ agree-
ment with certain positively worded items and, accord-
ingly, their disagreement with negatively worded
items on the same scale have higher scale values or
scores reflecting their agreement. Usually, the values
obtained from each item in the instrument are summed
to obtain a single score for each subject. Although the
values of each item are technically ordinal-level data,
the summed score is often treated as interval-level
data, allowing more sophisticated parametric statisti-
cal analyses (Nunnally & Bernstein, 1994).
The Center for Epidemiological Studies Depres-
sion Scale (CES-D) is an example of a 4-point Likert
scale that is commonly used to measure depression in
nursing studies (Figure 17-8). The CES-D was devel-
oped by Radloff in 1977 and has shown to be a reliable
and valid measure of depression. Beckie, Beckstead,
Schocken, Evans, and Fletcher (2011) implemented a
tailored cardiac rehabilitation program to determine its
effect on the depressive symptoms of women with
coronary heart disease. In this randomized clinical
trial, the CES-D was used to measure depressive
symptoms in the women and is described in the fol-
lowing excerpt.
Psychological Outcome
“The 20-item Center for Epidemiological Studies
Depression Scale (CES-D) (Radloff, 1977) measured
depressive symptoms [see Figure 17-8]. Participants
reported the frequency of occurrence of depressive
symptoms during the past week ranging from 0
([Rarely or none of the time or] < a day) to 3 (Most
of the time [5-7 days]). Scores range from 0 to 60
with higher scores reflecting greater depressive
symptoms. Used extensively in CHD [coronary heart
disease] populations (Beckie, Fletcher, Beckstead,
Schocken, & Evans, 2008; Dunn, Corser, Stommel, &
Holmes-Rovner, 2006; Scholz, Knoll, Sniehotta, &
Schwarzer, 2006; Swardfager et al., 2008), a score of
16 is typically the cut-off score for an elevated level
of depressive symptoms. Evidence of concurrent
validity, construct validity, and reliability of the CES-D
has also been provided for community samples
(Radloff, 1977). Cronbach’s alpha in the current study
at all three time-points was over .90.” (Beckie et al.,
2011, p. 6)
Beckie et al. (2011) clearly described the CES-D
used to measure depression in their study. The scoring
of the scale was discussed with a score of 16 indicat-
ing elevated depressive symptoms in women with
coronary heart disease. The reliability of the scale for
this study was strong (r = .90). The discussion of the
scale would have been strengthened by expanding the
validity and reliability information from previous
research. Beckie et al. (2011) found that the gender-
tailored cardiac rehabilitation program significantly
reduced the depressive symptoms of the women com-
pared with the traditional cardiac rehabilitation
program.
Semantic Differential Scale
The semantic differential scale was developed by
Osgood, Suci, and Tannenbaum (1957) to measure
attitudes or beliefs. It is now used more broadly to
measure variations in views of a concept. A semantic
differential scale consists of two opposite adjectives
with a 7-point scale between them. The subject is to
select 1 point on the scale that best describes his or
her view of the concept being examined. The scale is
designed to measure the connotative meaning of the
concept to the subject. Although the adjectives may
not seem to be particularly related to the concept being
examined, the technique can be used to distinguish
varying degrees of positive and negative attitudes
432 UNIT TWO  The Research Process

Center for Epidemiologic Studies Depression Scale

DEPA
THESE QUESTIONS ARE ABOUT HOW YOU HAVE BEEN FEELING LATELY.
AS I READ THE FOLLOWING STATEMENTS, PLEASE TELL ME HOW OFTEN YOU FELT OR
BEHAVED THIS WAY IN THE LAST WEEK. [Hand card ]. FOR EACH STATEMENT, DID YOU FEEL
THIS WAY: [Interviewer: You may help respondent focus on the whichever “style” answer is easier]
0 = Rarely or none of the time (or less than 1 day)?
1 = Some or a little of the time (or 1-2 days)?
2 = Occasionally or a moderate amount of time (or 3-4 days)?
3 = Most or all of the time (or 5-7 days)?

1. I WAS BOTHERED BY THINGS THAT USUALLY DON’T R S O M NR

BOTHER ME. 0 1 2 3 --
2. I DID NOT FEEL LIKE EATING; MY APPETITE WAS POOR. 0 1 2 3 --
3. I FELT THAT I COULD NOT SHAKE OFF THE BLUES EVEN
WITH HELP FROM MY FAMILY AND FRIENDS. 0 1 2 3 --
4. I FELT THAT I WAS JUST AS GOOD AS OTHER PEOPLE. 0 1 2 3 --
5. I HAD TROUBLE KEEPING MY MIND ON WHAT I WAS DOING. 0 1 2 3 --
6. I FELT DEPRESSED. 0 1 2 3 --
7. I FELT THAT EVERYTHING I DID WAS AN EFFORT. 0 1 2 3 --
8. I FELT HOPEFUL ABOUT THE FUTURE. 0 1 2 3 --
9. I THOUGHT MY LIFE HAD BEEN A FAILURE. 0 1 2 3 --
10. I FELT FEARFUL. 0 1 2 3 --
11. MY SLEEP WAS RESTLESS. 0 1 2 3 --
12. I WAS HAPPY. 0 1 2 3 --
13. I TALKED LESS THAN USUAL. 0 1 2 3 --
14. I FELT LONELY. 0 1 2 3 --
15. PEOPLE WERE UNFRIENDLY. 0 1 2 3 --
16. I ENJOYED LIFE. 0 1 2 3 --
17. I HAD CRYING SPELLS. 0 1 2 3 --
18. I FELT SAD. 0 1 2 3 --
19. I FELT PEOPLE DISLIKED ME. 0 1 2 3 --
20. I COULD NOT GET GOING. 0 1 2 3 --

Figure 17-8  Center for Epidemiologic Studies Depression Scale (CES-D). (From Radloff, L. S. [1977]. The CES-D scale: A self-report depression scale
for research in the general population. Applied Psychological Measures, 1, 385–394.)

toward a concept. Figure 17-9 illustrates the form used
for this type of scale.
In a semantic differential scale, values from 1 to 7
are assigned to each of the spaces, with 1 being the
most negative response and 7 the most positive. Place-
ment of negative responses to the left or right of the
scale should be randomly varied to avoid global
responses (in which the subject places checks in the
same column of each scale). Each line is considered
one scale, and the values for the scales are summed to
obtain one score for each subject. Factor analysis is
used to determine the factor structure, which is
expected to reflect three factors or dimensions:
(1) evaluation, (2) potency, and (3) activity (Osgood
et al., 1957). Researchers need to explain theoretically
why particular items on the scale cluster together in
the factor analysis. Thus, development of the instru-
ment contributes to theory development. Factor analy-
sis is also used to evaluate the construct validity of the
instrument. With some of these instruments, three
 CHAPTER 17  Measurement Methods Used in Developing Evidence-Based Practice 433
Nursing Research
Logical
Insignificant
Structured
Active
Organized
Scientific
Important to
Practice
Lacking Rigor
Detailed
Boring
factor scores, each representing one of the dimensions,
are used to describe the subject’s responses and
provide for further analysis (Nunnally & Bernstein,
1994).
Chase (2011) conducted a quasi-experimental study
to examine the effect of an intergenerational email
project on the attitudes of college students toward
older adults. The college students were paired with
older adults with whom they exchanged emails for 6
weeks. The students’ attitudes toward the older adults
were measured with the Aging Semantic Differential
(ASD). The ASD includes 32 bipolar adjectives with
a 7-point linear scale between them. The college stu-
dents indicated their degree of agreement with each
pair of the adjectives on a scale from 1 to 7. Chase
found that the emailing intervention significantly
improved the college students’ attitudes toward older
adults in the experimental group versus the students
in the comparison group.
Visual Analogue Scale
One of the problems with scaling procedures is the
difficulty of obtaining a fine discrimination of values.
In an effort to resolve this problem, the visual ana-
logue scale was developed to measure magnitude,
strength, and intensity of an individual’s sensations or
Illogical
Significant
Unstructured
Passive
Disorganized Figure 17-9  Example items from a semantic
differential scale to measure nursing research.
Unscientific
Unimportant to
Practice
Rigorous
Vague
Exciting
feelings (Wewers & Lowe, 1990). The VAS is referred
to as magnitude scaling (Gift, 1989). This technique
seems to provide interval-level data, and some
researchers argue that it provides ratio-level data
(Sennott-Miller, Murdaugh, & Hinshaw, 1988). It is
particularly useful in scaling stimuli. This scaling
technique has been used to measure pain, mood,
anxiety, alertness, craving for cigarettes, quality of
sleep, attitudes toward environmental conditions,
functional abilities, and severity of clinical symptoms
(Waltz et al., 2010; Wewers & Lowe, 1990).
The stimuli must be defined in a way that the
subject clearly understands. Only one major cue
should appear for each scale. The scale is a line 100
mm in length with right-angle stops at each end. The
line may be horizontal or vertical as shown in Figure
17-10. Bipolar anchors are placed beyond each end
of the line. The anchors should not be placed under-
neath or above the line before the stop. These end
anchors should include the entire range of sensations
possible in the phenomenon being measured. Exam-
ples include “all” and “none,” “best” and “worst,”
and “no pain” and “worst pain imaginable” (see
Figure 17-10).
The VAS is frequently used in healthcare research
because it is easy to construct, administer, and score.
434 UNIT TWO  The Research Process
Figure 17-10  Example of a visual analogue
scale to measure pain. No pain
A VAS can be administered using a drawn, printed, or
computer-generated 100-mm line (Raven et al., 2008;
Waltz et al., 2010). The research participant is asked
to place a mark through the line to indicate the inten-
sity of the sensation or stimulus. A ruler is used to
measure the distance between the left end of the line
and the mark placed by the subject. This measure is
the value of the subject’s sensation. With a computer-
generated VAS, research participants can touch the
VAS line on the computer screen to indicate the degree
of their sensations, such as pain. The computer can
determine the value of the sensation for each subject
and store it in a database (Raven et al., 2008). The
scale is designed to be used while the subject is seated.
Whether use of the scale from the supine position
influences the results by altering perception of the
length of the line has yet to be determined (Gift,
1989). A VAS can be developed for children by using
pictorial anchors at each end of the line rather than
words (Lee & Kieckhefer, 1989).
Wewers and Lowe (1990) published an extensive
evaluation of the reliability and validity of VAS,
although reliability is difficult to determine. Reliabil-
ity of the VAS is most often determined with the test-
retest method, which is effective if the variable being
measured is fairly stable, such as chronic pain. Because
most of the variables measured with the VAS are
labile, test-retest consistency is not applicable, and
because a single measure is obtained, internal consis-
tency cannot be examined. The VAS is more sensitive
to small changes than numerical and rating scales and
can discriminate between two dimensions of pain.
Validity of the VAS has most commonly been
determined by comparing VAS scores with other mea-
sures of a concept. Winkelman, Norman, Maloni, and
Kless (2008) compared VAS scores with dermatome
assessment in measuring pain during labor. The fol-
lowing study excerpt describes the agreement between
these two measures of pain in laboring women who
received an epidural analgesia.
“The Visual Analogue Sensation of Pain Scale was
used to measure patient reports of labor pain. The
solid line was vertical in this study, with no pain
anchoring the bottom of the line and worst pain
Visual Analogue Scale
Worst pain
imaginable
imaginable at the top of the line (Gagliese, Weizblit,
Ellis, & Chan, 2005). A mark at the bottom (i.e., zero)
indicated no pain; 100 was the greatest value of pain.
Each subject placed a mark on the vertical line to
represent the level of pain or discomfort. A superim-
posed transparency marked in 1-mm increments was
used to record the distance of each subject’s mark
from the bottom of the line. Intrarater reliability for
VAS values was set at 2 mm; 90% agreement was
maintained.… A standard dermatome chart was also
used in data collection.… Dermatomes represent
sensory input from spinal nerves to specific areas of
the skin.” (Winkelman et al., 2008, pp. 105-106)
“VAS and dermatome levels were moderately cor-
related at each time point with Pearson’s r ranging
from .331 to .546 (p > .05).… Overall, there was
insufficient agreement on the intensity of pain sensa-
tion between dermatome level and VAS scores.…
The lack of agreement between dermatome levels
and VAS scores indicates that one value cannot be
substituted for another. Specifically, the assessment of
dermatome level provided an equivalent substitute
for the VAS as a measure of pain in laboring women
only at 20 minutes following epidural analgesia admin-
istration.” (Winkelman et al., 2008, pp. 107-108)
Winkelman et al. (2008) clearly described the VAS
used in their study and how the scale was administered
and scored. These researchers found that both the VAS
and the dermatome assessment were easy to use during
labor and were good measures of pain shortly after
epidural anesthesia. However, one measure could not
be substituted for the other, and the best measure of
pain in laboring women is currently dermatome
assessment. Other studies comparing the VAS with
other instruments measuring the same construct have
had varying positive and negative results. Additional
research is needed with the VAS to ensure it is a reli-
able and valid measure of certain patients’ sensations
(Waltz et al., 2010).
Q-Sort Methodology
Q-sort methodology is a technique of comparative
rating that preserves the subjective point of view of
 CHAPTER 17  Measurement Methods Used in Developing Evidence-Based Practice 435
the individual (McKeown & Thomas, 1988). Cards
are used to categorize the importance placed on
various words or phrases in relation to the other
words or phrases in the list. Each phrase is placed on
a separate card. The number of cards should range
from 40 to 100 (Tetting, 1988). The subject is
instructed to sort the cards into a designated number
of piles, usually 7 to 10 piles ranging from the most
to the least important. However, the subject is limited
in the number of cards that may be placed in each
pile. If the subject must sort 60 cards, category 1 (of
greatest importance) may allow only 2 cards; cate-
gory 2, 5 cards; category 3, 10 cards; category 4, 26
cards; category 5, 10 cards; category 6, 5 cards; and
category 7 (the least important), 2 cards. Placement
of the cards fits the pattern of a normal curve. The
subject is usually advised to select first the cards that
he or she wishes to place in the two extreme catego-
ries and then work toward the middle category
(which contains the largest number of cards), rear-
ranging cards until he or she is satisfied with the
results. When sorting the cards, subjects might be
encouraged to make comments about the statements
on the cards and provide a rationale for the catego-
ries where they placed the cards (Akhtar-Danesh,
Baumann, & Cordingley, 2008; Dariel, Wharrad, &
Windle, 2010).
The Q-sort method can also be used to determine
the priority of items or the most important items to
include in the development of a scale. In the previ-
ously mentioned example, the behaviors sorted into
categories 1, 2, and 3 might be organized into a
17-item scale. Correlational or factor analysis is used
to analyze the data (Dariel et al., 2010; Dennis, 1986;
Tetting, 1988). Simpson (1989) suggested using the
Q-sort method for cross-cultural research, with pic-
tures rather than words used for nonliterate groups.
Dariel et al. (2010) used the Q-sort methodology to
examine faculty views toward the use of technology
in nursing education. They described the Q-sort meth-
odology in-depth in their article, and the following
excerpt includes the Q-sort methodology they used in
their study.
“The Q-sort [methodology] typically consists of a
number of statements printed on small cards, which
participants rank according to a ‘condition of instruc-
tion.’ This act of ranking each statement in relation to
others, rather than evaluating them individually, is
designed to capture the way people think about ideas
in relation to other ideas rather than in isolation
(Akhtar-Danesh et al., 2008).” (Dariel et al., 2010,
p. 60)
“Statements with which participants most agree
are placed on the far right of the grid, whereas those
with which they most disagree are placed on the far
left. Cards are then placed in each subsequent column
based on their views towards the previous cards.
While participants read the statements, they are
asked to make comments about their interpretation
of the statements and their placement on the grid.
“During the pilot study, participants were asked to
read the statements using a specific ‘condition of
instruction’ to guide the sort and provide a lens
through which to read each statement. The guiding
statement was ‘Think about the issues which might
be influencing your approach to using technology in
your teaching practice as you read and sort the state-
ments according to how you most agree or most
disagree with how they each impact your use (or
decision not to use) e-learning.’ Participants then
placed each statement into a quasi-normal distribu-
tion grid with 11 categories ranging from −5 to +5.”
(Dariel et al., 2010, pp. 67-68)
Delphi Technique
The Delphi technique measures the judgments of a
group of experts for the purpose of making decisions,
assessing priorities, or making forecasts (Vernon,
2009). Using this technique allows a wide variety of
experts to express opinions and provide feedback,
nationally and internationally, without meeting to­
gether. When the Delphi technique is used, the
opinions of individuals cannot be altered by the per-
suasive behavior of a few people at a meeting. Three
types of Delphi techniques have been identified:
classic or consensus Delphi, dialectic Delphi, and
decision Delphi. In classic Delphi, the focus is on
reaching consensus. Dialectic Delphi is sometimes
called policy Delphi, and the aim is not consensus
but rather to identify and understand a variety of
viewpoints and resolve disagreements. In decision
Delphi, the panel consists of individuals in decision-
making positions. The purpose is to come to a deci-
sion (Vernon, 2009; Waltz et al., 2010). Mitchell
(1998) assessed the validity of the Delphi technique
in nursing education planning and found that 98.1%
of the predicted events had either occurred or were
still expected to occur.
To implement the Delphi technique, researchers
identify a panel of experts, who have a variety of
436 UNIT TWO  The Research Process
Experimenter
(start)
Design initial
questionnaire
Redesign
succeeding
questionnaire(s)
Pilot test questionnaire
Finalize questionnaire
Analyze data and
summarize comments
Mail questionnaire
Collect responses
Respond to questionnaires
with numerical data and
comments
Respondent group
Figure 17-11  Delphi technique sequence model. Multiple arrows indicate repeated cycles of review by experts.
perceptions, personalities, interests, and demograph-
ics to reduce biases in the process. Members of the
panel usually remain anonymous to each other. A
questionnaire is developed that addresses the topics of
concern. Although most questions call for closed-
ended responses, the questionnaire usually contains
opportunities for open-ended responses by the expert.
Once they have completed the questionnaires, the
respondents return them to the researcher, who then
analyzes and summarizes the results. The statistical
analyses usually include measures of central tendency
and measures of dispersion. The role of the researcher
is to maintain objectivity. The outcome of the statisti-
cal analysis is returned to the panel of experts, along
with a second questionnaire. Respondents with
extreme responses to the first round of questions may
be asked to justify their responses. The respondents
return the second round of questionnaires to the
researcher for analysis. This procedure is repeated
until the data reflect a consensus among the panel.
Limiting the process to two or three rounds is not a
good idea if consensus is the goal. In some studies,
true consensus is reached, whereas in others, “major-
ity rules.” Some authors question whether the
Perform
final analysis and
report findings
Consensus
reached
Plan and
provide feedback
No consensus reached
agreement reached is genuine (Vernon, 2009; Waltz et
al., 2010). Couper (1984) developed a model of the
Delphi technique, which is presented in Figure 17-11.
This model might assist you in implementing a Delphi
technique in a study.
Vernon (2009) identified benefits and limitations of
the Delphi technique. The benefits include increased
access to experts and usually good response rates. The
Delphi design has simplicity and flexibility in its use;
it is easily understood and implemented by research-
ers. Because the participants are anonymous, views
can be expressed freely without direct persuasion from
others.
There are also several potential problems that
researchers could encounter when using the Delphi
technique. There has been no documentation that the
responses of “experts” are different from responses
one would receive from a random sample of subjects.
Because the panelists are anonymous, they have no
accountability for their responses. Respondents could
make hasty, ill-considered judgments because they
know that no negative feedback would result. Feed-
back on the consensus of the group tends to centralize
opinion, and traditional analysis with the use of means
 CHAPTER 17  Measurement Methods Used in Developing Evidence-Based Practice 437
and medians may mask the responses of individuals
who are resistant to the consensus sentiment. Conclu-
sions could be misleading (Vernon, 2009).
Some nursing specialty organizations have estab-
lished their research priorities by using Delphi tech-
niques. Lindeman (1975) conducted one of the initial
studies using the Delphi technique to determine
research priorities in clinical nursing. She used a panel
of 433 experts, nurses and nonnurses, with a wide
range of interests. The panel was sent four rounds of
a 150-item questionnaire. The report, published in
Nursing Research, had an important influence on the
research conducted in nursing for clinical practice.
Wilkes, Mohan, Luck, and Jackson (2010) used the
Delphi technique to develop a violence assessment
tool for a hospital emergency department. The follow-
ing excerpt from their study describes the methodol-
ogy they used.
“The aim of this study was to develop a violence
assessment tool by refining a list of predictive cues
identified from both a previous study and existing
literature. Using the Delphi technique, a panel of 11
expert nurse academics and clinicians developed a
37-item questionnaire and used three rounds of
Delphi to refine the violence assessment question-
naire. The resulting tool comprises 17 cues of poten-
tial violence that can be easily observed and requires
no prior knowledge of the perpetrators’ medical
history.” (Wilkes et al., 2010, p. 70)
Diaries
A diary is a recording of events over time by an indi-
vidual to document experiences, feelings, or behavior
patterns. Diaries are also called logs or journals and
have been used since the 1950s to collect data for
research from various populations including children,
patients with acute and chronic illness, pregnant
women, and elderly adults (Aroian & Wal, 2007;
Nicholl, 2010). A diary, which allows recording
shortly after an event, is thought to be more accurate
than obtaining the information through recall during
an interview. In addition, the reporting level of inci-
dents is higher, and one tends to capture the partici-
pant’s immediate perception of situations.
The diary technique gives nurse researchers a
means to obtain data on topics of particular interest
within nursing that have not been accessible by other
means. Some potential topics for diary collection
include expenses related to a healthcare event (par-
ticularly out-of-pocket expenses), self-care activities
(frequency and time required), symptoms of disease,
eating behavior, exercise behavior, sexual activities,
the child development process, and care provided by
family members in a home-care situation. Although
diaries have been used primarily with adults, they are
also an effective means of collecting data from school-
age children.
Health diaries have been used to document health
problems, responses to symptoms, and efficacy of
responses. Diaries may also be used to determine how
people spend their days; this information could be
particularly useful in managing the care needs of indi-
viduals with chronic illnesses. In experimental studies,
diaries may be used to determine responses of subjects
to experimental treatments. Diaries can take a variety
of forms and might include filling in blanks, selecting
the best response from a list of options, or checking a
column. Figure 17-12 shows a page from a diary for
patients to record their symptoms and how they were
managed. This diary includes blanks to identify the
symptoms and an option to check how the symptoms
were managed. This type of diary is used to collect
numerical data for a quantitative study. Validity and
reliability have been examined by comparing the
results with data obtained through interviews and have
been found to be acceptable. Participation in studies
using health diaries has been good, and attrition rates
are reported as low. Some diaries include the collec-
tion of narrative data and are more common in qualita-
tive studies (Alaszewski, 2006).
Nicholl (2010) and Burman (1995) provide some
key points to consider when selecting a diary for col-
lecting data in a study:
1. Analyze the phenomenon of interest to determine
if it can be adequately captured using a diary. Also,
determine if a diary is the best data collection
approach when compared with interviews, ques-
tionnaires, and scales.
2. Decide if the diary will be used alone or with other
measurement methods.
3. Determine which format of the diary to use so that
the most valid information can be obtained to
address the study purpose without burdening the
study participants. Diaries can be paper, online, or
phone text-messaging formats. Some researchers
are using blogs as a way to collect diary data (Lim,
Sacks-Davis, Aitken, Hocking, & Hellard, 2010).
The format of the questions in diaries can also vary
based on the purpose of the study. Diaries with
closed-ended questions are usually used in quanti-
tative research, and participants are provided spe-
cific direction on the data to be recorded. Diaries
with open-ended questions are more common in
438 UNIT TWO  The Research Process
Date What symptom
did you have?
Figure 17-12  Sample diary page.
qualitative research with the narrative data requir-
ing content analysis (Alaszewski, 2006; Nicholl,
2010).
4. Pilot-test any new or refined diary with the target
population of interest to identify possible prob-
lems, determine if the instructions and terminology
are clear, ensure that the data can be recorded with
this approach, and examine the ability of partici-
pants to complete diaries.
5. Determine the period of time that the diary will be
completed to accomplish the purpose of the study,
taking into consideration the burden on the partici-
pants. Typical diary periods are 2 to 8 weeks.
6. Provide clear instructions to all participants on the
use of a diary before the study begins to enhance
the quality of data collected. Participants need to
know how to use the diaries, what types of events
are to be reported, and how to contact the researcher
or clinician with questions.
7. Use follow-up procedures, such as phone calls or
emails, during data collection to enhance comple-
tion rates. Diaries might be emailed, mailed, or
picked up by the researchers. Picking up the diary
in person promotes a higher completion rate than
mailing.
8. Plan data analysis procedures during diary devel-
opment and refine these plans to ensure the most
appropriate analyses are used. Diary data are very
dense and rich, and carefully prepared analysis
plans can minimize problems (Burman, 1995;
Nicholl, 2010).
The use of diaries has some disadvantages. In some
cases, keeping the diary may alter the behavior or
Did you talk Did you talk Did you take
with a family with a health any pills or
member or professional treatments
friend about about it? for the
the symptom? symptom?
No Yes No Yes No Yes,
Specify
events under study. For example, if a person were
keeping a diary of the nursing care that he or she was
providing to patients, the insight that the person gained
from recording the information in the diary might lead
to changes in care. In addition, patients can become
more sensitive to items (e.g., symptoms or problems)
reported in the diary, which could result in overreport-
ing. Subjects may also become bored with keeping the
diary and become less thorough in recording items,
which could result in underreporting (Aroian & Wal,
2007; Nicholl, 2010).
Lim et al. (2010) conducted a randomized con-
trolled trial to determine the best diary format for
collecting sexual behavior information from adoles-
cents. The three formats for the diaries were paper,
online, and phone text messaging (short message
service); these were compared for response rate, time-
liness, completeness of data, and acceptability. The
following excerpt describes the use of the diaries for
data collection and the outcomes.
“Participants were recruited by telephone and ran-
domized into one of three groups. They completed
weekly sexual behavior diaries for 3 months by SMS
[short message service], online, or paper (by post).
An online survey was conducted at the end of 3
months to compare retrospective reports with the
diaries and assess opinions on the diary collection
method.… Conclusions were that the SMS is a con-
venient and timely method of collecting brief
behavioral data, but online data collection was
 CHAPTER 17  Measurement Methods Used in Developing Evidence-Based Practice 439
preferable to most participants and more likely to be
completed. Data collected in retrospective sexual
behavior questionnaires were found to agree substan-
tially with data collected through weekly self-report
diaries.” (Lim et al., 2010, p. 885)
Lim et al. (2010) provided some valuable informa-
tion about the formats for collecting data with diaries.
Researchers might want to consider using online or
phone text messaging to collect diary data from
younger populations. These formats could signifi-
cantly increase the response rate and the completeness
of the data collected. The paper format for collecting
diary data also provides quality information and might
be better for populations with limited access to
technology.
Measurement Using
Existing Databases
Nurse researchers are increasing their use of existing
databases to address the research problems they have
identified as essential in generating evidence for prac-
tice. The reasons for using these databases in studies
are varied. With the computerization of healthcare
information, more large data sets have been developed
internationally, nationally, regionally, at the state level,
and within clinical agencies. These databases include
large amounts of information that have relevance in
developing research evidence needed for practice
(Brown, 2009; Melnyk & Fineout-Overholt, 2011).
The costs and technology for storage of data have
improved over the last 10 years making these large
data sets more reliable and accessible. Using existing
databases makes it possible to conduct complex analy-
ses to expand understanding of healthcare outcomes
(Doran, 2011). Another reason is that the primary col-
lection of data in a study is limited by the availability
of participants and the expense of the data collection
process. By using existing databases, researchers are
able to have larger samples, conduct more longitudinal
studies, experience less costs during the data collec-
tion process, and limit the burdens placed on study
participants (Johantgen, 2010).
There are also problems with using data from exist-
ing databases. The data in the database might not
clearly address the researchers’ study purpose. Most
researchers identify a study problem and purpose and
then develop a methodology to address these. The data
collected are specific to the study and clearly focused
on answering the research questions or testing the
study hypotheses. However, with existing databases,
researchers need to ensure that the data they require
for their study is in the database they are planning to
use. Sometimes researchers must revise their study
questions and variables based on what data exist in the
database. The level of measurement of the study vari-
ables might limit the analysis techniques that can be
used. There is also the question of the validity and
reliability of the data in existing databases; unless
these are specifically reported, researchers using these
data files need to be cautious in their interpretation of
findings.
Existing Healthcare Data
Existing healthcare data consist of two types: second-
ary and administrative. Data that are collected for a
particular study are considered primary data. Data that
are collected from previous research and stored in a
database are considered secondary data when used by
other researchers to address their study purposes.
Because these data were collected as part of research,
details can be obtained about the data collection and
storage process. Researchers usually clearly indicate
in their article in the methodology section when sec-
ondary data analyses were conducted as part of their
research (Johantgen, 2010).
Data that are collected for reasons other than
research are considered administrative data. Adminis-
trative data are collected within clinical agencies;
obtained by national, state, and local professional
organizations; and collected by federal, state, and
local agencies. The processes for collection and
storage of administrative data are more complex and
often more unclear than the data collection process
for research (Johantgen, 2010). The data in adminis-
trative databases are collected by different people
in different sites using different methods. However,
the data elements collected for most administrative
databases include demographics, organizational char-
acteristics, clinical diagnosis and treatment, and geo-
graphical information. These database elements were
standardized by the Health Insurance Portability and
Accountability Act (HIPAA) of 1996, which improved
the quality of the databases. The HIPAA regulations
can be viewed online at http://www.hhs.gov/ocr/
privacy/ (U.S. Department of Health & Human Ser-
vices, 2011).
When using secondary data and administrative
data in a study, researchers need to determine the reli-
ability and validity of the data in the database they
plan to access. They also need to ensure that the data
in the data set address the research questions or
440 UNIT TWO  The Research Process
hypotheses of their study Lake, Shang, Klaus, and
Dunton (2010) used an existing database entitled the
National Database of Nursing Quality Indicators
(NDNQI) for the conduct of their study. The NDNQI
was designed to measure nursing quality and patient
safety. Lake et al. examined the relationships between
patient falls, nursing unit staffing, and hospital Magnet
status. The following excerpt describes the database
they used.
“NDNQI Database Overview
The NDNQI, a unique database that was well-suited
to our study aims, is part of the American Nurses
Association’s (ANA) Safety and Quality Initiative. This
initiative started in 1994 with information gathering
from an expert panel and focus groups to specify a
set of 10 nurse-sensitive indicators to be used in the
database.… The database was pilot tested in 1996
and 1997 and was established in 1998 with 35 hospi-
tals. Use of the NDNQI has grown rapidly.… In 2009,
1,450 hospitals—one out of every four general hos-
pitals in the U.S.—participated in it.
The NDNQI has served as a unit-level benchmark-
ing resource, but research from this data repository
has been limited. NDNQI researchers have published
two studies on the association between characteris-
tics of nursing workforce and fall rates.… The scope
of work on this topic was extended in the current
study by: (a) specifying nurse staffing separately for
RNs [registered nurses], LPNs [licensed practical
nurses], and NAs [nursing assistants], (b) using the
entire NDNQI database, (c) selecting the most
detailed level of observation (month), and (d) applying
more extensive patient risk adjustment than had been
evaluated previously.” (Lake et al., 2010, p. 415)
Lake et al. (2010) provide a detailed description of
the national database that they used in their study. This
database was selected because it was “well-suited” to
the study’s aims and because the focus of the study
was to expand on previous research that had been done
using this database information. The development of
the database was very structured, which increases the
reliability and validity of the field data.
Selection of an
Existing Instrument
Selecting an instrument to measure the variables in a
study is a critical process in research. The method of
measurement selected must fit closely the conceptual
definition of the variable. Researchers need to conduct
an extensive search of the literature to identify appro-
priate methods of measurement. In many cases, they
find instruments that measure some of the needed ele-
ments but not all, or the content may be related to but
somehow different from what is needed for a study.
Instruments found in the literature may have little or
no documentation of their validity and reliability.
Beginning researchers often conclude that no
appropriate method of measurement exists and that
they must develop a tool. At the time, this solution
seems to be the most simple because the researcher
has a clear idea of what needs to be measured. This
solution is not recommended unless all else fails. Tool
development is a lengthy process and requires sophis-
ticated research. Using a new instrument in a study
without first evaluating its validity and reliability can
be problematic and lead to questionable findings.
For novice researchers developing their first study,
it is essential to identify existing instruments to
measure study variables. Jones (2004) developed a
flow chart that might help you to select an existing
instrument for your study (Figure 17-13). The major
steps include (1) identifying an instrument from the
literature, (2) determining if the instrument is appro-
priate for measuring a study variable, and (3) examin-
ing the performance of the instrument by evaluating
its reliability and validity. These steps are detailed in
the following sections.
Locating Existing Instruments
Locating existing measurement methods has become
easier in recent years. A computer database, the Health
and Psychological Instruments Online (HAPI), is
available in many libraries and can be used to search
for instruments that measure a particular concept or
for information on a particular instrument. Sometimes
a search on Medline or CINAHL might uncover an
instrument that is useful. Many reference books have
compiled published measurement tools, some that are
specific to instruments used in nursing research. Dis-
sertations often contain measurement tools that have
never been published, so a review of Dissertation
Abstracts online might be helpful.
Another important source of recently developed
measurement tools is word-of-mouth communication
among researchers. Information on tools is often pre-
sented at research conferences years before publica-
tion. There are usually networks of researchers
conducting studies on similar nursing phenomena.
These researchers are frequently associated with
nursing organizations and keep in touch through
 CHAPTER 17  Measurement Methods Used in Developing Evidence-Based Practice 441

Existing Tool New Tool

Identification
Perform literature search Develop new tool

Design
Potentially No Write study protocol
relevant
tool exists?

Implementation
Yes Collect data
Appropriateness
Define criteria

Analysis
Perform univariate
and multivariate Figure 17-13  Flow chart depicting the identification and
Tool No analyses assessment of an existing tool and development of a new
satisfies tool.
criteria?

Yes Tool
Devise allocation rule
Performance
Evaluate reliability and validity

Performance
Evaluate goodness
of fit, reliability, and
Is tool validity
No
valid and
reliable?

Yes

Use existing tool

newsletters, correspondence, telephone, email, com-

puter discussion boards, and web pages. Questioning
available nurse investigators can lead to a previously
unknown tool. These researchers can often be reached
by telephone, letter, or email and are usually willing
to share their tools in return for access to the data to
facilitate work on developing validity and reliability
information. The Sigma Theta Tau Directory of Nurse
Researchers provides email address and phone infor-
mation on nurse researchers. In addition, it lists nurse
researchers by category according to their area of
research.
Waltz and colleagues (2010) made the following
suggestions to facilitate locating existing instruments
for studies:
“(1) [S]earch computerized databases by using the
name of the instrument or keywords or phrases;
(2) generalize the search to the specific area of inter-
est and related topics (research reports are particu-
larly valuable); (3) search for summary articles
describing, comparing, contrasting, and evaluating the
instruments used to measure a given concept;
(4) search journals, such as Journal of Nursing Measure-
ment, that are devoted specifically to measurement;
(5) after identifying a publication in which relevant
instruments are used, use citation indices to locate
other publications that used them; (6) examine
computer-based and print indices, and compendia of
442 UNIT TWO  The Research Process
instruments developed by nursing, medicine, and
other disciplines; and (7) examine copies of published
proceedings and abstracts from relevant scientific
meetings.” (Waltz et al., 2010, pp. 393-394)
Evaluating Existing Instruments for
Appropriateness and Performance
You may need to examine several instruments to find
the one most appropriate for your study. When select-
ing an instrument for research, carefully consider how
the instrument was developed, what the instrument
measures, and how to administer it. Before you review
existing instruments, be sure you have conceptually
defined your study variable and are clear on what you
desire to measure. You then need to address the fol-
lowing questions to determine the best instrument for
measuring your study variable:
1. Does this instrument measure what you want to
measure?
2. Does the instrument reflect your conceptual defi-
nition of the variable?
3. Is the instrument well constructed?
4. Does your population resemble populations previ-
ously studied with the instrument? (Waltz et al.,
2010)
5. Is the readability level of the instrument appropri-
ate for your population?
6. How sensitive is the instrument in detecting small
differences in the phenomenon you want to
measure (what is the effect size)?
7. What is the process for obtaining, administering,
and scoring the instrument? Are there costs asso-
ciated with the instrument?
8. What skills are required to administer the instru-
ment? Do you need training or a particular cre-
dential to administer the instrument?
9. How are the scores interpreted?
10. What is the time commitment of the study partici-
pants and researcher for administration of the
instrument?
11. What evidence is available related to the reliabil-
ity and validity of the instrument? Have multiple
types of validity been examined (content validity,
validity from factor analysis, validity from exam-
ining measures assessing for convergence and
divergence, or evidence of validity from predic-
tion of concurrent and future events)? Chapter 16
provided a detailed discussion of reliability and
validity (Bartlett & Frost, 2008; Bialocerkowski
et al., 2010; DeVon et al., 2007; Fawcett & Garity,
2009).
Assessing Readability Levels of Instruments
The readability level of an instrument is a critical
factor when selecting an instrument for a study.
Regardless of how valid and reliable the instrument is,
it cannot be used effectively if study participants do
not understand the items. Calculating readability is
easy and can be done in a few minutes. Many word
processing programs and computerized grammar
checkers report the readability level of written mate-
rial. The Fog formula described in Chapter 16 pro-
vides a quick and easy way to assess readability. If the
reading level of an instrument is beyond the reading
level of the study population, you need to select
another instrument for use in your study. Changing the
items on an instrument to reduce the reading level can
alter the validity and reliability of the instrument.
Constructing Scales
Scale construction is a complex procedure that should
not be undertaken lightly. There must be firm evidence
of the need for developing another instrument to
measure a particular phenomenon important to nursing
practice. However, in many cases, measurement
methods have not been developed for phenomena of
concern to nurse researchers, or measurement tools
that have been developed may be poorly constructed
and have insufficient evidence of validity to be accept-
able for use in studies. It is possible for the researcher
to carry out instrument development procedures on an
existing scale with inadequate evidence of validity
before using it in a study. Neophyte nurse researchers
could assist experienced researchers in carrying out
some of the field studies required to complete the
development of scale validity and reliability.
The procedures for developing a scale have been
well defined. The following discussion describes this
theory-based process and the mathematical logic
underlying it. The theories on which scale construc-
tion is most frequently based include classic test
theory (Nunnally & Bernstein, 1994), item response
theory (Hulin, Drasgow, & Parsons, 1983), multi­
dimensional scaling (Borg & Groenen, 2010; Kruskal
& Wish, 1990), and unfolding theory (Coombs, 1950).
Most existing instruments used in nursing research
have been developed with classic test theory, which
assumes a normal distribution of scores.
Constructing a Scale by Using Classic
Test Theory
In classic test theory, the following process is used to
construct a scale:
 CHAPTER 17  Measurement Methods Used in Developing Evidence-Based Practice 443
1. Define the concept. A scale cannot be constructed
to measure a concept until the nature of the
concept has been delineated. The more clearly the
concept is defined, the easier it is to write items
to measure it (Spector, 1992). Concepts are
defined through the process of concept analysis,
a procedure discussed in Chapter 7.
2. Design the scale. Items should be constructed to
reflect the concept as fully as possible. The
process of construction differs depending on
whether the scale is a rating scale, Likert scale, or
semantic differential scale. Items previously
included in other scales can be used if they have
been shown empirically to be good indicators of
the concept (Hulin et al., 1983). A blueprint may
ensure that all elements of the concept are covered.
Each item must be stated clearly and concisely
and express only one idea. The reading level of
items must be identified and considered in terms
of potential respondents. The number of items
constructed must be considerably larger than
planned for the completed instrument because
items are discarded during the item analysis step
of scale construction. Nunnally and Bernstein
(1994) suggested developing an item pool at least
twice the size of that desired for the final scale.
3. Review the items. As items are constructed, it is
advisable to ask qualified individuals to review
them. Crocker and Algina (1986) recommended
asking for feedback in relation to accuracy, appro-
priateness, or relevance to test specifications;
technical flaws in item construction; grammar;
offensiveness or appearance of bias; and level of
readability. The items should be revised according
to the critical appraisal.
4. Conduct preliminary item tryouts. While items are
still in draft form, it is helpful to test them on a
limited number of subjects (15 to 30) who repre-
sent the target population. The reactions of
respondents should be observed during testing to
note behaviors such as long pauses, answer chang-
ing, or other indications of confusion about spe-
cific items. After testing, a debriefing session
needs to be held during which respondents are
invited to comment on items and offer sugges-
tions for improvement. Descriptive and explor-
atory statistical analyses are performed on data
from these tryouts while noting means, response
distributions, items left blank, and outliers. Items
need to be revised based on this analysis and com-
ments from respondents.
5. Perform a field test. All the items in their final
draft form are administered to a large sample of
subjects who represent the target population.
Spector (1992) recommended a sample size of
100 to 200 subjects. However, the sample size
needed for the statistical analyses to follow
depends on the number of items in the instrument.
Some experts recommend including 10 subjects
for each item being tested. If the final instrument
was expected to have 20 items, and 40 items were
constructed for the field test, 400 subjects could
be required.
6. Conduct item analyses. The purpose of item anal-
ysis is to identify items that form an internally
consistent or reliable scale and to eliminate items
that do not meet this criterion. Internal reliability
implies that all the items are consistently mea­
suring a concept. Before these analyses are
conducted, negatively worded items must be
reverse-scored or given a score as though the item
was stated positively. For example, the item might
read “I do not believe exercise is important to
health,” with the responses of 1 = strongly dis-
agree, 2 = disagree, 3 = uncertain, 4 = agree, and
5 = strongly agree. If the subject marked a 1 for
strongly disagree, this item would be reverse-
scored and given a 5, indicating the subject thinks
exercise is very important to health. The analyses
examine the extent of intercorrelation among the
items. The statistical computer programs cur-
rently providing the set of statistical procedures
needed to perform item analyses (as a package)
are SPSS, SPSS/PC, and SYSTAT. These pack-
ages perform item-item correlations and item-
total correlations. In some cases, the value of the
item being examined is subtracted from the total
score, and an item-remainder coefficient is calcu-
lated. This latter coefficient is most useful in
evaluating items for retention in the scale.
7. Select items to retain. Depending on the number
of items desired in the final scale, items with the
highest coefficients are retained. Alternatively, a
criterion value for the coefficient (e.g., 0.40) can
be set, and all items greater than this value are
retained. The greater the number of items retained,
the smaller the item-remainder coefficients can be
and still have an internally consistent scale. After
this selection process, a coefficient alpha is calcu-
lated for the scale. This value is a direct function
of the number of items and the magnitude of inter-
correlation. Thus, one can increase the value of
a coefficient alpha by increasing the number of
items or raising the intercorrelations through
inclusion of more highly intercorrelated items.
Values of coefficient alphas range from 0 to 1. The
444 UNIT TWO  The Research Process
alpha value should be at least 0.70 to indicate suf-
ficient internal consistency in a new tool (Nunnally
& Bernstein, 1994). An iterative process of remov-
ing or replacing items or both, recalculating item-
remainder coefficients, and recalculating the alpha
coefficient are repeated until a satisfactory alpha
coefficient is obtained. Deleting poorly correlated
items raises the alpha coefficient, but decreasing
the number of items lowers it (Spector, 1992). The
initial attempt at scale development may not
achieve a sufficiently high coefficient alpha. In
this case, additional items need to be written, more
data collected, and the item analysis redone. This
scenario is most likely to occur when too few items
were developed initially or when many of the
initial items were poorly written. It may also be a
consequence of attempts to operationalize an inad-
equately defined concept (Spector, 1992).
8. Conduct validity studies. When scale development
is judged to be satisfactory, studies must be per-
formed to evaluate the validity of the scale. (See
the discussion of validity in Chapter 16.) These
studies require the researcher to collect additional
data from large samples. As part of this process,
scale scores must be correlated with scores on
other variables proposed to be related to the
concept being put into operation. Hypotheses must
be generated regarding variations in mean values
of the scale in different groups. Exploratory and
confirmatory factor analysis (discussed in Chap-
ters 16 and 23) is usually performed as part of
establishing the validity of the instrument. As
many different types of evidence of validity as
possible should be collected (Spector, 1992).
9. Evaluate the reliability of the scale. Various sta-
tistical procedures are performed to determine the
reliability of the scale. These analyses can be per-
formed on the data collected to evaluate validity
for this scale. (See Chapter 16 for a discussion of
the procedures performed to examine reliability.)
10. Compile norms on the scale. To determine norms,
the scale must be administered to a large sample
that is representative of the groups to which the
scale is likely to be administered. Norms should
be acquired for as many diverse groups as possi-
ble. Data acquired during validity and reliability
studies can be included for this analysis. To obtain
the large samples needed for this purpose, many
researchers permit others to use their scale with
the condition that data from these studies be pro-
vided for compiling norms.
11. Publish the results of development of the scale.
Scales are often not published for many years
after the initial development because of the length
of time required to validate the instrument. Some
researchers never publish the results of this work.
Studies using the scale are published, but the
instrument development process may not be
available except by writing to the author. This
information needs to be added to the body of
knowledge, and colleagues should encourage
instrument developers to complete the work and
submit it for publication (Lynn, 1989; Norbeck,
1985). Klein et al. (2010) provided a detailed dis-
cussion of their development of the NPAT that
was presented earlier in this chapter. The validity
and reliability of the tool were addressed and a
copy of the tool was included in the article (see
Figure 17-4).
Constructing a Scale by Using Item
Response Theory
Using item response theory to construct a scale pro-
ceeds initially in a fashion similar to that of classic test
theory. There is an expectation of a well-defined
concept to operationalize. Items are initially written in
a manner similar to that previously described, and
item tryouts and field testing are also similar. However,
the process changes with the initiation of item analy-
sis. The statistical procedures used are more sophisti-
cated and complex than the procedures used in classic
test theory. Using data from field testing, item charac-
teristic curves are calculated by using logistic regres-
sion models (Hulin et al., 1983; Nunnally & Bernstein,
1994). After selecting an appropriate model based on
information obtained from the analysis, item param-
eters are estimated. These parameters are used to
select items for the scale. This strategy is used to avoid
problems encountered with classic test theory
measures.
Scales developed by using classic test theory effec-
tively measure the characteristics of subjects near the
mean. The statistical procedures used assume a linear
distribution of scale values. Items reflecting responses
of respondents closer to the extremes tend to be dis-
carded because of the assumption that scale values
should approximate the normal curve. Scales devel-
oped in this manner often do not provide a clear under-
standing of study participants at the high or low end
of values.
One purpose of item response theory is to choose
items in such a way that estimates of characteristics at
each level of the concept being measured are accurate.
To accomplish this goal, researchers use maximal like-
lihood estimates. A curvilinear distribution of scale
values is assumed. Rather than choosing items on the
 CHAPTER 17  Measurement Methods Used in Developing Evidence-Based Practice 445
basis of the item remainder coefficient, the researcher
specifies a test information curve. The scale can be
tailored to have the desired measurement accuracy. By
comparing a scale developed by classic test theory
with one developed from the same items with item
response theory, one would find differences in some
of the items retained. Biserial correlations would be
lower in the scale developed from item response
theory than in the scale developed from classic test
theory. Item bias is lower in scales developed by using
item response theory and occurs when respondents
from different subpopulations having the same amount
of an underlying trait have different probabilities of
responding to an item positively (Hambleton &
Swaminathan, 2010; Hulin et al., 1983).
Constructing a Scale by Using
Multidimensional Scaling
Multidimensional scaling is used when the concept
being operationalized is actually an abstract construct
believed to be represented most accurately by multiple
dimensions. The scaling techniques used allow the
researcher to uncover the hidden structure in the con-
struct. The analysis techniques use proximities among
the measures as input. The outcome of the analysis is
a spatial representation, or a geometrical configuration
of data points, that reveals the hidden structure. The
procedure tends to be used to examine differences in
stimuli rather than differences in people. A researcher
might use this method to measure differences in per-
ception of light or pain. Scales developed by using this
procedure reveal patterns among items. The procedure
is used in the development of rating scales and seman-
tic differentials (Borg & Groenen, 2010; Kruskal &
Wish, 1990).
Constructing a Scale by Using
Unfolding Theory
oasis-ebl|Rsalles|1476321091
When a scale is being constructed with the use of
unfolding theory, researchers ask study participants to
respond to the items in the rating scale. Next, partici-
pants are asked to rank the various response options
in relation to the response option that they selected
for that item. This procedure is followed for each item
in the scale. By using this procedure, the underlying
continuum for each scale item is “unfolded.” As
an example, suppose researchers developed the fol-
lowing item:
My preference for a diet to lose weight is
1. A low-fat diet
2. A low-calorie diet
3. A low-carbohydrate diet
4. A vegetarian diet
Study participants would be asked to select their
response to the item and then rank the other options
according to the proximity to their choice. The partici-
pant might choose a low-calorie diet as number 1, a
low-carbohydrate diet as number 2, a low-fat diet as
number 3, and a vegetarian diet as number 4. Although
the preferences of other study participants would
differ, the results can be plotted to reveal patterns of
an underlying continuum. Items selected for the scale
would be the items with evidence of a pattern of
responses.
Translating a Scale
to Another Language
Contrary to expectations, translating an instrument
from the original language to a target language is a
complex process. By translating a scale, researchers
can compare concepts among respondents of different
cultures. The comparison requires that they first infer
and then validate that the conceptual meaning in
which the scale was developed is the same in both
cultures. This process is highly speculative, and con-
clusions about the similarities of meanings in a
measure must be considered tentative (Hulin et al.,
1983).
Four types of translations can be performed: prag-
matic translations, aesthetic-poetic translations, ethno-
graphic translations, and linguistic translations.
Pragmatic translations communicate the content from
the source language accurately in the target language.
The primary concern is the information conveyed. An
example of this type of translation is the use of trans-
lated instructions for assembling a computer. Aesthetic-
poetic translations evoke moods, feelings, and affect
in the target language that are identical to those evoked
by the original material. In ethnographic translations,
the purpose is to maintain meaning and cultural
content. In this case, translators must be familiar with
both languages and cultures. Linguistic translations
strive to present grammatical forms with equivalent
meanings. Translating a scale is generally done in the
ethnographic mode (Hulin et al., 1983).
One strategy for translating scales is to translate
from the original language to the target language and
then back-translate from the target language to the
original language by using translators not involved in
the original translation (Fawcett & Garity, 2009). Dis-
crepancies are identified, and the procedure is repeated
until troublesome problems are resolved. After this
procedure, the two versions are administered to bilin-
gual subjects and scored by standard procedures. The
446 UNIT TWO  The Research Process
resulting sets of scores are examined to determine the
extent to which the two versions yield similar informa-
tion from the subjects. This procedure assumes that
the subjects are equally skilled in both languages. One
problem with this strategy is that bilingual subjects
may interpret meanings of words differently from
monolingual subjects. This difference in interpretation
is a serious concern because the target subjects for
most cross-cultural research are monolingual (Hulin
et al., 1983).
Yu, Lee, and Woo (2004) provided an excellent
description of their process of translating the Medical
Outcomes Study Social Support Survey (MOS-SSS)
from English to Chinese. These researchers used the
forward and backward translation process previously
discussed, and the steps they took are outlined in the
following excerpt.
“This translation model includes a cycle of four steps
as follows.
Forward translation of the MOS-SSS by
a bilingual health professional. The translation
process began with forward translation of the original
source language (SL) version (English) of the MOS-SSS
into the target language (TL) of Chinese by a bilingual
native Chinese registered nurse.…
Review of the Chinese MOS-SSS by a mono-
lingual reviewer. The Chinese version of the
MOS-SSS was then reviewed by a Chinese monolin-
gual reviewer for incomprehensible or ambiguous
wordings.…
Backward translation of the Chinese MOS-SSS
by a bilingual health professional. In this step, the
reviewed Chinese version of the MOS-SSS (as dis-
cussed in Step 2) was back translated by another
bilingual nurse, who was ‘blinded’ to the original
English version.…
Comparison of the SL version and back-
translated version. The researcher, at this stage,
compared the back-translated version of the MOS-SSS
with its original version for linguistic congruence and
cultural relevancy. Items with apparent discrepancies
were examined to ascertain whether the problems
originated in the forward translation or the backward
translation. The error in items resulting from the
forward translation had to go through the whole-
cycle again from Steps 1 to 4, whereas the latter type
of error was subjected to further back translation.
This process was repeated until a maximum equiva-
lence between the SL and back-translated versions
was achieved.” (Yu et al., 2004, pp. 309-310)
In 1997, the Medical Outcomes Trust introduced
new translation criteria that are much more compre-
hensive. The discussion of these criteria is available at
www.outcomes-trust.org/bulletin/0797blltn.htm.
Hulin et al. (1983) suggested the use of item response
theory procedures to address some of the problems of
translation. These procedures can provide direct evi-
dence about the meanings of items in the two lan-
guages. Item characteristic curves for an item in the
two languages can be compared, as can scale scores in
the two languages. This procedure eliminates the need
for bilingual samples. It also eliminates the need
for the two populations to be equivalent in terms of
the distributions of their scores on the trait being
measured.
Rather than translating an instrument into each lan-
guage, Turner, Rogers, Hendershot, Miller, and Thorn-
berry (1996) tested the use of electronic technology
involving multilingual audio computer-assisted self-
interviewing (Audio-CASI) to enable researchers to
include multiple linguistic minorities in nationally
representative studies and clinical studies. The Audio-
CASI system uses electronic translation from one lan-
guage to another. In the funded project to develop and
test Audio-CASI, a backup phone bank was available
to provide multilingual assistance if needed. Whether
this strategy will provide equivalent validity of a trans-
lated tool is unclear.
KEY POINTS
• Measurement approaches used in nursing research
include physiological measures; observations; in-
terviews; questionnaires; scales; and specialized
instruments such as Q-sort method, Delphi tech-
nique, diaries, and analyses using existing data-
bases.
• Measurements of physiological variables can be
either direct or indirect and sometimes require
the use of specialized equipment or laboratory
analysis.
• The Human Genome Project has increased the
opportunities for nurses to be involved in genetic
research and to include the measurement of nucleic
acids in their studies.
• To measure observations, every variable is observed
in a similar manner in each instance, with careful
attention given to training data collectors.
• In structured observational studies, category
systems must be developed; checklists or rating
scales are developed from the category systems and
used to guide data collection.
 CHAPTER 17  Measurement Methods Used in Developing Evidence-Based Practice 447
• Interviews involve verbal communication between
the researcher and the study participant, during
which the researcher acquires information. Inter-
viewers must be trained in the skills of interview-
ing, and the interview protocol must be pretested.
• A questionnaire is a printed or electronic self-report
form designed to elicit information through the
responses of a study participant. The information
obtained through questionnaires is similar to infor-
mation obtained by interview, but the questions
tend to have less depth. An item on a questionnaire
usually has two parts: a lead-in question and a
response set.
• Scales, another form of self-reporting, are more
precise in measuring phenomena than question-
naires and have been developed to measure psycho-
social and physiological variables. The types
of scales included in this text are rating scale,
Likert scale, semantic differential scale, and visual
analogue scale.
• A rating scale is a crude form of measurement that
includes a list of an ordered series of categories of
a variable, which are assumed to be based on an
underlying continuum. A numerical value is
assigned to each category.
• The Likert scale contains declarative statements
with a scale after each statement to determine the
opinion or attitude of a study participant.
• A semantic differential scale consists of two oppo-
site adjectives with a 7-point scale between them
and measures the connotative meaning of a concept
to a subject.
• The visual analogue scale, sometimes referred to as
magnitude scaling, is a 100-mm line with right-
angle stops at each end with bipolar anchors placed
beyond each end of the line. These end anchors
must cover the entire range of sensations possible
in the phenomenon being measured.
• The Delphi technique measures the judgments of a
group of experts to assess priorities or make fore-
casts. It provides a means for researchers to obtain
the opinions of a wide variety of experts across the
United States without the need for the experts
to meet.
• A diary, which allows one to record an experience
shortly after an event, is more accurate than obtain-
ing the information through recall at an interview.
In addition, the reporting level of incidents is
higher, and one tends to capture the participant’s
immediate perception of situations.
• Nurse researchers are expanding their use of data
from existing databases to answer their research
questions and test their research hypotheses. Health
data are usually categorized into secondary data
and administrative data.
• The choice of tools for use in a particular study is
a critical decision that can have a major impact on
the significance of the study.
• The researcher first must conduct an extensive
search for existing tools. Once found, the tools
must be carefully evaluated. Tools that are selected
for a study need to be described in great detail in
the proposal and in the final research report for
publication.
• Scale construction is a complex procedure that
should not be undertaken lightly. Theories on which
scale construction is most frequently based include
classic test theory, item response theory, multidi-
mensional scaling, and unfolding theory. Most
existing instruments used in nursing research have
been developed through the use of classic test
theory.
• Translating a scale to another language is a complex
process that allows concepts among respondents of
different cultures to be compared if care is taken to
ensure that concepts have the same or similar
meanings across cultures.
REFERENCES
Akhtar-Danesh, N., Baumann, A., & Cordingley, L. (2008).
Q-methodology in nursing research: A promising method for the
study of subjectivity. Western Journal of Nursing Research, 30(6),
759–773.
Alaszewski, A. (2006). Using diaries for social research. London,
UK: Sage.
Aroian, K. J. & Wal, J. S. V. (2007). Measuring elders’ symptoms
with daily diaries and retrospective reports. Western Journal of
Nursing Research, 29(3), 322–337.
Bartlett, J. W. & Frost, C. (2008). Reliability, repeatability and
reproducibility: Analysis of measurement errors in continuous
variables. Ultrasound Obstetric Gynecology, 31(4), 466–475.
Beckie, T. M., Beckstead, J. W., Schocken, D. D., Evans, M. E., &
Fletcher, G. F. (2011). The effects of a tailored cardiac rehabilita-
tion program on depressive symptoms in women: A randomized
clinical trial. International Journal of Nursing Studies, 48(1),
3–12.
Beckie, T. M., Fletcher, G. F., Beckstead, J. W., Schocken, D. D.,
& Evans, M. E. (2008). Adverse baseline physiological and psy-
chosocial profiles of women enrolled in a cardiac rehabilitation
clinical trial. Journal of Cardiopulmonary Rehabilitation & Pre-
vention, 28(1), 52–60.
Berdie, D. R., Anderson, J. F., & Niebuhr, M. A. (1992). Question-
naires: Design and use (2nd ed.). Metuchen, NJ: Scarecrow Press.
Bhengu, B. R., Ncama, B. P., McInerney, P. A., Wantland, D. J.,
Nicholas, P. D., Corless, I. B., et al. (2011). Symptoms experi-
enced by HIV-infected individuals on antiretroviral therapy in
KwaZulu-Natal, South Africa. Applied Nursing Research, 24(1),
1–9.
448 UNIT TWO  The Research Process
Bialocerkowski, A., Klupp, N., & Bragge, P. (2010). Research
methodology series: How to read and critically appraise a reli-
ability article. International Journal of Therapy & Rehabilitation,
17(3), 114–120.
Borg, J. & Groenen, P. J. (2010). Modern multidimensional scaling:
Theory and application (2nd ed.). New York, NY: Springer.
Brener, N. D., Kann, L., Kinchen, S. A., Grunbaum, J. A., Whalen,
L., Eaton, D.; Department of Health and Human Services, Centers
for Disease Control and Prevention. (2004). Methodology of the
Youth Risk Surveillance System. Morbidity & Mortality Weekly
Report, 53(Sept. 24), 1–13.
Briggs, C. L. (1986). Learning how to ask: A sociolinguistic
appraisal of the role of the interview in social science research.
Cambridge, MA: Cambridge University Press.
Brown, S. J. (2009). Evidence-based nursing: The research-practice
connection. Boston, MA: Jones & Bartlett.
Burman, M.E. (1995). Health diaries in nursing research and prac-
tice. Image Journal of Nursing Scholarship, 27(2), 147–152.
Cazzell, M. (2010). College student risk behavior: The implications
of religiosity and impulsivity. Ph.D. dissertation, The University
of Texas at Arlington, United States: Texas. Proquest Dissertations
& Theses. (Publication No. AAT 3391108).
Cazzell, M. (personal communication, March 20, 2011). Self-report
college student risk behavior questionnaire. The University of
Texas at Arlington, Arlington, TX.
Chase, C. A. (2011). An intergenerational e-mail pal project on
attitudes of college students toward older adults. Educational Ger-
ontology, 37(1), 27–37.
Clinical and Laboratory Standards Institute (CLSI). (2011). Harmo-
nized terminology database. Retrieved from http://www.clsi.org/
Content/NavigationMenu/Resources/HarmonizedTerminology
Database/Harmonized_Terminolo.htm.
Converse, J. M., & Presser, S. (1986). Survey questions: Handcraft-
ing the standardized questionnaire. Newbury Park, CA: Sage.
Coombs, C. H. (1950). Psychological scaling without a unit of
measurement. Psychological Review, 57(3), 145–158.
Couper, M. R. (1984). The Delphi technique: Characteristics and
sequence model. Advances in Nursing Science, 7(1), 72–77.
Cowan, M. J., Heinrich, J., Lucas, M., Sigmon, H., & Hinshaw, A.
S. (1993). Integration of biological and nursing sciences: A
10-year plan to enhance research and training. Research in
Nursing & Health, 16(1), 3–9.
Craig, J. V., & Smyth, R. L. (2012). The evidence-base practice
manual for nurses (3rd ed.). Edinburgh, UK: Churchill
Livingstone.
Crocker, L., & Algina, J. (1986). Introduction to classical modern
test theory. New York, NY: Holt, Rinehart & Winston.
Dariel, O. P., Wharrad, H., & Windle, R. (2010). Developing
Q-methodology to explore staff views toward the use of technol-
ogy in nurse education. Nurse Researcher, 18(1), 58–71.
Dennis, K. E. (1986). Q methodology: Relevance and application
to nursing research. Advances in Nursing Science, 8(3), 6–17.
Department of Education Genomic Science. (2011). Human
Genome Project information. Retrieved from http://www.ornl.gov/
sci/techresources/Human_Genome/home.shtml.
DeVon, H. A., Block, M. E., Moyle-Wright, P., Ernst, D. M.,
Hayden, S. J., Lazzara, D. J., et al. (2007). A psychometric toolbox
for testing validity and reliability. Journal of Nursing Scholarship,
39(2), 155–164.
Dillman, D. A., Smyth, J. D., & Christian, L. M. (2009). Internet,
mail, and mixed-mode surveys: The tailored design method.
Hoboken, NJ: John Wiley & Sons.
Dillon, J. T. (1990). The practice of questioning. New York, NY:
Routledge.
Doran, D. M. (2011). Nursing outcomes: The state of the science
(2nd ed.). Sudbury, MA: Jones & Bartlett Learning.
Dubbert, P. M., White, J. D., Grothe, K. B., O’Jile, J., & Kirchner,
K. A. (2006). Physical activity in patients who are severely men-
tally ill: Feasibility of assessment for clinical and research appli-
cations. Archives of Psychiatric Nursing, 20(5), 205–209.
Dunn, S. L., Corser, W., Stommel, M., & Holmes-Rovner, M.
(2006). Hopelessness and depression in the early recovery period
after hospitalization for acute coronary syndrome. Journal of Car-
diopulmonary Rehabilitation, 26(3), 152–159.
Fawcett, J., & Garity, J. (2009). Evaluating research for evidence-
based nursing practice. Philadelphia: F.A. Davis.
Foddy, W. H. (1993). Constructing questions for interviews and
questionnaires: Theory and practice in social research. Cam-
bridge, UK: Cambridge University Press.
Fowler, F. J. (1990). Standardized survey interviewing: Minimizing
interviewer-related error. Newbury Park, CA: Sage.
Gagliese, L., Weizblit, N., Ellis, W., & Chan, V. W. (2005). The
measurement of postoperative pain: A comparison in younger and
older surgical patients. Pain, 117(3), 412–420.
Gelinas, C., Choiniere, M., Ranger, M., Denault, A., Deschamps,
A., & Johnston, C. (2010). Toward a new approach for the detec-
tion of pain in adult patients undergoing cardiac surgery: Near-
infrared spectroscopy—a pilot study. Heart & Lung, 39(6),
485–943.
Gift, A. G. (1989). Visual analogue scales: Measurement of subjec-
tive phenomena. Nursing Research, 38(5), 286–288.
Gorden, R. L. (1987). Interviewing: Strategy, techniques, and
tactics. Chicago, IL: Dorsey Press.
Gorden, R. L. (1998). Basic interviewing skills. Chicago, IL: Dorsey
Press.
Grove, S. K. (2007). Statistics for health care research: A practical
workbook. Philadelphia, PA: Saunders.
Hambleton, R. K., & Swaminathan, H. (2010). Item response
theory: Principles and applications. Boston, MA: Kluwer Aca-
demic Publishers Group.
Harralson, T. L. (2007). Factors influencing delay in seeking treat-
ment for acute ischemic symptoms among lower income, urban
women. Heart & Lung, 36(2), 96–104.
Hockenberry, M. J., & Wilson, D. (2009). Wong’s essentials of
pediatric nursing (8th ed.). St. Louis, MO: Mosby.
Holaday, B., & Turner-Henson, A. (1989). Response effects in
surveys with school-age children. Nursing Research, 38(4),
248–250.
Holzemer, W. L., Henry, S. B., Nokes, K. M., Corless, I. B., Brown,
M. A., Powell-Cope, G. M., et al. (1999) Validation of the Sign
and Symptom Checklist for persons with HIV disease (SSC-HIV).
Journal of Advanced Nursing, 30(5), 1041–1049.
Holzemer, W. L., Hudson, A., Kirksey, K. M., Hamilton, M. J., &
Bakken, S. (2001). The revised Sign and Symptom Checklist for
 CHAPTER 17  Measurement Methods Used in Developing Evidence-Based Practice 449
HIV (SSC-HIVrev). Journal of the Association of Nurses in AIDS
Care, 12(1), 60–70.
Hulin, C. L., Drasgow, F., & Parsons, C. K. (1983). Item response
theory: Application to psychological measurement. Homewood,
IL: Dow Jones-Irwin.
International Organization for Standardization. (ISO). (2011).
Standards development. Retrieved from http://www.iso.org/iso/
standards_development.htm.
Johantgen, M. (2010). Using existing administrative and national
databases. In C. F. Waltz, O. L. Strickland, & E. R. Lenz (Eds.),
Measurement in nursing and health research (4th ed., pp. 241–
250). New York, NY: Springer Publishing Company.
Jones, D. J., Munro, C. L., Grap, M. J., Kitten, T., & Edmond, M.
(2010). Oral care and bacteremia risk in mechanically ventilated
adults. Heart & Lung, 39(6S), S57–S65.
Jones, J. M. (2004). Nutritional methodology: Development of a
nutritional screening or assessment tool using a multivariate tech-
nique. Nutrition, 20(3), 298–306.
Klein, D. G., Dumpe, M., Katz, E., & Bena, J. (2010). Pain assess-
ment in the intensive care unit: Development and psychometric
testing of the nonverbal pain assessment tool. Heart & Lung,
39(6), 521–528.
Kruskal, J. B., & Wish, M. (1990). Multidimensional scaling.
Newbury Park, CA: Sage.
Lake, E. T., Shang, J., Klaus, S., & Dunton, N. E. (2010).
Patient falls: Association with hospital Magnet status and nursing
unit staffing. Research in Nursing & Health, 33(5), 413–425.
Lee, K. A., & Kieckhefer, G. M. (1989). Measuring human responses
using visual analogue scales. Western Journal of Nursing
Research, 11(1), 128–132.
Lim, M., Sacks-Davis, R., Aitken, C. K., Hocking, J. S., & Hellard,
M. E. (2010). Randomized controlled trial of paper, online, and
SMS diaries for collecting sexual behavior information from
young people. Journal of Epidemiology & Community Health,
64(10), 885–889.
Lindeman, C. A. (1975). Delphi survey of priorities in clinical
nursing research. Nursing Research, 24(6), 434–441.
Lynn, M. R. (1989). Instrument reliability: How much needs to be
published? Heart & Lung, 18(4), 421–423.
Marshall, C., & Rossman, G. B. (2011). Designing qualitative
research (5th ed.). Thousand Oaks, CA: Sage.
McKeown, B., & Thomas, D. (1988). Q methodology. Newbury
Park, CA: Sage.
McLaughlin, P. (1990). How to interview: The art of asking ques-
tions (2nd ed.). North Vancouver, BC: International Self-Counsel
Press.
Melnyk, B. M., & Fineout-Overholt, E. (2011). Evidence-based
practice in nursing and healthcare: A guide to best practice (2nd
ed.). Philadelphia, PA: Lippincott Williams & Wilkins.
Mishler, E. G. (1986). Research interviewing: Context and narra-
tive. Cambridge, MA: Harvard University Press.
Mitchell, M. P. (1998). Nursing education planning: A Delphi study.
Journal of Nursing Education, 37(7), 305–307.
Munhall, P. L. (2012). Nursing research: A qualitative perspective
(5th ed.). Miami, FL: Jones & Bartlett.
National Institute of Nursing Research (NINR). (2012). Summer
Genetics Institute (SGI). Retrieved from http://www.ninr.nih
.gov/Training/TrainingOpportunitiesIntramural/Summer
GeneticsInstitute.
National Institute of Nursing Research (NINR). (2011). Strategic
plan National Institute of Nursing Research: Areas of research
emphasis. Retrieved from http://www.ninr.nih.gov/AboutNINR/
NINRMissionandStrategicPlan/.
Ng, K., Wong, S., Lim, S., & Goh, Z. (2010). Evaluation of the
Cadi ThermoSENSOR wireless skin-contact thermometer against
ear and axillary temperatures in children. Journal of Pediatric
Nursing, 25(3), 176–186.
Nicholl, H. (2010). Diaries as a method of data collection in
research. Paediatric Nursing, 22(7), 16–20.
Norbeck, J. S. (1985). What constitutes a publishable report
of instrument development? Nursing Research, 34(6), 380–381.
Nunnally, J. C., & Bernstein, I. H. (1994). Psychometric theory (3rd
ed.). New York, NY: McGraw-Hill.
Osgood, C. E., Suci, G. J., & Tannenbaum, P. H. (1957). The
measurement of meaning. Urbana, IL: University of Illinois
Press.
Pugh, L. C., & DeKeyser, F. G. (1995). Use of physiologic variables
in nursing research. Image Journal of Nursing Scholarship, 27(4),
273–276.
Radloff, L. S. (1977). The CES-D scale: A self-report depression
scale for research in the general population. Applied Psychologi-
cal Measures, 1, 385–394.
Raven, E. E., Haverkamp, D., Sierevelt, I. N., Van Montfoort,
D. O., Poll, R. G., Blankevoort, L., et al. (2008). Construct valid-
ity and reliability of the disability of arm, shoulder, and hand
questionnaire for upper extremity complaints in rheumatoid
arthritis. Journal of Rheumatology, 35(12), 2334–2338.
Rudy, E., & Grady, P. (2005). Biological researchers: Building
nursing science. Nursing Outlook, 53(2), 88–94.
Ryan-Wenger, N. A. (2010). Evaluation of measurement precision,
accuracy, and error in biophysical data for clinical research and
practice. In C. F. Waltz, O. L. Strickland, & E. R. Lenz (Eds.),
Measurement in nursing and health research (4th ed., pp. 371–
383). New York, NY: Springer Publishing Company.
Saris, W. E. (1991). Computer-assisted interviewing. Newbury
Park, CA: Sage.
Saris, W. E., & Gallhofer, I. N. (2007). Design, evaluation, and
analysis of questionnaires for survey research. Hoboken, NJ: John
Wiley & Son.
Scholz, J., Knoll, N., Sniehotta, F. F., & Schwarzer, R. (2006).
Physical activity and depressive symptoms in cardiac rehabilita-
tion: Long-term effects of a self-management intervention. Social
Science & Medicine, 62(12), 3109–3120.
Sennott-Miller, L., Murdaugh, C, & Hinshaw, A. S. (1988). Magni-
tude estimation: Issues and practical applications. Western Journal
of Nursing Research, 10(4), 414–424.
Shelley, S. I. (1984). Research methods in nursing and health.
Boston, MA: Little, Brown.
Silfen, M. E., Manibo, A. M., McMahon, D. J., Levine, L. S.,
Murphy, A. R., & Oberfield, S. E. (2001). Comparison of simple
measures of insulin sensitivity in young girls with premature
adrenarche: The fasting glucose to insulin ratio may be a simple
and useful measure. Journal of Clinical Endocrinology & Metab-
olism, 86(6), 2863–2868.
450 UNIT TWO  The Research Process
Simpson, S. H. (1989). Use of Q-sort methodology in cross-cultural
nutrition and health research. Nursing Research, 38(5),
289–290.
Smith, A. E., Annesi, J. J., Walsh, A. M., Lennon, V., & Bell, R. A.
(2010). Association of changes in self-efficacy, voluntary physical
activity, and risk factors for type 2 diabetes in a behavioral treat-
ment for obese preadolescents: A pilot study. Journal of Pediatric
Nursing, 25(5), 393–399.
Spector, P. E. (1992). Summated rating scale construction: An intro-
duction. Newbury Park, CA: Sage.
Stone, K. S., & Frazier, S. K. (2010). Measurement of physiological
variables using biomedical instrumentation. In C. F. Waltz, O. L.
Strickland, & E. R. Lenz (Eds.), Measurement in nursing and
health research (4th ed., pp. 335–370). New York, NY: Springer
Publishing Company.
Swardfager, W., Herrmann, N., Dowlati, Y., Oh, P., Kiss, A., &
Lanctot, K. L. (2008). Relationship between cardiopulmonary
fitness and depressive symptoms in cardiac rehabilitation patients
with coronary artery disease. Journal of Rehabilitation Medicine,
40(3), 213–218.
Tetting, D. W. (1988). Q-sort update. Western Journal of Nursing
Research, 10(6), 757–765.
Thomas, S. J. (2004). Using web and paper questionnaires for data-
based decision making: From design to interpretation of the
results. Thousand Oaks, CA: Corwin Press.
Turner, C. F., Rogers, S. M., Hendershot, T. P., Miller, H. G., &
Thornberry, J. P. (1996). Improving representation of linguistic
minorities in health surveys. Public Health Reports, 111(3),
276–279.
U.S. Department of Health and Human Services. (2011). Health
information privacy. Retrieved from http://www.hhs.gov/ocr/
privacy/.
Vernon, W. (2009). The Delphi technique: A review. International
Journal of Therapy & Rehabilitation, 16(2), 69–76.
Waltz, C. F., Strickland, O. L., & Lenz, E. R. (2010). Measurement
in nursing and health research (4th ed.). New York, NY: Springer
Publishing Company.
Wewers, M. E., & Lowe, N. K. (1990). A critical review of visual
analogue scales in the measurement of clinical phenomena.
Research in Nursing & Health, 13(4), 227–236.
Wilkes, L., Mohan, S., Luck, L., & Jackson, D. (2010). Develop-
ment of a violence tool in the emergency hospital setting. Nurse
Researcher, 17(4), 70–82.
Winkelman, C., Norman, D., Maloni, J. A., & Kless, J. R. (2008).
Pain measurement during labor: Comparing the visual analog
scale with dermatome assessment. Applied Nursing Research,
21(2), 104–109.
Yu, D. S. F., Lee, D. T. F., & Woo, J. (2004). Issues and challenges
of instrument translation. Western Journal of Nursing Research,
26(3), 307–320.
UNIT THREE
Putting It All Together for Evidence-Based Health Care
Critical Appraisal of Nursing Studies
  http://evolve.elsevier.com/Grove/practice/
T
he nursing profession continues to strive for
evidence-based practice, which includes criti-
cally appraising studies, synthesizing research
findings, and applying sound scientific evidence in
practice. Researchers also critically appraise studies in
a selected area, develop a summary of current knowl-
edge, and identify areas for future studies. Critically
appraising research is essential for evidence-based
nursing practice and the conduct of future research.
The critical appraisal of research involves a system-
atic, unbiased, careful examination of all aspects of
studies to judge their strengths, weaknesses, meaning,
and significance. The ability of a nurse to appraise
studies critically is based on the nurse’s previous
research experience and knowledge of a topic. To
conduct a critical appraisal of research, one must
possess analysis and logical reasoning skills to
examine the credibility and integrity of a study. This
chapter provides a background for critically apprais-
ing studies in nursing and other healthcare disciplines.
The expanding roles of nurses in conducting critical
appraisals of research are addressed. Detailed guide-
lines are provided to direct you in critically appraising
both quantitative and qualitative studies.
Evolution of Critical Appraisal
of Research in Nursing
The process for critically appraising research has
evolved gradually in nursing from a few to now many
nurses who are prepared to conduct comprehensive,
scholarly critiques. During the 1940s and 1950s, pre-
sentations of nursing research were followed by cri-
tiques of the studies. These critiques often focused on
the weaknesses or limitations of the studies and tended
18
CHAPTER
to be harsh and traumatic for the researcher (Meleis,
2007). As a consequence of these early unpleasant
experiences, nurse researchers began to protect and
shelter their nurse scientists from the threat of criti-
cism. Public critiques, written or verbal, were rare in
the 1960s and 1970s. Nurses responding to research
presentations focused on the strengths of studies, and
the limitations were either critically not mentioned or
were minimized. Thus, the effects of the limitations
on the meaning, validity, and significance of studies
were often lost.
Incomplete critiques or the absence of critiques
may have served a purpose as nurses gained basic
research skills. However, the nursing discipline
has moved past this point, and it is recognized that
comprehensive critical appraisals of research are
essential to strengthen the scientific investigations
needed for evidence-based practice (EBP; Brown,
2009; Craig & Smyth, 2012; Fawcett & Garity, 2009).
As a result of advances in the nursing profession over
the last 30 years, many nurses now have the educa-
tional preparation and expertise to conduct critical
appraisals of research. Nursing research textbooks
provide detailed information on the critical appraisal
process. Skills in critical appraisal are introduced at
the baccalaureate level of nursing education and are
expanded at the master’s and doctoral levels. Spe-
cialty organizations provide workshops on the critical
appraisal process to promote the use of scientific evi-
dence in practice.
The critical appraisal of studies is essential for the
development and refinement of nursing knowledge.
Nurses need these skills to examine the meaning
and credibility of study findings and to ask searching
questions. Was the methodology of a study sound to
produce credible findings? Are the findings an
451
452 UNIT THREE  Putting It All Together for Evidence-Based Health Care
accurate reflection of reality? Do they increase our
understanding of the nature of phenomena that are
important in nursing? Are the findings from the present
study consistent with findings from previous studies?
The answers to these questions require careful exami-
nation of the research problem and purpose, the theo-
retical or philosophical basis of the study, and the
study’s methodology. Not only must the mechanics of
conducting the study be evaluated, but also the abstract
and logical reasoning the researchers used to plan
and implement the study (Fawcett & Garity, 2009;
Munhall, 2012). If the reasoning process used to
develop a study has flaws, there are probably flaws in
interpreting the meaning of the findings, decreasing
the credibility of the study.
All studies have flaws, but if all flawed studies were
discarded, there would be no scientific knowledge
base for practice. In fact, science itself is flawed.
Science does not completely or perfectly describe,
explain, predict, or control reality. However, improved
understanding and an increased ability to predict and
control phenomena depend on recognizing the flaws
in studies and in science. New studies can then be
planned to minimize the flaws or limitations of earlier
studies. A researcher must critically analyze previous
studies to determine their limitations and then inter-
pret the study findings in light of those limitations. The
limitations can lead to inaccurate data, inaccurate out-
comes of analysis, and decreased ability to generalize
the findings. You must decide if a study is too flawed
to be used in a systematic review of knowledge in an
area. Although we recognize that knowledge is not
absolute, we need to have confidence in the research
evidence synthesized for practice.
All studies have strengths as well as limitations.
Recognition of these strengths is also essential to the
generation of sound research evidence for practice. If
only weaknesses are identified, nurses might discount
the value of studies and refuse to invest time in reading
and examining research. The continued work of the
researcher also depends on recognizing the strengths
of the study. If no study is good enough, why invest
time conducting research? The strong points of a study,
added to the strong points from multiple other studies,
slowly build solid research evidence for practice.
When Are Critical Appraisals of Research
Implemented in Nursing?
In general, research is critically appraised to broaden
understanding, summarize knowledge for practice,
and provide a knowledge base for future studies. In
addition, critical appraisals are often conducted after
verbal presentations of studies, after a published
research report, for an abstract section for a confer-
ence, for article selection for publication, and for
evaluation of research proposals for implementation
or funding. Nursing students, practicing nurses, nurse
educators, and nurse researchers all are involved in the
critical appraisal of research.
Critical Appraisal of Studies by Students
In nursing education, conducting a critical appraisal
of a study is often seen as a first step in learning the
research process. Part of learning this process is being
able to read and comprehend published research
reports. However, conducting a critical appraisal of a
study is not a basic skill, and the content presented in
previous chapters is essential for implementing this
process. Nurses usually acquire basic knowledge of
the research process and critical appraisal skills early
in their nursing education. Advanced analysis skills
are usually taught at the master’s and doctoral levels.
Performing a critical appraisal of a study involves
(1) identifying the elements or steps of the study,
(2) determining the study strengths and limitations,
and (3) evaluating the credibility and meaning of
study findings for nursing knowledge and practice. By
critically appraising studies, students expand their
analysis skills, strengthen their knowledge base, and
increase their use of research evidence in practice.
Critical Appraisal of Research by Practicing Nurses
Practicing nurses need to appraise studies critically so
that their practice is based on current research evi-
dence and not tradition and trial and error (Brown,
2009; Melnyk & Fineout-Overholt, 2011). Nursing
actions must be updated in response to the current
evidence that is generated through research and theory
development. Practicing nurses need to design methods
for remaining current in their practice areas. Reading
research journals and posting or emailing current
studies at work can increase nurses’ awareness of
study findings but are insufficient for the purposes of
critical appraisal. Nurses need to question the quality
of the studies and the credibility of the findings and
share their concerns with other nurses. For example,
nurses may form a research journal club in which
studies are presented and critically appraised by
members of the group (Gloeckner & Robinson, 2010).
Skills in critical appraisal of research enable practic-
ing nurses to synthesize the most credible, significant,
and appropriate evidence for use in their practice. EBP
is essential in agencies either seeking or maintaining
Magnet status. The Magnet Recognition Program®
 CHAPTER 18  Critical Appraisal of Nursing Studies 453
was developed by the American Nurses Credentialing
Center (ANCC, 2012) to recognize healthcare organi-
zations that provide nursing excellence with care
based on the most current research evidence (see
http://www.nursecredentialing.org/Magnet/Program
Overview.aspx).
Critical Appraisal of Research by Nurse Educators
Educators critically appraise research to expand their
knowledge base and to develop and refine the educa-
tional process. The careful analysis of current nursing
studies provides a basis for updating curriculum
content for use in clinical and classroom settings. Edu-
cators act as role models for their students by examin-
ing new studies, evaluating the information obtained
from research, and indicating what research evidence
to use in practice. In addition, educators collaborate
in the conduct of studies, which requires a critical
appraisal of previous relevant research.
Critical Appraisal of Studies by Nurse Researchers
Nurse researchers critically appraise previous research
to plan and implement their next study. Many research-
ers have a program of research in a selected area, and
they update their knowledge base by critiquing new
studies in this area. The outcomes of these appraisals
influence the selection of research problems and pur-
poses, the implementation of research methodologies,
and the interpretations of findings in future studies.
The critical appraisal and synthesis of previous studies
for the literature review section in a research proposal
or report were described in Chapter 6.
Critical Appraisal of Research Presentations
and Publications
Critical appraisals following research presentations
can assist researchers in identifying the strengths and
weaknesses of their studies and generating ideas
for further research. Participants listening to study
critiques might gain insight into the conduct of
research. Experiencing the critical appraisal process
can increase the ability of participants to evaluate
studies and judge the usefulness of the research evi-
dence for practice.
At the present time, at least two nursing research
journals, Scholarly Inquiry for Nursing Practice:
An International Journal and Western Journal of
Nursing Research, include commentaries after the
research articles. In these journals, other researchers
critically appraise the authors’ studies, and the authors
have a chance to respond to these comments. Pub-
lished research critical appraisals often increase the
reader’s understanding of the study and the quality of
the study findings (American Psychological Associa-
tion [APA], 2010; Pyrczak, 2008). Another, more
informal critique of a published study might appear in
a letter to the editor, in which readers have the oppor-
tunity to comment on the strengths and weaknesses of
published studies by writing to the journal editor.
Critical Appraisal of Abstracts
for Conference Presentations
One of the most difficult types of critical appraisal is
examining abstracts. The amount of information avail-
able is usually limited because many abstracts are
restricted to 100 to 250 words. Nevertheless, reviewers
must select the best-designed studies with the most
significant outcomes for presentation at nursing confer-
ences. This process requires an experienced researcher
who needs few cues to determine the quality of a study.
Critical appraisal of an abstract usually addresses the
following criteria: (1) appropriateness of the study for
the program; (2) completeness of the research project;
(3) overall quality of the study problem, purpose, meth-
odology, and results; (4) contribution of the study to
the knowledge base of nursing; (5) contribution of the
study to nursing theory; (6) originality of the work (not
previously published); (7) implication of the study
findings for practice; and (8) clarity, conciseness, and
completeness of the abstract (APA, 2010).
Critical Appraisal of Research Articles
for Publication
Nurse researchers who serve as peer reviewers for
professional journals evaluate the quality of research
articles submitted for publication. The role of these
scientists is to ensure that the studies accepted for
publication are well designed and contribute to the
body of knowledge. Most of these reviews are con-
ducted anonymously so that friendships or reputa-
tions do not interfere with the selection process
(Pyrczak, 2008; Tilden, 2002). In most refereed jour-
nals, the experts who examine the research report
have been selected from an established group of peer
reviewers. Their comments or summaries of their
comments are sent to the researcher. The editor also
uses these comments to make selections for publica-
tion. The process for publishing a study is described
in Chapter 27.
Critical Appraisal of Research Proposals
Critical appraisals of research proposals are conducted
to approve student research projects; to permit data
collection in an institution; and to select the best
454 UNIT THREE  Putting It All Together for Evidence-Based Health Care
studies for funding by local, state, national, and inter-
national organizations and agencies. The process
researchers use to seek the approval to conduct a study
is presented in Chapter 28. The peer review process
in federal funding agencies involves an extremely
complex critical appraisal. Nurses are involved in this
level of research review through the national funding
agencies, such as the National Institute of Nursing
Research (NINR, 2012), National Institutes of Health,
and the Agency for Healthcare Research and Quality.
Some of the criteria used to evaluate the quality of a
proposal for possible funding include the (1) signifi-
cance of the research problem and purpose for nursing,
(2) appropriate use of methodology for the types of
questions that the research is designed to answer, (3)
appropriate use and interpretation of analysis proce-
dures, (4) evaluation of clinical practice and forecast-
ing of the need for nursing or other appropriate
interventions, and (5) construction of models to direct
the research and interpret the findings. The NINR
(2012) website (http://www.ninr.nih.gov/ResearchAnd
Funding/) provides details on grant development and
research funding, and Chapter 29 focuses on seeking
funding for research.
Nurses’ Expertise in Critical
Appraisal of Research
Conducting a critical appraisal of a study is a complex
mental process that is stimulated by raising questions.
The level of critique conducted is influenced by the
sophistication of the individual appraising the study
(Table 18-1). The initial critical appraisal of research
by an undergraduate student often involves the iden-
tification of the steps of the research process in a
quantitative study. Some baccalaureate programs
include more in-depth research courses that include
critical appraisals of the steps of quantitative studies
and identification of the aspects of qualitative studies.
A critical appraisal of research conducted by a student
at the master’s level usually involves description of
study strengths and weaknesses and evaluation of the
credibility and meaning of the study findings for
nursing knowledge and practice. Critical appraisals
might focus on quantitative, qualitative, and outcomes
studies.
At the doctoral level, students often critically
appraise several studies in an area of interest and
perform a complex synthesis of the research findings
to determine the current empirical knowledge base for
the phenomenon. These complex syntheses of quanti-
tative, qualitative, outcomes, and intervention research
TABLE 18-1  Educational Level and Expected
Level of Expertise in Critical
Appraisal of Research
Educational Expected Level of Expertise in
Level Critical Appraisal of Research
Baccalaureate Identify the steps of the quantitative
research process in a study
Identify the elements of a qualitative
study
Master’s Determine study strengths and weaknesses
in quantitative and qualitative studies
Evaluate the credibility and meaning of a
study and its contribution to nursing
knowledge and practice
Doctorate or Synthesize multiple studies in systematic
postdoctorate reviews, meta-analyses, meta-syntheses,
and mixed-methods systematic reviews
include (1) systematic review of research, (2) meta-
analysis, (3) meta-synthesis, and (4) mixed methods
systematic review (see Table 18-1). These summaries
of current research evidence are essential for provid-
ing EBP and directing future research (Craig & Smyth,
2012; Higgins & Green, 2008; Sandelowski & Barroso,
2007; Whittemore, 2005). Definitions of these types
of complex syntheses are presented in Chapter 2, and
Chapter 19 provides guidelines for critically apprais-
ing and conducting these syntheses.
The major focus of this chapter is conducting
critical appraisals of quantitative and qualitative
studies. These critical appraisals involve implement-
ing some initial guidelines that are outlined in Box
18-1. These guidelines stress the importance of
examining the expertise of the authors, reviewing
the entire study, addressing the strengths and weak-
nesses of the study, and evaluating the credibility of
the study findings (Fawcett & Garity, 2009; Marshall
& Rossman, 2011; Munhall, 2012; Shadish, Cook, &
Campbell, 2002). These guidelines provide a basis
for the critical appraisal process for quantitative
research that is discussed in the next section and the
critical appraisal process for qualitative research dis-
cussed later.
Critical Appraisal Process for
Quantitative Research
The critical appraisal process for quantitative
research includes three steps: (1) identifying the steps
of the research process, (2) determining study strengths
 CHAPTER 18  Critical Appraisal of Nursing Studies 455
Box 18-1 Guidelines for Conducting Critical Appraisals of Quantitative
and Qualitative Research
1. Read and evaluate the entire study. A
research appraisal requires comprehension of
a study that includes the identification and
examination of all steps of the research
process.
2. Examine the research, clinical, and educational
background of the authors. The authors need
a clinical and scientific background that is
appropriate for the study conducted.
3. Examine the organization and presentation of
the research report. The title of the research
report needs to indicate clearly the focus of
the study. The report usually includes an
abstract, introduction, methods, results,
discussion, and references. The abstract of the
study needs to present the purpose of the
study clearly and to highlight the
methodology and major results. The body of
the report needs to be complete, concise,
clearly presented, and logically organized. knowledge. Discuss the consistency of the
The references need to be complete and
presented in a consistent format.
4. Identify the strengths and weaknesses of
a study. All studies have strengths and
and weaknesses, and (3) evaluating the credibility and
meaning of a study to nursing knowledge and practice.
These steps occur in sequence, vary in depth, and
presume accomplishment of the preceding steps.
However, an individual with critical appraisal experi-
ence frequently performs several steps of this process
simultaneously.
This section includes the three steps of the quantita-
tive research critical appraisal process and provides
relevant questions for each step. These questions are
not comprehensive but have been selected as a means
for stimulating the logical reasoning and analysis nec-
essary for conducting a study review. Persons experi-
enced in the critical appraisal process formulate
additional questions as part of their reasoning pro-
cesses. We cover the identification of the steps of the
research process separately because persons who are
new to critical appraisal start with this step. The ques-
tions for determining the study strengths and weak-
nesses are covered together because this process
occurs simultaneously in the mind of the person con-
ducting the critical appraisal. Evaluation is covered
separately because of the increased expertise needed
to perform this step.
weaknesses, and you can use the questions
in this chapter to facilitate identification of
them. Address the quality of the problem,
purpose, methodology, results, and findings
of quantitative and qualitative studies.
5. Provide specific examples of the strengths and
weaknesses of a study. These examples
provide a rationale and documentation for
your critical appraisal of the study.
6. Be objective and realistic in identifying a
study’s strengths and weaknesses. Do not
be overly critical when identifying the
weaknesses of a study or overly flattering
when identifying the strengths.
7. Suggest modifications for future studies.
Modifications should increase the strengths
and decrease the weaknesses in the study.
8. Evaluate the study. Indicate the overall quality
of the study and its contribution to nursing
findings of this study with the findings of
previous studies. Discuss the need for further
research and the potential implications of the
findings for practice.
Step I: Identifying the Steps of the Research
Process in Studies
Initial attempts to comprehend research articles are
often frustrating because the terminology and stylized
manner of the report are unfamiliar. Identification of
the steps of the research process in a quantitative study
is the first step in critical appraisal. It involves under-
standing the terms and concepts in the report; identify-
ing study elements; and grasping the nature,
significance, and meaning of the study elements. The
following guidelines are presented to direct you in
identifying the elements or steps of a study.
Guidelines for Identifying the Steps
of the Research Process
The first step involves reviewing the abstract and
reading the study from beginning to end (see the
guidelines in Box 18-1). As you read, address the
following questions about the presentation of
the study: Does the title clearly identify the focus
of the study by including the major study variables
and the population? Does the title indicate the type
of study conducted—descriptive, correlational, quasi-
experimental, or experimental (Kerlinger & Lee,
456 UNIT THREE  Putting It All Together for Evidence-Based Health Care
2000; Shadish et  al., 2002)? Was the abstract clear?
Was the writing style of the report clear and concise?
Were the different parts of the research report plainly
identified (APA, 2010)? Were relevant terms defined?
You might underline the terms you do not understand
and determine their meaning from the glossary at
the end of this textbook. Read the article a second
time and highlight or underline each step of the quan-
titative research process. An overview of these steps
is presented in Chapter 3. To write a critical appraisal
identifying the study steps, you need to identify each
step of the research process concisely and respond
briefly to the following guidelines and questions:
I. Introduction
A. Describe the qualifications of the authors
to conduct the study, such as research
expertise, clinical experience, and
educational preparation. Doctoral
education, such as a PhD, provides
experience in conducting research. Have
the researchers conducted previous
studies, especially studies in this area?
Are the authors involved in clinical
practice or certified in their area of
clinical expertise?
B. Discuss the clarity of the article title
(type of study, variables, and population
identified).
C. Discuss the quality of the abstract
(includes purpose; highlights design,
sample, and intervention [if applicable];
and presents key results).
II. State the problem.
A. Significance of the problem
B. Background of the problem
C. Problem statement
III. State the purpose.
IV. Examine the literature review.
A. Are relevant previous studies and
theories described?
B. Are the references current? (Number and
percentage of sources in the last 10 years
and in the last 5 years?)
C. Are the studies described, critically
appraised, and synthesized (Fawcett &
Garity, 2009)?
D. Is a summary provided of the current
knowledge (what is known and not
known) about the research problem?
V. Examine the study framework or theoretical
perspective.
A. Is the framework explicitly expressed, or
must the reviewer extract the framework
from implicit statements in the
introduction or literature review?
B. Is the framework based on tentative,
substantive, or scientific theory? Provide
a rationale for your answer.
C. Does the framework identify, define, and
describe the relationships among the
concepts of interest? Provide examples
of this.
D. Is a model of the framework provided for
clarity? If a model is not presented,
develop one that represents the
framework of the study and describe it.
E. Link the study variables to the relevant
concepts in the model.
F. How is the framework related to the
body of knowledge of nursing (Alligood,
2010; Smith & Liehr, 2008)?
VI. List any research objectives, questions, or
hypotheses.
VII. Identify and define (conceptually and
operationally) the study variables or concepts
that were identified in the objectives,
questions, or hypotheses. If objectives,
questions, or hypotheses are not stated,
identify and define the variables in the study
purpose and the results section of the study. If
conceptual definitions are not found, identify
possible definitions for each major study
variable. Indicate which of the following
types of variables were included in the study.
A study usually includes independent and
dependent variables or research variables but
not all three types of variables.
A. Independent variables: Identify and
define conceptually and operationally.
B. Dependent variables: Identify and define
conceptually and operationally.
C. Research variables or concepts: Identify
and define conceptually and operationally.
VIII. Identify demographic variables and other
relevant terms.
IX. Identify the research design.
A. Identify the specific design of the study.
Draw a model of the design by using the
sample design models presented in
Chapter 11.
B. Does the study include a treatment or
intervention? If so, is the treatment
clearly described with a protocol and
consistently implemented, which
indicates intervention fidelity (Forbes,
2009; Mittlbock, 2008)?
 CHAPTER 18  Critical Appraisal of Nursing Studies 457

C. If the study has more than one group,
how were subjects assigned to groups
(Mittlbock, 2008; Shadish et al., 2002)?
D. Are extraneous variables identified and
controlled? Extraneous variables are
usually discussed as a part of quasi-
experimental and experimental studies
(Shadish et al., 2002).
E. Were pilot study findings used to design
this study? If yes, briefly discuss the
pilot and the changes made in this study
based on the pilot.
X. Describe the sample and setting.
A. Identify inclusion or exclusion sample or
eligibility criteria.
B. Identify the specific type of probability
or nonprobability sampling method that
was used to obtain the sample. Did the
researchers identify the sampling frame
for the study (Thompson, 2002)?
C. Identify the sample size. Discuss the
refusal rate and include the rationale for
refusal if presented in the article. Discuss
the power analysis if this process was
used to determine sample size (Aberson,
2010).
D. Identify the sample attrition (number and
percentage) and rationale for the study.
E. Identify the characteristics of the sample.
F. Discuss the institutional review board
approval. Describe the informed consent
process used in the study.
G. Identify the study setting, and indicate if
it is appropriate for the study purpose.
XI. Identify and describe each measurement
strategy used in the study. The following
table includes the critical information about
two measurement methods, the Beck Likert
scale to measure depression and the
physiological instrument to measure blood
pressure. Completing this table allows you to
cover essential measurement content for a
study (Waltz, Strickland, & Lenz, 2010).
A. Identify each study variable that was
measured.
B. Identify the name and author of each
measurement strategy.
C. Identify the type of each measurement
strategy (e.g., Likert scale, visual
analogue scale, physiological measure,
and existing database).
D. Identify the level of measurement
(nominal, ordinal, interval, or ratio)
achieved by each measurement method
used in the study (Grove, 2007).
E. Describe the reliability of each scale for
previous studies and this study. Identify
the precision of each physiological
measure (Bialocerkowski, Klupp, &
Bragge, 2010; DeVon et al., 2007).
F. Identify the validity of each scale and the
accuracy of physiological measures
(DeVon et al., 2007; Ryan-Wenger,
2010).

Name of
Measurement Type of
Variable Method/ Measurement Level of Reliability or
Measured Author Method Measurement Precision Validity or Accuracy
Depression Beck Depression Likert scale Interval Cronbach alpha of Construct validity: Content
Inventory/Beck 0.82-0.92 from validity from concept analysis,
previous studies and literature review, and reviews
0.84 for this study. of experts. Convergent validity
Reading level at 6th with Zung Depression Scale.
grade. Prediction validity of patients’
future depression episodes.
Successive use validity with
previous studies and this study.
Blood Omron BP Physiological Ratio Test-retest values of Documented accuracy of
pressure equipment/ measurement BP measurements in systolic and diastolic BPs to
(BP) Health Care method previous studies. BP 1 mm Hg by company
Equipment equipment new and developing Omron BP cuff.
Agency recalibrated every 50 Designated protocol for taking
BP readings in this BP. Average 3 BP readings to
study. Average 3 BP determine BP.
readings to
determine BP.
458 UNIT THREE  Putting It All Together for Evidence-Based Health Care
XII. Describe the procedures for data collection.
XIII. Describe the statistical techniques performed
to analyze study data.
A. List the statistical procedures conducted
to describe the sample.
B. Was the level of significance or alpha
identified? If so, indicate what it was
(0.05, 0.01, or 0.001).
C. Complete the following table with the
analysis techniques conducted in the
Purpose of Analysis
Description of subjects’ pulse rate
Difference between men and women in systolic and
diastolic blood pressures respectively
Differences of diet group, exercise group, and
comparison group for pounds lost by adolescents
Relationship of depression and anxiety in elderly adults
XIV. Describe the researcher’s interpretation of
findings.
A. Are the findings related back to the study
framework? If so, do the findings support
the study framework?
B. Which findings are consistent with the
expected findings?
C. Which findings were not expected?
D. Are the findings consistent with previous
research findings (Fawcett & Garity,
2009)?
XV. What study limitations did the researcher
identify?
XVI. How did the researcher generalize the
findings?
XVII. What were the implications of the findings
for nursing practice?
XVIII. What suggestions for further study were
identified?
XIX. Is the description of the study sufficiently
clear for replication?
Step II: Determining Study Strengths
and Weaknesses
The next step in critically appraising a quantitative
study requires determining the strengths and weak-
nesses of the study. To do this, you must have knowl-
edge of what each step of the research process should
be like from expert sources such as this textbook and
other research sources (Aberson, 2010; Bartlett &
Frost, 2008; Bialocerkowski et al., 2010; Borglin &
study: (1) identify the focus (description,
relationships, or differences) for each
analysis technique; (2) list the statistical
analysis technique performed; (3) list
the statistic; (4) provide the specific
results; and (5) identify the probability
(p) of the statistical significance
achieved by the result (Corty, 2007;
Grove, 2007).
Analysis Technique Statistic Results Probability (p)
Mean M 71.52
Standard deviation SD 5.62
Range Range 58-97
t-test t 3.75 0.001
t-test t 2.16 0.042
Analysis of variance F 4.27 0.04
Pearson correlation r 0.46 0.03
Richards, 2010; Burns & Grove, 2011; DeVon et al.,
2007; Doran, 2011; Fawcett & Garity, 2009; Grove,
2007; Houser, 2008; Morrison et al., 2009; Ryan-
Wenger, 2010; Santacroce, Maccarelli, & Grey, 2004;
Shadish et al., 2002; Thompson, 2002; Waltz et al.,
2010). The ideal ways to conduct the steps of the
research process are compared with the actual study
steps. During this comparison, you examine the extent
to which the researcher followed the rules for an ideal
study and identify the study elements that are strengths
or weaknesses.
You also need to examine the logical links connect-
ing one study element with another. For example, the
problem needs to provide background and direction
for the statement of the purpose. In addition, you need
to examine the overall flow of logic in the study. The
variables identified in the study purpose need to be
consistent with the variables identified in the research
objectives, questions, or hypotheses. The variables
identified in the research objectives, questions, or
hypotheses need to be conceptually defined in light of
the study framework. The conceptual definitions
provide the basis for the development of operational
definitions. The study design and analyses need to be
appropriate for the investigation of the study purpose
and for the specific objectives, questions, or hypoth-
eses (Fawcett & Garity, 2009). Most of the limitations
or weaknesses in a study result from breaks in logical
reasoning. For example, biases caused by sampling
and design impair the logical flow from design to
interpretation of findings (Borglin & Richards, 2010).
 CHAPTER 18  Critical Appraisal of Nursing Studies 459
The previous level of critical appraisal addressed con-
crete aspects of the study. During analysis, the process
moves to examining abstract dimensions of the study,
which requires greater familiarity with the logic
behind the research process and increased skill in
abstract reasoning.
You also need to gain a sense of how clearly the
researcher grasped the study situation and expressed
it. The clarity of the researchers’ explanation of study
elements demonstrates their skill in using and express-
ing ideas that require abstract reasoning. With this
examination of the study, you can determine which
aspects of the study are strengths and which are weak-
nesses and provide rationale and documentation for
your decisions.
Guidelines for Determining Study Strengths
and Weaknesses
The following questions were developed to assist you
in examining the different aspects of a study and deter-
mining if it is a strength or weakness. The intent is not
to answer each of these questions but to read the ques-
tions and make judgments about the elements or steps
in the study. You need to provide a rationale for your
decisions and document from relevant research sources
such as those listed in the previous section and in the
references at the end of this chapter. For example, you
might decide the study purpose is a strength because
it addresses the study problem, clarifies the focus of
the study, and is feasible to investigate (Burns &
Grove, 2011; Fawcett & Garity, 2009; Pyrczak, 2008).
I. Research problem and purpose
A. Is the problem sufficiently delimited in
scope so that it is researchable but not
trivial?
B. Is the problem significant to nursing and
clinical practice (Brown, 2009)?
C. Does the purpose narrow and clarify the
aim of the study?
D. Was this study feasible to conduct in terms
of money commitment; the researchers’
expertise; availability of subjects,
facilities, and equipment; and ethical
considerations?
II. Review of literature
A. Is the literature review organized to show
the progressive development of evidence
from previous research?
B. Is a theoretical knowledge base developed
for the problem and purpose?
C. Is a clear, concise summary presented of
the current empirical and theoretical
knowledge in the area of the study
(Craig & Smyth, 2012; Fawcett & Garity,
2009)?
D. Does the literature review summary
identify what is known and not known
about the research problem and provide
direction for the formation of the research
purpose?
III. Study framework
A. Is the framework presented with clarity? If
a model or conceptual map of the
framework is present, is it adequate to
explain the phenomenon of concern?
B. Is the framework linked to the research
purpose? If not, would another framework
fit more logically with the study?
C. Is the framework related to the body of
knowledge in nursing and clinical
practice?
D. If a proposition or relationship from a
theory is to be tested, is the proposition
clearly identified and linked to the study
hypotheses (Alligood, 2010; Fawcett &
Garity, 2009; Smith & Liehr, 2008)?
IV. Research objectives, questions, or hypotheses
A. Are the objectives, questions, or
hypotheses expressed clearly?
B. Are the objectives, questions, or hypotheses
logically linked to the research purpose?
C. Are hypotheses stated to direct the conduct
of quasi-experimental and experimental
research (Kerlinger & Lee, 2000; Shadish
et al., 2002)?
D. Are the objectives, questions, or
hypotheses logically linked to the concepts
and relationships (propositions) in the
framework (Fawcett & Garity, 2009; Smith
& Liehr, 2008)?
V. Variables
A. Are the variables reflective of the concepts
identified in the framework?
B. Are the variables clearly defined
(conceptually and operationally) and based
on previous research or theories (Smith &
Liehr, 2008)?
C. Is the conceptual definition of a variable
consistent with the operational definition?
VI. Design
A. Is the design used in the study the most
appropriate design to obtain the needed
data?
B. Does the design provide a means to
examine all the objectives, questions, or
hypotheses?
460 UNIT THREE  Putting It All Together for Evidence-Based Health Care
C. Is the treatment clearly described (Forbes,
2009)? Is the treatment appropriate for
examining the study purpose and
hypotheses? Does the study framework
explain the links between the treatment
(independent variable) and the proposed
outcomes (dependent variables) (Sidani &
Braden, 1998)? Was a protocol developed
to promote consistent implementation of
the treatment to ensure intervention fidelity
(Morrison et al., 2009)? Did the researcher
monitor implementation of the treatment to
ensure consistency (Santacroce et al.,
2004)? If the treatment was not
consistently implemented, what might be
the impact on the findings?
D. Did the researcher identify the threats to
design validity (statistical conclusion
validity, internal validity, construct validity,
and external validity) and minimize them
as much as possible (Shadish et al., 2002)?
E. Is the design logically linked to the
sampling method and statistical analyses?
F. If more than one group is used, do the
groups appear equivalent (Borglin &
Richards, 2010)?
G. If a treatment was implemented, were the
subjects randomly assigned to the
treatment group, or were the treatment and
comparison groups matched? Were the
treatment and comparison group
assignments appropriate for the purpose of
the study (Borglin & Richards, 2010)?
VII. Sample, population, and setting
A. Is the sampling method adequate to
produce a representative sample?
B. What are the potential biases in the
sampling method? Are any subjects
excluded from the study because of age,
socioeconomic status, or ethnicity without
a sound rationale (Borglin & Richards,
2010; Thompson, 2002)?
C. Did the sample include an understudied
population, such as young, elderly, or
minority subjects?
D. Were the sampling criteria (inclusion and
exclusion) appropriate for the type of study
conducted?
E. Is the sample size sufficient to avoid a type
II error? Was a power analysis conducted
to determine sample size? If a power
analysis was conducted, were the results of
the analysis clearly described and used to
determine the final sample size? Was the
attrition rate projected in determining the
final sample size (Aberson, 2010)?
F. Are the rights of human subjects
protected?
G. Is the setting used in the study typical of
clinical settings (Borglin & Richards,
2010)?
H. Was the refusal to participate rate a
problem? If so, how might this weakness
influence the findings?
I. Was sample attrition a problem? Did the
researchers provide a rationale for the
attrition of study participants? How did
attrition influence the final sample and the
study results and findings (Aberson, 2010;
Fawcett & Garity, 2009)?
VIII. Measurements
A. Do the measurement methods selected for
the study adequately measure the study
variables?
B. Are the measurement methods sufficiently
sensitive to detect small differences
between subjects? Should additional
measurement methods have been used to
improve the quality of the study outcomes
(Waltz et al., 2010)?
C. Do the measurement methods used in the
study have adequate validity and
reliability? What additional reliability or
validity testing is needed to improve the
quality of the measurement methods
(Bartlett & Frost, 2008; Bialocerkowski
et al., 2010; DeVon et al., 2007; Roberts &
Stone, 2004)?
D. Respond to the following questions, which
are relevant to the measurement
approaches used in the study:
1. Scales and questionnaires
(a) Are the instruments clearly
described?
(b) Are techniques to complete and
score the instruments provided?
(c) Are validity and reliability of the
instruments described (DeVon et al.,
2007)?
(d) Did the researcher reexamine the
validity and reliability of
instruments for the present sample?
(e) If the instrument was developed for
the study, is the instrument
development process described
(Waltz et al., 2010)?
 CHAPTER 18  Critical Appraisal of Nursing Studies 461
2. Observation
(a) Is what is to be observed clearly
identified and defined?
(b) Is interrater reliability described?
(c) Are the techniques for recording
observations described (Waltz et al.,
2010)?
3. Interviews
(a) Do the interview questions address
concerns expressed in the research
problem?
(b) Are the interview questions relevant
for the research purpose and
objectives, questions, or
hypotheses?
(c) Does the design of the questions
tend to bias subjects’ responses?
(d) Does the sequence of questions
tend to bias subjects’ responses
(Waltz et al., 2010)?
4. Physiological measures
(a) Are the physiological measures or
instruments clearly described
(Ryan-Wenger, 2010)? If
appropriate, are the brand names,
such as Space Labs or Hewlett-
Packard, of the instruments
identified?
(b) Are the accuracy, precision, and
error of the physiological
instruments discussed (Ryan-
Wenger, 2010)?
(c) Are the physiological measures
appropriate for the research purpose
and objectives, questions, or
hypotheses?
(d) Are the methods for recording data
from the physiological measures
clearly described? Is the recording
of data consistent?
IX. Data collection
A. Is the data collection process clearly
described (Fawcett & Garity, 2009;
Kerlinger & Lee, 2000)?
B. Are the forms used to collect data
organized to facilitate computerizing the
data?
C. Is the training of data collectors clearly
described and adequate?
D. Is the data collection process conducted in
a consistent manner (Borglin & Richards,
2010)?
E. Are the data collection methods ethical?
F. Do the data collected address the research
objectives, questions, or hypotheses?
G. Did any adverse events occur during data
collection, and were these appropriately
managed?
IX. Data analysis
A. Are data analysis procedures appropriate
for the type of data collected (Corty, 2007;
Grove, 2007)?
B. Are data analysis procedures clearly
described? Did the researcher address any
problems with missing data and how this
problem was managed?
C. Do the data analysis techniques address the
study purpose and the research objectives,
questions, or hypotheses (Fawcett &
Garity, 2009)?
D. Are the results presented in an
understandable way by narrative, tables, or
figures, or a combination of methods
(APA, 2010)?
E. Are the statistical analyses logically linked
to the design (Borglin & Richards, 2010)?
F. Is the sample size sufficient to detect
significant differences if they are present?
G. Was a power analysis conducted for
nonsignificant results (Aberson, 2010)?
H. Are the results interpreted appropriately?
X. Interpretation of findings
A. Are findings discussed in relation to each
objective, question, or hypothesis?
B. Are various explanations for significant
and nonsignificant findings examined?
C. Are the findings clinically significant
(Gatchel & Mayer, 2010; LeFort, 1993;
Melnyk & Fineout-Overholt, 2011)?
D. Are the findings linked to the study
framework (Smith & Liehr, 2008)?
E. Are the study findings an accurate
reflection of reality and valid for use in
clinical practice (Houser, 2008)?
F. Do the conclusions fit the results from the
data analyses? Are the conclusions based
on statistically significant and clinically
important results (Gatchel & Mayer,
2010)?
G. Does the study have limitations not
identified by the researcher?
H. Did the researcher generalize the findings
appropriately?
I. Were the identified implications for
practice appropriate based on the study
findings and the findings from previous
462 UNIT THREE  Putting It All Together for Evidence-Based Health Care
research (Brown, 2009; Fawcett & Garity,
2009)?
J. Were quality suggestions made for further
research?
Step III: Evaluating a Study
Evaluation involves determining the validity, credi-
bility, significance, and meaning of the study by exam-
ining the links between the study process, study
findings, and previous studies. The steps of the study
are evaluated in light of previous studies, such as an
evaluation of present hypotheses based on previous
hypotheses, present design based on previous designs,
and present methods of measuring variables based on
previous methods of measurement. The findings of the
present study are also examined in light of the findings
of previous studies. Evaluation builds on conclusions
reached during the first two stages of the critical
appraisal so that the credibility, validity, and meaning
of the study findings can be determined for nursing
knowledge and practice.
Guidelines for Evaluating a Study
You need to reexamine the findings, conclusions, and
implications sections of the study and the researchers’
suggestions for further study. Using the following
questions as a guide, summarize your evaluation of
the study, and document your responses.
I. What rival hypotheses can be suggested for the
findings?
II. Do you believe the study findings are valid?
How much confidence can be placed in the
study findings (Borglin & Richards, 2010;
Fawcett & Garity, 2009)?
III. To what populations can the findings be
generalized (Thompson, 2002)?
IV. What questions emerge from the findings, and
does the researcher identify them?
V. What future research can be envisioned?
VI. Could the limitations of the study have been
corrected?
VII. When the findings are examined in light of
previous studies, what is now known and
not known about the phenomenon under
study?
You need to read previous studies conducted in the
area of the research being examined and summarize
your responses to the following questions:
I. Are the findings of previous studies used to
generate the research problem and purpose?
II. Is the design an advancement over previous
designs (Borglin & Richards, 2010; Shadish
et al., 2002)?
III. Do sampling strategies show an improvement
over previous studies? Does the sample
selection have the potential for adding
diversity to samples previously studied
(Aberson, 2010; Thompson, 2002)?
IV. Does the current research build on previous
measurement strategies so that measurement is
more precise or more reflective of the
variables (DeVon et al., 2007; Waltz et al.,
2010)?
V. How do statistical analyses compare with
analyses used in previous studies (Corty,
2007)?
VI. Do the findings build on the findings of
previous studies?
VII. Is current knowledge in this area identified?
VIII. Does the author indicate the implication of the
findings for practice?
The evaluation of a research report should also
include a final discussion of the quality of the report.
This discussion should include an expert opinion of
the contribution of the study to nursing knowledge and
the need for additional research in selected areas. You
also need to determine if the empirical evidence gener-
ated by this study and previous research is ready for
use in practice (Brown, 2009; Craig & Smyth, 2012;
Fawcett & Garity, 2009; Melnyk & Fineout-Overholt,
2011; Whittemore, 2005).
Critical Appraisal Process for
Qualitative Studies
Critical appraisal of qualitative studies requires a dif-
ferent approach than the steps and processes used
when appraising a quantitative study (Sandelowski,
2008). However, appraisal of quantitative and qualita-
tive studies has a common purpose—determining the
rigor with which the methods were applied. The integ-
rity of the design and methods affects the credibility
and meaningfulness of the findings and their useful-
ness in clinical practice (Pickler & Butz, 2007). Burns
(1989) first described the standards for rigorous quali-
tative research 20 years ago. Since that time, other
criteria have been published (Cesario, Morin, & Santa-
Donato, 2002; Clissett, 2008; Fossey, Harvey, McDer-
mott, & Davidson, 2002; Morse, 1991; Pickler &
Butz, 2007) and have been the source of considerable
debate (Cohen & Crabtree, 2008; Mackey, 2012;
Nelson, 2008; Stige, Malterud, & Midtgarden, 2009;
Whittemore, Chase, & Mandle, 2001). Nurses criti-
cally appraising qualitative studies need three prereq-
uisite characteristics in applying rigorous appraisal
 CHAPTER 18  Critical Appraisal of Nursing Studies 463
standards. Without these prerequisites, nurses may
miss potential valuable contributions qualitative
studies might make to the knowledge base of nursing
for practice. These required prerequisite characteris-
tics are addressed in the following section.
Prerequisites for Critical Appraisal of
Qualitative Studies
The first prerequisite for appraising qualitative studies
is an appreciation for the philosophical foundation of
qualitative research. Qualitative researchers design
their studies to be congruent with one of a wide range
of philosophies, such as phenomenology, symbolic
interactionism, and hermeneutics, each of which
espouses slightly different approaches to gaining new
knowledge (Liamputtong & Ezzy, 2005). Although
unique, the qualitative philosophies are similar in their
view of the uniqueness of the individual and the value
of the individual’s perspective. Without an apprecia-
tion for the philosophical perspective supporting the
study being critically appraised, the appraiser may not
appropriately apply the standards of rigor consistent
with that perspective (Sale, 2008). Chapter 4 contains
more information on the different philosophies that
are foundational to qualitative research.
Guided by an appreciation of qualitative philosoph-
ical perspectives, nurses appraising a qualitative study
can evaluate the approach used to gather, analyze, and
interpret the data. A basic knowledge of different
qualitative approaches is as essential for appraisal of
qualitative studies as knowledge of quantitative
research designs is for appraising quantitative studies.
Spending time in the culture, organization, or setting
that is the focus of the study is an expectation
for ethnography studies but would not be expected
for a phenomenological study. A researcher using a
grounded theory approach is expected to analyze
data to extract social processes and construct connec-
tions among emerging concepts. Phenomenological
researchers are expected to produce a rich, detailed
description of a lived experience. Knowing these dis-
tinctions is a prerequisite to fair and objective critical
appraisal of qualitative studies. What one expects to
find in a qualitative research report may be the most
important influence on one’s appraisal of the study
(Sandelowski & Barroso, 2007).
Appreciating philosophical perspectives and know­
ing qualitative approaches are superficial, however,
without empathy for the participant’s perspective.
Empathy creates an openness to knowing a participant
within a naturalistic holistic framework. This open-
ness allows qualitative researchers and nurses apply-
ing the findings to acknowledge the depth, richness,
and complexity inherent in the lives of the patients we
serve. These prerequisites of philosophical founda-
tion, type of qualitative study, and openness to
study participants direct the implementation of the
following guidelines for critically appraising qualita-
tive studies.
Critical Appraisal Guidelines
for Qualitative Studies
Problem Statement
1. Identify the clinical problem and research problem
that led to the study. Were the clinical problem and
research problem explicitly stated?
2. How did the author establish the significance of the
study? In other words, why should the reader care
about this study? Look for statements about human
suffering, costs of treatment, or the number of
people affected by the clinical problem.
3. Did the researcher identify a personal connection
or motivation for selecting this topic to study? For
example, the researcher may choose to study the
lived experience of men undergoing radiation for
prostate cancer after the researcher’s father under-
went the same treatment. Acknowledging motives
and potential biases is an expectation for qualita-
tive researchers (Munhall, 2012).
Purpose and Research Questions
1. Identify the purpose of the study. Is the purpose a
logical approach to addressing the research problem
of the study (Fawcett & Garity, 2009; Munhall,
2012)? Does the purpose have an intuitive fit with
the problem?
2. List research questions that the study was designed
to answer. If the author does not explicitly provide
the questions, attempt to infer the questions from
the answers. What information did the author
include as findings?
3. Were the research questions related to the problem
and purpose?
4. Were qualitative methods appropriate to answer the
research questions?
Literature Review
1. Did the author cite quantitative and qualitative
studies relevant to the focus of the study? What
other types of literature did the author include?
2. Are the references current? For qualitative studies,
the author may have included studies older than the
5-year limit typically used for quantitative studies.
Findings of older qualitative studies may be rele-
vant to a qualitative study.
464 UNIT THREE  Putting It All Together for Evidence-Based Health Care
3. Identify the disciplines of the authors of studies
cited in the article. Does it appear that the author
searched databases outside of the Cumulative Index
to Nursing and Allied Health Literature for relevant
studies? Research publications in other disciplines
as well as literary works in the humanities may have
relevance for some qualitative studies.
4. Did the author evaluate or indicate the weaknesses
of the available studies?
5. Did the literature review include adequate synthe-
sized information to build a logical argument?
Another way to ask the question: Did the author
provide enough evidence to support the verdict that
the study was needed?
Philosophical Foundation
The methods used by qualitative researchers are deter-
mined by the philosophical foundation of their work.
The researcher may not have stated explicitly the
philosophical stance on which the study is based.
Despite this omission, a knowledgeable reader can
recognize the philosophy through the description of
the problem, formulation of the research questions,
and selection of the methods to address the research
questions. A well-designed qualitative study is con-
gruent at each stage with the underlying philosophical
perspective, and it is clearest if the researcher identi-
fies this perspective in his or her research report
(Fawcett & Garity, 2009; Marshall & Rossman, 2011;
Munhall, 2012).
1. Did the author identify a specific perspective from
which the study was developed? If so, what was it?
2. If a broad philosophy, such as phenomenology, was
identified, did the researcher identify the specific
philosopher, such as Husserl or Heidegger?
3. Did the researcher cite a primary source for the
philosophical foundation?
Qualitative Approach
1. Identify the stated or implied research approach
used for the study.
2. Provide a paraphrased description of the research
approach used. See Chapter 4 for descriptions of
the different qualitative research perspectives or
traditions.
3. Were the methods of the study consistent with the
research tradition?
Sampling and Sample
1. Identify how study participants were selected.
2. At what sites were participants recruited for the
study? Did the sites for recruitment fit the sampling
needs of the study?
3. What were the inclusion and exclusion criteria for
the sample?
4. Were the selected participants able to provide data
relevant to the study purpose and research
questions?
5. How many people participated in the study? Did
any potential study participants refuse to partici-
pate? Did any of the participants start but not
finish the study, which determines the study attri-
tion rate?
Data Collection
1. How were data collected in this study? What ratio-
nale did the author provide for using this data col-
lection method?
2. Identify the period of time during which data col-
lection occurred.
3. Describe the sequence of data collection events for
a participant. For example, were data collected
from one interview or a series of interviews? Were
focus group participants given an opportunity to
provide additional data or review the preliminary
conclusions of the researcher?
4. Did the researcher describe changes that were
made in the methods in response to the context and
early data collection? Were data collection proce-
dures proscriptively applied or allowed to emerge
with some flexibility? Flexibility within parameters
of the method is considered appropriate for qualita-
tive studies (Fossey et al., 2002).
Protection of Human Study Participants
Qualitative studies tend to address areas of human life
that are “sensitive” and require care in the way they
are addressed by researchers (Cowles, 1988). For
example, many of the topics studied by qualitative
researchers are social or moral issues not talked about
in common society. Other qualitative studies explore
emotional topics that may cause the participant dis-
comfort, anxiety, or grief. A qualitative report should
provide clues that the researcher was aware of these
concerns and addressed them appropriately.
1. Identify the benefits and risks of participation
addressed by the authors. Were there benefits or
risks the authors do not identify?
2. How were recruitment and consent techniques
adjusted to accommodate the sensitivity of the
subject matter and psychological distress of poten-
tial participants?
3. How were data collection and management tech-
niques adapted in acknowledgment of participant
sensitivity and vulnerability? For example, did
the authors have a counselor or other resources
 CHAPTER 18  Critical Appraisal of Nursing Studies 465
available for participants who might become upset
or disturbed by the interview?
Data Management and Analysis
1. Describe the data management and analysis
methods used in the study (Marshall & Rossman,
2011; Munhall, 2012).
2. Did the author discuss how the rigor of the process
was ensured? For example, does the author describe
maintaining a paper trail of critical decisions that
were made during the analysis of the data?
3. Analysis of qualitative data is influenced by the
experiences and perspectives of the individuals
doing the analysis. What measures were used to
minimize or allow for the effects of researcher
bias? For example, did two researchers indepen-
dently analyze the data and compare their analy-
ses? Some qualitative researchers believe that
comparability of interpretation across researchers
is more consistent with a positivist philosophy
(Fossey et al., 2002) and would not include this as
a criterion for appraisal.
4. Did the data management and analysis methods fit
the research purposes and data?
Findings
1. Did the findings address the purpose of the study
(Marshall & Rossman, 2011; Munhall, 2012)?
2. Were the data analyzed sufficiently? Findings in a
qualitative study are expected to be more than the
words that participants said. The researcher is
expected to identify themes or abstract concepts
that emerged from the data (Fawcett & Garity,
2009).
3. Were the interpretations of data congruent with
data collected?
4. Did the researcher address variations in the find-
ings by relevant sample characteristics?
Discussion
1. Did the results offer new information about the
target phenomenon?
2. Were the findings linked to findings in other studies
or other relevant literatures (Fawcett & Garity,
2009; Munhall, 2012)?
3. Describe the clinical, policy, theoretical, and other
significance of the findings. Does the author explore
these applications?
Logic and Form of Findings
The study report is the means of communicating the
findings. Producing knowledge can occur only through
communication (Sandelowski & Barroso, 2007); the
logic and form of the findings are critical to the
appraisal of the study.
1. Were readers able to hear the voice of the partici-
pants and gain an understanding of the phenome-
non studied?
2. Were readers able to identify easily the elements of
the research report?
3. Did the overall presentation of the study fit its
purpose, method, and findings (Fawcett & Garity,
2009; Marshall & Rossman, 2011; Munhall, 2012)?
4. Was there a coherent logic to the presentation of
findings?
Evaluation Summary
“The sense of rightness and feeling of comfort readers
experience reading the report of a study constitute the
very judgments they make about the validity or trust-
worthiness of the study itself” (Sandelowski &
Barroso, 2007, p. xix). Critical appraisal is not
complete without making judgments about the validity
of the study. Synthesis of the evaluative criteria for
qualitative studies can be reframed as philosophical
congruence, methodological coherence, intuitive com-
prehension, and intellectual contribution (Cesario
et al., 2002; Clissett, 2008; Fossey et al, 2002; Morse,
1991; Nelson, 2008; Pickler & Butz, 2007; Stige et al.,
2009).
1. Philosophical congruence: Were the development
and implementation of the study congruent with the
philosophical foundation of the study?
2. Methodological coherence: Did the data collection,
analysis, and interpretation processes fit together to
form a coherent approach to address the research
problem?
3. Intuitive comprehension: Do the findings provide a
credible reflection of reality and expand the read-
er’s comprehension of the study topic? If so, how
can the findings be used in nursing practice?
4. Intellectual contribution: What do the findings
contribute to the current body of knowledge?
5. State the conclusion of the critical appraisal of the
study.
KEY POINTS
• Critical appraisal of research involves carefully
examining all aspects of a study to judge its merits,
limitations, meaning, validity, and significance in
light of previous research experience, knowledge
of the topic, and clinical expertise.
• Critical appraisals of research are conducted (1) to
summarize evidence for practice, (2) to provide a
466 UNIT THREE  Putting It All Together for Evidence-Based Health Care
basis for future research, (3) to evaluate presenta-
tions and publications of studies, (4) for abstract
selection for a conference, (5) to select an article
for publication, and (6) to evaluate research pro-
posals for funding and implementation in clinical
agencies.
• Nurses’ levels of expertise in conducting critical
appraisals depend on their educational preparation;
nurses with baccalaureate, master’s, doctorate, and
postdoctorate preparation all have a role in examin-
ing the quality of research.
• The critical appraisal process for quantitative
research includes the following steps: identifying
the steps of the research process in a study; deter-
mining the study strengths and weaknesses; and
evaluating the credibility, validity, and meaning of
a study to nursing knowledge and practice.
• The identification step involves understanding the
terms and concepts in the report and identifying
study elements.
• The second step of determining study strengths and
weaknesses involves comparing what each step of
the research process should be like with how the
steps of the study were conducted. The logical
development and implementation of the study steps
also need to be examined for strengths and
weaknesses.
• Study strengths and weaknesses need to be clearly
identified, supported with a rationale, and docu-
mented with current research sources.
• The evaluation step involves examining the
meaning, validity, and significance of the study
according to set criteria.
• To perform fair critical appraisals of qualitative
studies, nurses need the prerequisites of an appre-
ciation for the philosophical foundations of qualita-
tive research, knowledge of different qualitative
approaches, and empathy for the study participant’s
perspective.
• Each aspect of a qualitative study, such as problem,
purpose, research questions, sample, data collec-
tion and analysis, and findings, needs to be exam-
ined for strengths and weaknesses.
• Philosophical congruence, methodological coher-
ence, intuitive comprehension, and intellectual
contribution are evaluative standards for qualitative
studies.
• Philosophical congruence is the extent to which the
development and implementation of the study
reflected the philosophical underpinnings of the
qualitative approach used for the study.
• Methodological coherence is the extent to which
the data collection, implementation, and analysis
formed a coherent whole to address the research
problem.
• Intuitive comprehension is the enhanced under-
standing gained by readers of the participant per-
spective on the study topic.
• Intellectual contribution is the degree to which the
findings add to the knowledge of the discipline.
REFERENCES
Aberson, C. L. (2010). Applied power analysis for the behavioral
sciences. New York, NY: Routledge Taylor & Francis Group.
Alligood, M. R. (2010). Nursing theory: Utilization & application.
St Louis, MO: Mosby.
American Nurses Credentialing Center (ANCC). (2012).
Magnet program overview. Retrieved from http://www
.nursecredentialing.org/Magnet/ProgramOverview.aspx.
American Psychological Association (APA). (2010). Publication
manual of the American Psychological Association (6th ed.).
Washington, DC: Author.
Bartlett, J. W., & Frost, C. (2008). Reliability, repeatability and
reproducibility: Analysis of measurement errors in continuous
variables. Ultrasound in Obstetrics & Gynecology, 31(4), 466–
475.
Bialocerkowski, A., Klupp, N., & Bragge, P. (2010). Research
methodology series: How to read and critically appraise a reli-
ability article. International Journal of Therapy and Rehabilita-
tion, 17(3), 114–120.
Borglin, G., & Richards, D. A. (2010). Bias in experimental nursing
research: Strategies to improve the quality and explanatory power
of nursing science. International Journal of Nursing Studies,
47(1), 123–128.
Brown, S. J. (2009). Evidence-based nursing: The research-practice
connection. Sudbury, MA: Jones & Bartlett Publishers.
Burns, N. (1989). Standards for qualitative research. Nursing
Science Quarterly, 2(1), 44–52.
Burns, N., & Grove, S. K. (2011). Understanding nursing research
(5th ed.). Philadelphia, PA: Saunders.
Cesario, S., Morin, K., & Santa-Donato, A. (2002). Evaluating the
level of evidence of qualitative research. Journal of Obstetric,
Gynecologic, and Neonatal Nursing, 31(6), 708–714.
Clissett, P. (2008). Evaluating qualitative research. Journal of
Orthopedic Nursing, 12(2), 99–105.
Cohen, D. J., & Crabtree, B. F. (2008). Evaluative criteria for quali-
tative research in health care: Controversies and recommenda-
tions. Annals of Family Medicine, 6(4), 331–339.
Corty, E. W. (2007). Using and interpreting statistics: A practical
text for the health, behavioral, and social sciences. St. Louis, MO:
Mosby.
Cowles, K. V. (1988). Issues in qualitative research on sensitive
topics. Western Journal of Nursing Research, 10(2), 163–
179.
Craig, J. V., & Smyth, R. L. (2012). The evidence-based practice
manual for nurses (3rd ed.). Edinburgh, UK: Churchill
Livingstone.
DeVon, H. A., Block, M. E., Moyle-Wright, P., Ernst, D. M.,
Hayden, S. J., Lazzara, D. J., et al. (2007). A psychometric toolbox
 CHAPTER 18  Critical Appraisal of Nursing Studies 467
for testing validity and reliability. Journal of Nursing Scholarship,
39(2), 155–164.
Doran, D. M. (2011). Nursing-sensitive outcomes: State of the
science. Sudbury, MA: Jones & Bartlett.
Fawcett, J., & Garity, J. (2009). Evaluating research for evidence-
based nursing practice. Philadelphia, PA: F.A. Davis.
Forbes, A. (2009). Clinical intervention research in nursing. Inter-
national Journal of Nursing Studies, 46(4), 557–568.
Fossey, E., Harvey, C., McDermott, F., & Davidson, L. (2002).
Understanding and evaluating qualitative research. Australian and
New Zealand Journal of Psychiatry, 36(6), 717–732.
Gatchel, R. J., & Mayer, T. G. (2010). Testing minimal clinically
important difference: Consensus or conundrum? The Spine
Journal, 35(19), 1739–1743.
Gloeckner, M. B., & Robinson, C. B. (2010). A nursing journal club
thrives through shared governance. Journal for Nurses in Staff
Development, 26(6), 267–270.
Grove, S. K. (2007). Statistics for health care research: A practical
workbook. St. Louis, MO: Saunders.
Higgins, J. P. T., & Green, S. (2008). Cochrane handbook for sys-
tematic reviews of interventions. West Sussex, UK: Wiley-
Blackwell & The Cochrane Collaboration.
Houser, J. (2008). Nursing research: Reading, using, and creating
evidence. Sudbury, MA: Jones & Bartlett.
Kerlinger, F. N., & Lee, H. B. (2000). Foundations of behavioral
research (4th ed.). Fort Worth, TX: Harcourt College.
LeFort, S. M. (1993). The statistical versus clinical significance
debate. Image Journal of Nursing Scholarship, 25(1), 57–62.
Liamputtong, P., & Ezzy, D. (2005). Qualitative research methods
(2nd ed.). Melbourne, VIC, Australia: Oxford University Press.
Mackey, M. C. (2012). Evaluation of qualitative research. In P. L.
Munhall (Ed.). Nursing research: A qualitative perspective (5th
ed., pp. 517–532). Sudbury, MA: Jones & Bartlett.
Marshall, C., & Rossman, G. B. (2011). Designing qualitative
research (5th ed.). Los Angeles, CA: Sage.
Meleis, A. I. (2007). Theoretical nursing: Development and prog-
ress (4th ed.). Philadelphia, PA: Lippincott.
Melnyk, B. M., & Fineout-Overholt, E. (2011). Evidence-based
practice in nursing & healthcare: A guide to best practice (2nd
ed.). Philadelphia, PA: Lippincott Williams & Wilkins.
Mittlbock, M. (2008). Critical appraisal of randomized clinical
trials: Can we have faith in the conclusions? Breast Care, 3(5),
341–346.
Morrison, D. M., Hoppe, M. J., Gillmore, M. R., Kluver, C., Higa,
D., & Wells, E. A. (2009). Replicating an intervention: The
tension between fidelity and adaptation. AIDS Education and Pre-
vention, 21(2), 128–140.
Morse, J. M. (1991). Evaluating qualitative research. Qualitative
Health Research, 1(3), 283–286.
Munhall, P. L. (2012). Nursing research: A qualitative perspective
(5th ed.). Sudbury, MA: Jones & Bartlett Learning.
National Institute of Nursing Research (NINR, 2011). Research and
funding. Retrieved from http://www.ninr.nih.gov/ResearchAnd
Funding/.
Nelson, A. M. (2008). Addressing the threat of evidence-based prac-
tice to qualitative inquiry through increasing attention to quality:
A discussion paper. International Journal of Nursing Studies,
45(2), 316–322.
Pickler, R. H., & Butz, A. (2007). Evaluating qualitative research
studies. Journal of Pediatric Health Care, 21(3), 195–197.
Pyrczak, F. (2008). Evaluating research in academic journals: A
practical guide to realistic evaluation (4th ed.). Los Angeles, CA:
Pyrczak.
Roberts, W. D., & Stone, P. W. (2004). Ask an expert: How to
choose and evaluate a research instrument. Applied Nursing
Research, 16(10), 70–72.
Ryan-Wenger, N. A. (2010). Evaluation of measurement precision,
accuracy, and error in biophysical data for clinical research and
practice. In C. F. Waltz, O. L. Strickland, & E. R. Lenz (Eds.).
Measurement in nursing and health research (4th ed., pp. 371–
383). New York, NY: Springer Publishing Company.
Sale, J. E. M. (2008). How to assess rigor … or not in qualitative
research. Journal of Evaluation in Clinical Practice, 14(5),
912–913.
Sandelowski, M. (2008). Justifying qualitative research. Research
in Nursing & Health, 31(3), 193–195.
Sandelowski, M., & Barroso, J. (2007). Handbook for synthesizing
qualitative research. New York, NY: Springer Publishing
Company.
Santacroce, S. J., Maccarelli, L. M., & Grey, M. (2004). Methods:
Intervention fidelity. Nursing Research, 53(1), 63–66.
Shadish, W. R., Cook, T. D., & Campbell, D. T. (2002). Experimen-
tal and quasi-experimental designs for generalized causal infer-
ence. Chicago, IL: Rand McNally.
Sidani, S., & Braden, C. J. (1998). Evaluation of nursing interven-
tions: A theory-driven approach. Thousand Oaks, CA: Sage.
Smith, M. J., & Liehr, P. R. (2008). Middle range theory for nursing
(2nd ed.). New York, NY: Springer Publishing Company.
Stige, B., Malterud, K., & Midtgarden, T. (2009). Toward an agenda
for evaluation of qualitative research. Qualitative Health Research,
19(10), 1504–1516.
Thompson, S. K. (2002). Sampling (2nd ed.). New York, NY: John
Wiley & Sons.
Tilden, V. (2002). Editorial. Peer review: Evidence-based or sacred
cow? Nursing Research, 51(5), 275.
Waltz, C. F., Strickland, O. L., & Lenz, E. R. (2010). Measurement
in nursing and health research (4th ed.). New York, NY: Springer
Publishing Company.
Whittemore, R. (2005). Methods. Combining evidence in nursing
research: Methods and implications. Nursing Research, 54(1),
56–62.
Whittemore, R., Chase, S. K., & Mandle, C. L. (2001). Validity
in qualitative research. Qualitative Health Research, 11(4),
522–537.

19
CHAPTER
Evidence Synthesis and Strategies for
Evidence-Based Practice
R
esearch evidence has greatly expanded since
the 1990s as numerous quality studies in
nursing, medicine, and other healthcare disci-
plines have been conducted and disseminated. These
studies are commonly communicated via journal pub-
lications, the Internet, books, conferences, and televi-
sion. The expectations of society and the goals of
healthcare systems are the delivery of high-quality,
cost-effective health care to patients, families, and
communities nationally and internationally. To ensure
the delivery of quality health care, the care must be
based on the current, best research evidence available.
Healthcare agencies are emphasizing the delivery of
evidence-based health care, and nurses and physicians
are focused on evidence-based practice (EBP). With
the emphasis on EBP over the last 2 decades, out-
comes have improved for patients, healthcare provid-
ers, and healthcare agencies (Brown, 2009; Craig &
Smyth, 2012; Doran, 2011; Gerrish et al., 2011;
Higgins & Green, 2008; Melnyk & Fineout-Overholt,
2011; Sackett, Straus, Richardson, Rosenberg, &
Haynes, 2000).
Evidence-based practice (EBP) is an important
theme in this textbook and was defined earlier as the
conscientious integration of best research evidence
with clinical expertise and patient values and needs in
the delivery of quality, cost-effective health care (see
Chapter 1) (Craig & Smyth, 2012; Institute of Medi-
cine, 2001; Sackett et al., 2000). Best research evi-
dence is produced by the conduct and synthesis of
numerous high-quality studies in a selected health-
related area. The concept of best research evidence
was described in Chapter 2, and the processes for
synthesizing research evidence (systematic review,
meta-analysis, meta-synthesis, and mixed-methods
systematic review) are defined.
This chapter builds on previous EBP discussions to
provide you with strategies for implementing the best
research evidence in your practice and moving the
468
  http://evolve.elsevier.com/Grove/practice/
profession of nursing toward EBP. This chapter exam-
ines the benefits and barriers related to implementing
evidence-based care in nursing. Guidelines are pro-
vided for synthesizing research to determine the best
research evidence. Two nursing models developed to
facilitate evidence-based practice in healthcare agen-
cies are introduced. Expert researchers, clinicians, and
consumers—through government agencies, profes-
sional organizations, and healthcare agencies—have
developed an extensive number of evidence-based
guidelines. This chapter offers a framework for
reviewing the quality of these evidence-based guide-
lines and for using them in practice. The chapter con-
cludes with a discussion of the nationally designated
EBP centers and Institutional Clinical Translational
Science Awards. These centers and awards are sup-
ported by the U.S. government to expand the research
evidence generated, synthesized, and developed into
evidence-based guidelines for practice.
Benefits and Barriers Related to
Evidence-Based Nursing Practice
EBP is a goal for the profession of nursing and
each practicing nurse. At the present time, some
nursing interventions are evidence-based, or sup-
ported by the best research knowledge available from
systematic reviews, meta-analyses, meta-syntheses,
and mixed-methods systematic reviews. However,
many nursing interventions require additional research
to generate essential knowledge for making changes
in practice. Some nurses readily use research-based
interventions, and others are slower to make changes
in their practice based on research. Some clinical
agencies are supportive of EBP and provide resources
to facilitate this process, but other agencies have
limited support for the EBP process. This section iden-
tifies some of the benefits and barriers related to EBP
 CHAPTER 19  Evidence Synthesis and Strategies for Evidence-Based Practice 469
to assist you in promoting EBP in your agency and
delivering evidence-based care to your patients.
Benefits of Evidence-Based Practice
in Nursing
The greatest benefits of EBP are improved outcomes
for patients, providers, and healthcare agencies. Orga-
nizations and agencies nationally and internationally
have promoted the synthesis of the best research evi-
dence in thousands of healthcare areas by teams of
expert researchers and clinicians. These research syn-
theses, such as systematic reviews and meta-analyses,
have provided the basis for developing strong evidence-
based guidelines for practice. These guidelines identify
the best treatment plan or gold standard for patient care
in a selected area to promote quality health outcomes.
Healthcare providers have easy access to numerous
evidence-based guidelines to assist them in making
the best clinical decisions for their patients. These
evidence-based syntheses and guidelines are commu-
nicated by presentations and publications and can be
easily accessed online through the National Guideline
Clearinghouse (NGC, 2012b) in the United States
(http://www.guidelines.gov/), Cochrane Collaboration
(2012) in England (http://www.cochrane.org/), and
Joanna Briggs Institute (2012) in Australia (http://
www.joannabriggs.edu.au/).
Individual studies, research syntheses, and evidence-
based guidelines assist students, educators, registered
nurses (RNs), and advanced practice nurses (APNs) to
provide the best possible care. Expert APNs, such as
nurse practitioners, clinical nurse specialists, nurse
anesthetists, and nurse midwives, are resources to
other nurses and facilitate access to evidence-based
guidelines to ensure patient care is based on the best
research evidence available (Gerrish et al., 2011).
Nurse researchers and APNs are involved in the devel-
opment of systematic reviews, meta-analyses, meta-
syntheses, and evidence-based guidelines to manage
patient health problems, prevent illnesses, and promote
health.
Healthcare agencies are highly supportive of EBP
because it promotes quality, cost-effective care for
patients and families and meets accreditation require-
ments. The Joint Commission revised their accredita-
tion criteria to emphasize patient care quality achieved
through EBP. Approximately 25% of chief nursing
officers (CNOs) identified the movement toward
evidence-based nursing practice as their number one
priority (Nurse Executive Center, 2005; The Joint
Commission, 2012).
Many CNOs and healthcare agencies are trying
either to obtain or to maintain Magnet status, which
documents the excellence of nursing care in an agency.
Approval for Magnet status is obtained through the
American Nurses Credentialing Center (ANCC).
The national and international healthcare agencies
that currently have Magnet status can be viewed
online at the ANCC (2012) website (http://www
.nursecredentialing.org/FindaMagnetHospital.aspx).
The Magnet Recognition Program® recognizes EBP as
a way to improve the quality of patient care and to
revitalize the nursing environment. Selection criteria
for Magnet status that require healthcare agencies to
promote the conduct of research and the use of research
evidence in practice follow.
“FORCE 6: Quality Care
“Research and Evidence-Based Practice
“22. Describe how current literature, appropriate
to the practice setting, is available,
disseminated, and used to change
administrative and clinical practices.
“23. Discuss the institution’s policies and
procedures that protect the rights of
participants in research protocols. Include
evidence of consistent nursing involvement in
the governing body responsible for protection
of human subjects in research.
“24. Provide evidence that research consultants are
actively involved in shaping nursing research
infrastructure, capacity, and mentorship.
“25. Provide a copy of the nursing budget or other
sources of funding for the past year, the
current year-to-date, and the future
projection, highlighting the allocation and
utilization of resources for nursing research.
“26. Supply documentation of all nursing research
activities that are ongoing, including internal
validation studies, internal and external
research, and participation in surveys
completed within the past twelve (12) month
period.
“27. Provide evidence of education and mentoring
activities that have effectively engaged staff
nurses in research- and/or evidence-based
practice activities.
“28. Describe resources available to nursing staff to
support participating in nursing research and
nursing research utilization activities.” (Nurse
Executive Center, 2005, p. 15)
470 UNIT THREE  Putting It All Together for Evidence-Based Health Care
These selection criteria include critical elements
for EBP, especially financial support for and out-
comes related to research activities. Important
research-related outcomes to be documented by agen-
cies for Magnet status include nursing studies
conducted and professional publications and presen-
tations by nurses. For each study, the following needs
to be documented: title of the study, principal investi-
gator or investigators, role of nurses in the study, and
study status (Horstman & Fanning, 2010). In working
toward EBP, nurses are encouraged to embrace the
benefits of EBP, use the evidence-based guidelines
available, synthesize current research evidence,
and support or participate in the research needed to
determine the effectiveness of selected nursing
interventions.
Barriers of Evidence-Based Practice
in Nursing
Barriers to the EBP movement have been both practi-
cal and conceptual. One of the most serious barriers
is the lack of research evidence available regarding the
effectiveness of many nursing interventions. EBP
requires synthesizing research evidence from random-
ized controlled trials (RCTs) and other types of inter-
vention studies, and these types of studies are still
limited in nursing. Mantzoukas (2009) reviewed the
research evidence in 10 high-impact nursing journals,
including Nursing Research, Research in Nursing &
Health, Western Journal of Nursing Research, Journal
of Nursing Scholarship, and Advances in Nursing
Science, between 2000 and 2006 and found that the
studies were 7% experimental, 6% quasi-experimental,
and 39% nonexperimental. However, RCTs and quasi-
experimental studies conducted to determine the
effectiveness of nursing interventions continue to
increase.
Systematic reviews and meta-analyses conducted
in nursing have been limited compared with other
disciples. In addition, nurse authors of these research
syntheses have sometimes indicated there is inade-
quate research evidence to support using certain
nursing interventions in practice (Craig & Smyth,
2012; Mantzoukas, 2009). Bolton, Donaldson, Rut-
ledge, Bennett, and Brown (2007, p. 123S) conducted
a review of “systematic/integrative reviews and meta-
analyses on nursing interventions and patient out-
comes in acute care settings.” Their literature search
covered 1999-2005 and identified 4000 systematic/
integrative reviews and 500 meta-analyses covering
the following seven topics selected by the authors:
staffing, caregivers, incontinence, elder care, symptom
management, pressure ulcer prevention and treatment,
and developmental care of neonates and infants. The
authors found a limited association between nursing
interventions and processes and patient outcomes
in acute care settings. Their findings included the
following.
“The strongest evidence was for the use of patient
risk-assessment tools and interventions implemented
by nurses to prevent patient harm. We observed sig-
nificant variation in the methods to measure the effect
of independent variables (nursing interventions) on
patient outcomes. Results indicate the need for more
research measuring the effect of specific nursing
interventions that may impact acute care patient out-
comes.” (Bolton et al., 2007, p. 123S)
Extensive evidence has been generated through
nursing research, but additional studies are needed that
focus on determining the effectiveness of nursing
interventions on patient outcomes (Bolton et al., 2007;
Craig & Smyth, 2012; Doran, 2011; Mantzoukas,
2009). Identifying the areas where research evidence
is lacking is an important first step in developing the
evidence needed for practice. Well-designed experi-
mental and quasi-experimental studies are needed to
test selected nursing interventions and to use that
understanding to generate sound evidence for practice
(see Chapter 14). Nurses also need to be more active
in conducting quality syntheses (systematic reviews,
meta-analyses, and meta-syntheses) of research evi-
dence in selected areas (Finfgeld-Connett, 2010;
Higgins & Green, 2008; Rew, 2011; Sandelowski &
Barroso, 2007). The next section of this chapter pro-
vides guidelines to direct different types of research
syntheses.
Another concern is that the research evidence is
generated based on population data and then is applied
in practice to individual patients. Sometimes it is dif-
ficult to transfer research knowledge to individual
patients, who respond in unique ways or have unique
needs (Biswas et al., 2007). More work is needed to
promote the use of evidence-based guidelines with
individual patients. The National Institutes of Health
(NIH) is supporting translational research to improve
the use of research evidence with different patient
populations in various settings. Patients who have
poor outcomes when managed according to an
evidence-based guideline need to be reported, and, if
possible, their circumstances should be published as a
case study. Electronic patient records now make it
 CHAPTER 19  Evidence Synthesis and Strategies for Evidence-Based Practice 471
possible to determine patient outcomes of care deliv-
ered using EBP guidelines.
Best research evidence is generated mainly from
RCTs and other intervention studies with limited
focus on the contributions of descriptive-correlational
studies, qualitative research, mixed-methods studies,
and theories. These types of studies do make contribu-
tions to the research evidence in many areas and need
to be synthesized for use in practice (Mantzoukas,
2009) (see Chapter 10 for mixed-methods studies).
Qualitative researchers have developed several syn-
thesis processes for qualitative studies, and these are
discussed later in this chapter.
Another concern of the EBP movement is that the
development of evidence-based guidelines has led to
a “cookbook” approach to health care. Health profes-
sionals are expected to follow these guidelines in their
practice as developed. However, the definition of EBP
describes it as the conscientious integration of best
research evidence with clinical expertise and patient
values and needs. Nurse clinicians have a major role
in determining how the best research evidence will be
implemented to achieve quality care and outcomes.
For example, a nurse practitioner uses the national
evidence-based guidelines for the treatment of patients
with hypertension (Joint National Committee on Pre-
vention, Detection, Evaluation, and Treatment of High
Blood Pressure [JNC 7]) (Chobanian et al., 2003) but
also makes clinical decisions based on the needs and
values of individual patients. If a patient has a dry,
persistent, irritating cough when taking angiotensin-
converting enzyme inhibitor medications, this type of
medication would not be used to manage the patient’s
high blood pressure if possible. If a patient refuses a
treatment based on cultural or religious reasons, these
reasons would be taken into consideration in develop-
ing the patient’s treatment plan. Evidence-based
guidelines provide the gold standard for managing a
particular health condition, but the healthcare provider
and patient individualize the treatment plan.
Another serious barrier is that some healthcare
agencies and administrators do not provide the
resources necessary for nurses to implement EBP.
Their lack of support might include the following: (1)
inadequate access to research journals and other
sources of synthesized research findings and evidence-
based guidelines, (2) inadequate knowledge on how to
implement evidence-based changes in practice, (3)
heavy workload with limited time to make research-
based changes in practice, (4) limited authority to
change patient care based on research findings, (5)
limited support from nursing administrators or medical
staff to make evidence-based changes in practice, (6)
limited funds to support research projects and research-
based changes in practice, and (7) minimal rewards
for providing evidence-based care to patients and
families (Butler, 2011; Eizenberg, 2010; Gerrish et al.,
2011). The success of EBP is determined by all
involved including healthcare agencies, administra-
tors, nurses, physicians, and other healthcare profes-
sionals. We all need to take an active role in ensuring
that the health care provided to patients and families
is based on the best research available.
Guidelines for Synthesizing
Research Evidence
Many nurses lack the expertise and confidence to syn-
thesize research evidence in a selected area of nursing.
They need additional knowledge and skills in criti-
cally appraising studies and synthesizing research evi-
dence. Synthesizing research evidence can focus on a
specific area or intervention for practice or on complex
clinical problems. Master’s and doctoral students
often focus on clearly defined interventions when con-
ducting research syntheses. Synthesis of research in
complex clinical areas is best done with a team of
expert researchers and clinicians. However, novice
researchers need to be included in these teams to
increase their understanding of the synthesis processes
for determining the best research evidence in an area.
In this section, guidelines are provided for conduct-
ing systematic reviews, meta-analyses, meta-syntheses,
and mixed-methods systematic reviews to guide you
in synthesizing research evidence for nursing practice.
Numerous research syntheses have been conducted in
nursing and medicine, so be sure to search for an exist-
ing synthesis or review of research in an area before
undertaking such a project. More recent data suggest
that at least 2500 new systematic reviews are reported
in English and indexed in MEDLINE each year (Libe-
rati et al., 2009). Table 19-1 identifies some common
databases and EBP organizational websites for nurses
to search for nursing syntheses of research. The
Cochrane Collaboration library of systematic reviews
is an excellent resource with more than 11,000
entries relevant to nursing and health care (http://
www.cochrane.org/cochrane-reviews). In 2009, the
Cochrane Nursing Care Field was developed to
support the conduct, dissemination, and use of system-
atic reviews in nursing. The Joanna Briggs Institute
also provides resources for locating and conducting
research syntheses in nursing (see Table 19-1). If you
can find no synthesis of research for a selected nursing
intervention or the review you find is outdated, you
472 UNIT THREE  Putting It All Together for Evidence-Based Health Care

TABLE19-1  Evidence-Based Practice Resources

Resource
Electronic Databases
CINAHL (Cumulative Index to
Nursing and Allied Health
Literature)
MEDLINE (PubMed—National
Library of Medicine)
MEDLINE with MeSH
PsychINFO
CANCERLIT
National Library Sites
Cochrane Library
National Library of Health (NLH)
Evidence-Based Practice Organizations
Cochrane Nursing Care Network
National Guideline Clearinghouse
(NGC)
National Institute for Health and
Clinical Excellence (NICE)
Joanna Briggs Institute
Description
CINAHL is an authoritative resource covering the English-language journal literature
for nursing and allied health. Database was developed in the U.S. and includes
sources published from 1982 forward
Database was developed by the National Library of Medicine in the U.S. and provides
access to >11 million MEDLINE citations back to the mid-1960s and additional life
science journals
Database provides authoritative medical information on medicine, nursing, dentistry,
veterinary medicine, the healthcare system, preclinical services, and more
Database was developed by the American Psychological Association and includes
professional and academic literature for psychology and related disciplines from
1887 forward
Database of information on cancer was developed by the U.S. National Cancer Institute
Cochrane Library provides high-quality evidence to inform people providing and
receiving health care and people responsible for research, teaching, funding, and
administration of health care at all levels. Included in the Cochrane Library is the
Cochrane Collaboration (2012), which has many systematic reviews of research.
Cochrane Reviews are available at http://www.cochrane.org/reviews/
NLH is located in the United Kingdom. You can search for evidence-based sources at
http://www.evidence.nhs.uk/
Cochrane Collaboration includes 11 different fields, one of which is the Cochrane
Nursing Care Field (CNCF), which supports the conduct, dissemination, and use of
systematic reviews in nursing; see http://cncf.cochrane.org/
Agency for Healthcare Research and Quality (AHRQ) developed NGC to house the
thousands of evidence-based guidelines that have been developed for use in clinical
practice. The guidelines can be accessed online at http://www.guidelines.gov
NICE was organized in the United Kingdom to provide access to the evidence-based
guidelines that have been developed. These guidelines can be accessed at http://
nice.org.uk
This international evidence-based organization, originating in Australia, has a search
website that includes evidence summaries, systematic reviews, systematic review
protocols, evidence-based recommendations for practice, best practice information
sheets, consumer information sheets, and technical reports; see Search the Joanna
Briggs Institute (2012) at http://www.joannabriggs.edu.au/Search.aspx

might use the following guidelines to conduct a sys-
tematic review of relevant research.
Guidelines for Implementing and Evaluating
Systematic Reviews
A systematic review is a structured, comprehensive
synthesis of the research literature to determine the
best research evidence available to address a health-
care question. A systematic review involves identify-
ing, locating, appraising, and synthesizing quality
research evidence for expert clinicians to use to
promote an EBP (Bettany-Saltikov, 2010a; Craig &
Smyth, 2012; Higgins & Green, 2008; Liberati et al.,
2009; Rew, 2011). Systematic reviews are often con-
ducted by two or more researchers or clinicians (or
researchers and clinicians) in a selected area of interest
to determine the best research knowledge in that area.
Systematic reviews need to be conducted with rigor-
ous research methodology to promote the accuracy of
the findings and minimize the reviewers’ bias.
Table 19-2 is a checklist for critically appraising
the steps of a systematic review, and these steps are
discussed in the following section. These steps are
based on the Preferred Reporting Items for Systematic
 CHAPTER 19  Evidence Synthesis and Strategies for Evidence-Based Practice 473

TABLE 19-2  Checklist for Critically Appraising Published Systematic Reviews

Step Complete Comments: Quality
Systematic Review Steps (Yes or No) and Rationale
  1.  Was the clinical question clearly expressed and significant? Was the PICOS
(participants, intervention, comparative interventions, outcomes, and study
design) format used to develop the question and focus the review?
  2.  Were the purpose and objectives or aims of the review clearly expressed
and used to direct the review?
  3.  Were the search criteria clearly identified? Was the PICOS format used to
identify the search criteria and were the years covered, language, and
publication status of sources identified in the search criteria?
  4.  Was a comprehensive, systematic search of the literature conducted using
explicit criteria identified in Step 3? Were the search strategies clearly
reported with examples? Did the search include published studies, grey
literature, and unpublished studies?
  5.  Was the process for the selection of studies for the review clearly identified
and consistently implemented? Was the selection process expressed in a
flow diagram such as Figure 19-1?
  6.  Were key elements (population, sampling process, design, intervention,
outcomes, and results) of each study clearly identified and presented in a
table?
  7.  Was a quality critical appraisal of the studies conducted? Were the results
related to participants, types of intervention, outcomes, outcome
measurement methods, and risks of bias clearly discussed related to each
study (i.e., in table and narrative format)?
  8.  Was a meta-analysis conducted as part of the systematic review? Was a
rationale provided for conducting the meta-analysis? Were the details of
the meta-analysis process and results clearly described?
  9.  Were the results of the review clearly described (i.e., in narrative and
table)? Were details of the study interventions compared and contrasted in
a table? Were the outcome variables clearly identified and the quality of
the measurement methods addressed?
10.  Did the report conclude with a clear discussion section?
a.  Were the review findings summarized to identify the current best
research evidence?
b.  Were the limitations of the review and how they might have affected the
findings addressed?
c.  Were the recommendations for further research, practice, and policy
development addressed?
11.  Did the authors of the review develop a clear, concise, quality report for
publication? Was the report inclusive of the items identified in the
PRISMA Statement (Liberati et al., 2009)?

Reviews and Meta-Analyses (PRISMA) Statement
and other relevant sources to guide nurses in conduct-
ing systematic reviews (Bettany-Saltikov, 2010a,
2010b; Higgins & Green, 2008; Rew, 2011). The
PRISMA Statement was developed in 2009 by an
international group of expert researchers and clini-
cians to improve the quality of reporting for system-
atic reviews and meta-analyses. The PRISMA
Statement includes 27 items, which can be found at
http://prisma-statement.org/ and are detailed in the
article by Liberati et al. (2009). If the review process
is clearly detailed in the report, others can replicate
the process and verify the findings (Rew, 2011). A
systematic review conducted by Goulding, Furze, and
Birks (2010) is presented as an example with the dis-
cussion of the review steps outlined in Table 19-2.
Goulding et al. conducted a systematic review of only
RCTs to determine the best interventions to use in
changing maladaptive illness beliefs of people with
coronary heart disease (CHD).
474 UNIT THREE  Putting It All Together for Evidence-Based Health Care
Step 1: Formulate a Relevant Clinical Question
to Direct the Review
A systemic review or meta-analysis is best directed by
a relevant clinical question that focuses the review
process and promotes the development of a quality
synthesis of research evidence. Formulating the ques-
tion involves identifying a relevant topic, developing
a question of interest that is worth investigating,
deciding if the question will generate significant infor-
mation for practice, and determining if the question
will clearly direct the review process and synthesis of
findings. A well-stated question will define the nature
and scope of the literature search, identify keywords
for the search, determine the best search strategy,
provide guidance in selecting articles for the review,
and guide the synthesis of results (Bettany-Saltikov,
2010a, 2010b; Higgins & Green, 2008; Liberati et al.,
2009).
The question developed might focus on an inter-
vention or therapy, health promotion action, illness
prevention strategy, diagnostic process, prognosis,
causation, or experiences (Bettany-Saltikov, 2010a).
One of the most common formats used to develop a
relevant clinical question to guide a systematic review
is the PICO or PICOS format described in the
Cochrane Handbook for Systematic Reviews of Inter-
ventions (Higgins & Green, 2008). PICOS format
includes the following elements:
P—Population or participants of interest (see Chapter
15 on sampling)
I—Intervention needed for practice (see Chapter 14
on intervention research)
C—Comparisons of the intervention with control,
placebo, standard care, variations of the same inter-
vention, or different therapies (see Chapter 14)
O—Outcomes needed for practice (see Chapter 13 on
outcomes research and Chapter 17 on measurement
methods)
S—Study design (see Chapter 11 on types of study
designs)
Goulding et al. (2010) noted that interventions to
change maladaptive illness beliefs were beneficial to
people with CHD because positive illness representa-
tions may lead to improved lifestyle behaviors of exer-
cise, smoking cessation, and balanced diet. What was
not known was “[w]hich types of intervention to
change illness cognitions (e.g. counseling, education,
or cognitive behavioural) are most effective” (Gould-
ing et al., 2010, p. 948) for people with CHD. The
population was people with CHD, and the intervention
was focused on changing maladaptive illness beliefs
of these individuals. The different types of this
intervention, including counseling, education, and
cognitive behavioral therapy, were compared. The
intervention group was compared with groups receiv-
ing standard care, no treatment, or a variation of the
treatment. The primary outcome measured was the
change in beliefs about CHD at follow-up. The study
design included synthesis of only RCTs using guide-
lines from the Cochrane Collaboration handbook
(Higgins & Green, 2008) (see Chapters 11 and 14
about conducting RCTs). The study design (RCTs)
clearly focused the literature review but might have
eliminated some important studies that could have
expanded the knowledge related to the intervention of
changing illness beliefs.
Step 2: State the Purpose and Objectives
or Aims of the Review
Most systematic reviews of research include a purpose
and specific aims or objects to guide the synthesis
process (Bettany-Saltikov, 2010a; Rew, 2011). The
purpose identifies the major goal or focus of the
review. Goulding et al. (2010, p. 948) identified their
purpose as follows: “This systematic review was
therefore necessary to collate and present evidence of
the effectiveness of maladaptive belief change inter-
ventions for people with CHD.” The specific aims
direct the remaining steps of the research synthesis
and become the focus of the discussion of the findings.
Goulding et al. (2010) identified the following.
“Aims
“The aims of the systematic review were to establish
whether interventions can significantly change mal-
adaptive illness cognitions in people with CHD and to
demonstrate which types of intervention are most
effective. We also aimed to assess whether change in
cognition was accompanied by changes in behav-
ioural, functional, and psychological outcomes.”
(Goulding et al., 2010, p. 948)
Step 3: Identify the Literature Search Criteria
and Strategies
Researchers conducting a systematic review or meta-
analysis need to identify the inclusion and exclusion
criteria to be used to direct their literature search. The
PICOS format might be used to develop the search
criteria with more detail being developed for each of
the elements. These search criteria might focus on the
following: (1) type of research methods, such as quan-
titative, qualitative, or outcomes research; (2) the
 CHAPTER 19  Evidence Synthesis and Strategies for Evidence-Based Practice 475
population or type of study participants; (3) study
designs, such as description, correlational, quasi-
experimental, experimental, qualitative, or mixed
methods; (4) sampling processes, such as probability
or nonprobability sampling methods; (5) intervention
and comparison interventions; and (6) specific out-
comes to be measured. The PICOS format is effective
in identifying the key terms to be included in the
search process. The search criteria also need to indi-
cate the years for the review, language, and publica-
tion status. The focus of the review might be narrowed
by limiting the years reviewed, the language to
English, and only studies in print (Bettany-Saltikov,
2010b; Higgins & Green, 2008; Rew, 2011).
Often searches have been limited to published
sources in common databases, which excludes the
grey literature from the research synthesis. Grey lit-
erature refers to studies that have limited distribu-
tions, such as theses and dissertations, unpublished
research reports, articles in obscure journals, articles
in some online journals, conference papers and
abstracts, conference proceedings, research reports to
funding agencies, and technical reports (Benzies,
Premji, Hayden, & Serrett, 2006; Conn, Valentine,
Cooper, & Rantz, 2003). Most grey literature is dif-
ficult to access through database searches and is often
not peer-reviewed with limited referencing informa-
tion. These are some of the main reasons for not
including grey literature in searches for systematic
reviews and meta-analyses. However, excluding grey
literature from these searches might result in mislead-
ing, biased results. Studies with significant findings
are more likely to be published than studies with non-
significant findings and are usually published in more
high-impact, widely distributed journals that are
indexed in computerized databases (Conn et al.,
2003). Studies with significant findings are more
likely to have duplicate publications that need to be
excluded when selecting studies to include in a
research synthesis. Benzies et al. made the following
recommendations related to including grey literature
in a systematic review or meta-analysis.
• “Intervention and outcome are complex with
multiple components.
• Lack of consensus is present concerning
measurement of outcome.
• Context is important to implementing the
intervention.
• Availability of research-based evidence is low
volume and quality.” (Benzies et al., 2006, p. 60)
Authors of systematic reviews also need to identify
the search strategies that they will use. Many sources
are identified through searches of electronic databases
using the criteria previously discussed. However, pub-
lication bias might best be reduced with more rigorous
searches of the following areas for grey literature and
other unpublished studies:
1. Review the references of identified studies for
additional studies. These are ancestry searches to
use citations in relevant studies to identify addi-
tional studies.
2. Hand search certain journals for selected years,
especially for older studies that were not identified
in the electronic search.
3. Identify expert researchers in an area and search
their names in the databases.
4. Contact the expert researchers regarding studies
they have conducted that have not been published
yet.
5. Search thesis and dissertation databases for rele-
vant studies.
6. Review abstracts and conference reports of rele-
vant professional organizations.
7. Search the websites of funding agencies for rele-
vant research reports (Bettany-Saltikov, 2010b;
Conn et al., 2003; Liberati et al., 2009).
Often it is best to construct a table that includes the
search criteria so that they can be applied consistently
throughout the search process (Liberati et al., 2009)
(see Chapter 6). Goulding et al. (2010) used the
PICOS format to determine inclusion criteria for the
studies in the search. The participants had to be adults
with at least one of the following: angina, CHD, myo-
cardial infarction, or eligible for or recently received
revascularization by percutaneous coronary interven-
tion or coronary artery bypass graft surgery. The
studies needed to focus on an intervention to change
knowledge, attitudes, perceptions, and misconceptions
about CHD. The interventions to change maladaptive
beliefs were compared with different interventions,
usual care, or no intervention. The primary outcome
was change in beliefs about CDH, and the secondary
outcomes were focused on quality of life, behavior
change, anxiety level, depression, psychological well-
being, and modifiable risk factors. The study design
was limited to only RCTs that included a comparison
of the intervention group with a control group or
another intervention.
Goulding et al. (2010) designed their literature
search strategies and their protocol for conducting
their systematic review using sources such as the
Cochrane Collaboration handbook (Higgins & Green,
2008) and the Quality of Reporting of Meta-analyses
476 UNIT THREE  Putting It All Together for Evidence-Based Health Care
(QUOROM) Statement (Moher et al., 1999). No date
restriction was applied to the search for studies, but
because of the lack of funds, only studies reported in
English were identified. The databases searched are
discussed in Step 4. The researchers did not include
grey literature in their review and recognized this as a
limitation in their discussion section.
Step 4: Conduct a Comprehensive Search
of the Research Literature
The next step for conducting a systematic review or
meta-analysis requires an extensive search of the lit-
erature focused by the criteria and strategies identified
in Step 3. The different databases searched, date of the
search, and search results need to be recorded for each
database (see Chapter 6 for details on conducting and
storing searches of databases). Table 19-1 identifies
common databases that are searched by nurses in con-
ducting syntheses of research and in searching for
evidence-based guidelines. Usually the key search
terms are identified in the report. Sometimes authors
of systematic reviews provide a table that identifies
the search terms and criteria. The PRISMA Statement
recommends presenting the full electronic search
strategy used for at least one major database such as
CINAHL or MEDLINE (Liberati et al., 2009). The
search strategies used to identify grey literature and
other unpublished studies need to be identified.
Goulding et al. (2010) searched the following elec-
tronic databases: MEDLINE, EMBASE, CINAHL,
BNI, PsychINFO, The Cochrane Library (including
the Cochrane Database of Systematic Reviews,
CENTRAL, and DARE), and Web of Knowledge.
They provided an extensive table that detailed their
electronic search of the five databases identified. The
search criteria included the participants with different
types of CHD; interventions of education and cogni-
tive and behavioral therapies; comparisons with
control, standard care, and placebo groups; outcomes
of health knowledge, attitudes, and illness percep-
tions; and study design of RCT. The authors noted that
the electronic search identified 3526 citations and that
they obtained one source from an expert researcher
and seven by reviewing the references of other studies.
Step 5: Selection of Studies for Review
The selection of studies for inclusion in the systematic
review or meta-analysis is a complex process that
initially involves review and removal of duplicate
sources. The abstracts of the remaining studies are
reviewed by two or more authors and sometimes an
external reviewer to ensure they meet the criteria
identified in Step 3. The abstracts might be excluded
based on the study participants, interventions, out-
comes, or design not meeting the search criteria.
Sometimes the abstracts are not in English, are incom-
plete, or are of studies not attainable. If contacting the
authors of the abstracts cannot produce essential infor-
mation, often the abstracts are excluded from the
review (Bettany-Saltikov, 2010b; Higgins & Green,
2008; Liberati et al., 2009).
After the abstracts meeting the designated criteria
are identified, the next step is to retrieve the full-text
citation for each study. It is best to enter these studies
into a table and document how each study meets the
eligibility criteria. If studies do not meet criteria, they
need to be removed with a rationale provided. Two or
more authors of the review need to examine the studies
to ensure that the eligibility or inclusion criteria are
consistently implemented. Often the study selection
process includes all members of the review team. This
selection process is best demonstrated by a flow
diagram that was developed by the PRISMA Group
(Liberati et al., 2009). Figure 19-1 shows this diagram,
which has four phases: (1) identification of the sources,
(2) screening of the sources based on set criteria, (3)
determining if the sources meet eligibility require-
ments, and (4) identifying the studies included in the
review.
Goulding et al. (2010) provided the following
description of their section of sources and a flow
diagram (see Figure 19-2) that documented the final
results of the 13 RCTs included in their systematic
review.
“Search Outcome
“The electronic search produced 3526 citations,
which were reduced to 115 on citation review. A
check of 10% of these citations was undertaken by
an independent researcher from another university,
with 100% concordance on abstracts to be retrieved.
A review of abstracts identified 74 papers to retrieve
in full. A further seven papers were identified from
reference checks, and an additional relevant study
was uncovered via contact with an expert in the field.
After a consensus meeting between all authors of the
review, 13 studies were included. Each of these was
a published journal article. The study selection flow-
chart shown in [Figure 19-2] documents this process.”
(Goulding et al., 2010, p. 950)
 CHAPTER 19  Evidence Synthesis and Strategies for Evidence-Based Practice 477

Identification
Records identified through Additional records identified
database searching through other sources
(n = ) (n = )

Records after duplicates removed

(n = )
Screening

Records screened Records excluded

(n = ) (n = )

Full-text articles assessed Full-text articles excluded,

Eligibility

for eligibility with reasons

(n = ) (n = )

Studies included in
qualitative synthesis
(n = )
Included

Studies included in
quantitative synthesis
(meta-analysis)
(n = )

Figure 19-1  PRISMA 2009 Flow Diagram. Identification, screening, eligibility, and inclusion of research sources in systematic reviews and meta-
analyses. (From Moher, D., Liberati, A., Tetzlaff, J., Altman, D. G., & PRISMA Group. [2009]. Preferred Reporting Items for Systematic Reviews and
Meta-Analyses: The PRISMA Statement. Retrieved from http://www.prisma-statement.org.)

Step 6: Critical Appraisal of the Studies Included
in Review
An initial critical appraisal of the methodological
quality of the studies occurs during the selection of
the studies to be included in the systematic review.
Once the studies are selected, a more thorough critical
appraisal takes place. This critical appraisal is best
done by constructing a table describing the character-
istics of the included studies, such as the purpose of
the studies, population, sampling method, sample size,
sample acceptance and attrition rates, design, inter-
vention (independent variable), dependent variables,
measurement methods for each dependent variable,
and major results (Bettany-Saltikov, 2010b; Higgins
& Green, 2008; Liberati et al., 2009).
It is best if two or more experts independently
review the studies and make judgments about their
quality. The authors of the review usually contact the
study investigators if needed to obtain important infor-
mation about the study design or results not included
in the publication. Chapter 18 provides guidelines
for critically appraising quantitative and qualitative
studies. The critical appraisal of the studies is often
difficult because of the differences in types of partici-
pants, designs, sampling methods, intervention proto-
cols, outcome variables and measurement methods,
478 UNIT THREE  Putting It All Together for Evidence-Based Health Care

Potentially relevant citations identified by a

sensitive electronic search: n = 3526

Citations excluded following

initial screen: n = 3406
Duplicates excluded: n = 5

Abstracts reviewed: n = 115

Abstracts not meeting

inclusion criteria: n = 41
Reasons for exclusion:
• Not English language: 7
• Abstract only: 2
• Design: 7
• Participants: 2
• Intervention: 6
• Outcome measure: 11
• No results: 4
Figure 19-2  Study selection flow chart. (From Goulding, L.,
• Paper unobtainable: 2
Furze, G., & Birks, Y. [2010]. Randomized controlled trials of
interventions to change maladaptive illness beliefs in people
with coronary heart disease: Systematic review. Journal of Studies retrieved in full from
Advanced Nursing, 66[5], 950.) electronic search: n = 74

Further studies identified

through reference search and
author contact: n = 8

Excluded after evaluation of

full text: n = 69
Reasons for exclusion:
• Design: 14
• Participants: 4
• Intervention: 9
• Outcome measure: 38
• Repeat data: 3
• No results: 1

Studies included in the

systematic review: n = 13

and presentation of results. The studies are often rank-
ordered based on their quality and contribution to the
development of the review (Bettany-Saltikov 2010b;
Liberati et al., 2009).
Goulding et al. (2010) developed a detailed table
of essential content from the 13 studies included in the
systematic review and labeled the headings of the
columns in the table as (1) authors, year, and country
of the study; (2) participants and setting; (3) design,
sample size per group, and follow-up; (4) intervention;
(5) control or comparison group; (6) study outcomes;
and (7) results with statistical significance set at
alpha = 0.05. Two reviewers independently assessed
the quality of each study. They provided a detailed
description of the appraisal process and their findings
for each study. You may want to access this systematic
 CHAPTER 19  Evidence Synthesis and Strategies for Evidence-Based Practice 479
review to view their table of studies and examine their
critical appraisal process.
Step 7: Conduct a Meta-Analysis If Appropriate
Some systematic reviews include published meta-
analyses as sources in the review. Because a meta-
analysis involves the use of statistics to summarize
results of different studies, it usually provides strong,
objective information about the effectiveness of an
intervention or solid knowledge about a clinical
problem. Other authors conduct a meta-analysis in the
synthesis of sources for their systematic review (Libe-
rati et al., 2009). The authors of the review need to
provide a rationale for conducting the meta-analysis
and detail the process they used to conduct this analy-
sis. For example, a meta-analysis might be conducted
on a small group of similar studies to determine the
effect of an intervention. The next section provides
more details on conducting a meta-analysis.
The systematic review conducted by Goulding
et al. (2010) did not include a meta-analysis as a
source, and a meta-analysis was not conducted during
the review process. A meta-analysis was probably not
appropriate because of the limited number of studies
that had been conducted to examine the effectiveness
of the intervention to change maladaptive illness
beliefs in people with CHD.
Step 8: Results of the Review
The results of the authors’ reviews need to include
a description of the study participants, types of inter-
ventions, and outcomes. The results of the different
types of interventions might be best summarized in
a table that includes the following: (1) study source;
(2) structure of the intervention (stand-alone or multi-
faceted); (3) specific type of intervention such as
physiological treatment, education, counseling, or
behavioral therapy; (4) delivery method such as dem-
onstration and return demonstration, verbal, video,
or self-administered; and (5) statistical difference
between the intervention and the control, standard
care, placebo, or alternative intervention groups
(Liberati et al., 2009).
The specific outcomes, including primary and sec-
ondary outcomes, of the studies might also be best
summarized in a table. This table might include (1)
the study source; (2) outcome variable, with an indica-
tion as to whether it was a primary or secondary
outcome in the study; (3) measurement method used
for each study outcome variable; and (4) the quality
of the measurement methods, such as the reliability
and validity of a scale or the precision and accuracy
of a physiological measure (see Chapter 16).
Goulding et al. (2010, p. 955) described their par-
ticipants for their 13 studies as “male and female
adults of all ages with a diagnosis of CHD (including
people diagnosed with MI [myocardial infarction] or
angina or those receiving revascularization). There
was no clear link between patient group and effective-
ness of interventions to change beliefs.” These authors
also provided two detailed tables: one addressing the
types of interventions and the other identifying the
measurement methods of the study outcomes and
quality of these methods. Numerous outcome mea-
sures were used in the 13 studies, and some had poor
reliability and validity, which limit the results of this
systematic review. Goulding et al. also summarized
key findings related to the effectiveness of the inter-
ventions and reached the following conclusion.
“Overall, the majority of interventions designed to
elicit positive and correct illness cognitions regarding
CHD were effective. Such interventions can be effec-
tive either as part of a multifaceted intervention or as
a stand-alone intervention. However, because of the
numerous differences in the structure of each inter-
vention, method of belief change, and method of
delivery, it is difficult to ascertain whether there is a
relationship between type of intervention and effect
on belief change.” (Goulding et al., 2010, p. 956)
Step 9: Discussion Section of the Review
In a systematic review or meta-analysis, the discussion
of the findings needs to include an overall evaluation
of the types of interventions implemented and the
outcomes measured. The methodological issues or
limitations of the review also need to be addressed.
The discussion section needs to include a theoretical
link back to the studies’ frameworks to indicate the
theoretical implications of the findings. Lastly, the
discussion section needs to provide recommendations
for further research, practice, and policy development
(Bettany-Saltikov, 2010b; Higgins & Green, 2008;
Liberati et al., 2009). Goulding et al. (2010) provided
the following discussion of their findings; review limi-
tations; and recommendations for research, practice,
and policy development.
“We found that interventions to change beliefs can be
successful, with cognitive behavioral interventions
being the most consistently effective. The evidence
on whether interventions to change maladaptive
beliefs can improve psychological, functional, and
behavioural outcomes was unclear. It is therefore not
480 UNIT THREE  Putting It All Together for Evidence-Based Health Care
possible to determine which types of intervention are
most effective in creating improvements in these
important outcomes.
Methodological Issues
The major weakness of the review methodology was
the lack of a search for non-English language litera-
ture, unpublished trials, and grey literature because
of time and resource constraints.… It is therefore
possible that the review has a publication bias.
However, as one of the authors (GF) is internationally
collaborative in the review area, it is unlikely that a
large number of studies of good quality was missed.…
The purpose of including only RCTs was to synthe-
size the results of the best quality research available,
yet some of the included studies did not meet the
anticipated quality standard. This said, the overall
review methodology was good.…
Leventhal’s CSM [Common Sense Model] can be
used to explain the relationships between illness
beliefs, coping, and medical, psychological, and social
outcomes. Four of the reports included in the review
explicitly mentioned the theoretical framework of the
CSM as informing the design of the intervention or
interpretation of results.…
Implications for Future Research
The review demonstrates the need for methodologi-
cally sound and adequately powered trials of interven-
tions to change maladaptive illness cognitions.… The
follow-up time of such interventions should be long
enough to determine whether any positive effects
remain stable over time.… Cognitive behavioral
interventions appear promising, and could perhaps
be used in conjunction with education and counseling.
It is important that researchers choose a valid and
reliable measure to assess change in the cognition(s)
of interest.” (Goulding et  al., pp. 957-959)
Goulding et al. stressed the importance of interven-
tions to change maladaptive illness beliefs of people
with CHD. However, the impact of changing those
beliefs on behavior, functional, and psychological out-
comes is unclear. Additional research is needed on the
use of cognitive behavioral interventions to change
maladaptive illness beliefs and how this affects people
with CHD health outcomes.
Step 10: Development of the Final Report
for Publication
The final step is the development of the systematic
review report for publication and presentation. The
report should include a title that identifies it as either
a systematic review or a meta-analysis for ease of
location in database searches. An abstract also needs
to be included that identifies background, purpose,
data sources, review methods, results, and conclu-
sions. The body of the report needs to include the
content discussed in the previous nine steps: (1) clini-
cal question to be addressed by the review; (2) purpose
and aims or objectives to be accomplished by the
review; (3) search criteria and strategies used to iden-
tify essential studies; (4) comprehensive search of the
literature and results; (5) selection of studies to be
included in the review; (6) critical appraisal process
and findings; (7) description of the meta-analysis con-
ducted and results if appropriate; (8) presentation of
the review results in table and narrative format; and
(9) discussion of the systematic review findings, limi-
tations, and implications for further research, practice,
and policy development.
The PRISMA Group developed a 27-item checklist
to use in developing the final reports for systematic
reviews and meta-analyses. This PRISMA Group
checklist is presented in Table 19-3 and discussed
on PRISMA Group website (http://www.prisma-
statement.org) and in the articles by Liberati et al.
(2009) and Moher, Liberati, Tetzlaff, Altman, and
PRISMA Group (2009). This checklist is an excellent
guide to use in developing the final report of a
systematic review or meta-analysis for publication.
This checklist was developed to improve the quality,
completeness, and consistency of the syntheses devel-
oped and published in nursing, medicine, and other
healthcare professions.
Critical Appraisal of a Published Systematic Review
Your critical appraisal of a systematic review might be
guided by the checklist provided in Table 19-2. This
table includes the steps of a systematic review, and you
need to assess if each step was completed or not in the
review. You also need to provide comments and ratio-
nale for the appraised strengths and limitations of the
review. Using this checklist, you could develop a
formal critical appraisal paper for a systematic review.
In developing and critically appraising systematic
reviews, you might also use the articles by Bettany-
Saltikov (2010a, 2010b), the Cochrane Collaboration
handbook (Higgins & Green, 2008), the EBP manual
for nurses by Craig and Smyth (2012), and other
sources identified by your faculty or experts in this area.
The critical appraisal of a systematic review or
meta-analysis also needs to include an assessment of
how current the literature synthesis is. This leads to
the following question: How quickly do systematic
reviews become out of date? Shojania et al. (2007,
 CHAPTER 19  Evidence Synthesis and Strategies for Evidence-Based Practice 481

TABLE 19-3  Checklist of Items to Include When Reporting a Systematic Review or Meta-Analysis
Reported on
Section/Topic No. Checklist Item Page No.
Title
Title 1 Identify the report as a systematic review, meta-analysis, or both
Abstract
Structured 2 Provide a structured summary including, as applicable, background;
summary objectives; data sources; study eligibility criteria, participants, and
interventions; study appraisal and synthesis methods; results; limitations;
conclusions and implications of key findings; systematic review
registration number
Introduction
Rationale 3 Describe the rationale for the review in the context of what is already
known
Objectives 4 Provide an explicit statement of questions being addressed with reference to
PICOS (participants, interventions, comparisons, outcomes, and study
design)
Methods
Protocol and 5 Indicate if a review protocol exists; if and where it can be accessed (e.g.,
registration Web address); and, if available, provide registration information
including registration number
Eligibility criteria 6 Specify study characteristics (e.g., PICOS, length of follow-up) and report
characteristics (e.g., years considered, language, publication status) used
as criteria for eligibility, giving rationale
Information 7 Describe all information sources (e.g., databases with dates of coverage,
sources contact with study authors to identify additional studies) in the search
and date last searched
Search 8 Present full electronic search strategy for at least one database, including
any limits used, such that it could be repeated
Study selection 9 State the process for selecting studies (i.e., screening, eligibility, included in
systematic review, and, if applicable, included in the meta-analysis)
Data collection 10 Describe method of data extraction from reports (e.g., piloted forms,
process independently, in duplicate) and any processes for obtaining and
confirming data from investigators
Data items 11 List and define all variables for which data were sought (e.g., PICOS,
funding sources) and any assumptions and simplifications made
Risk of bias in 12 Describe methods used for assessing risk of bias of individual studies
individual (including specification of whether this was done at the study or outcome
studies level), and how this information is to be used in any data synthesis
Summary 13 State the principal summary measures (e.g., risk ratio, difference in means)
measures
Synthesis of 14 Describe the methods of handling data and combining results of studies, if
results done, including measures of consistency (e.g., I2) for each meta-analysis

From Moher, D., Liberati, A., Tetzlaff, J., Altman, D. G., & PRISMA Group. (2009). Preferred Reporting Items for Systematic Reviews and Meta-Analyses:
The PRISMA Statement. Retrieved from http://www.prisma-statement.org.

p. 224) conducted a survival analysis on 100 quantita-
tive systematic reviews published from 1995-2005 “to
estimate the average time to changes in evidence that
is sufficiently important to warrant updating system-
atic reviews.” They found the average time before a
systematic review should be updated was 5.5 years.
However, 23% of the reviews signaled a need for
updating within 2 years, and 15% needed updating
within 1 year. Shojania et al. (2007) stressed that high-
quality systematic reviews that were directly relevant
to clinical practice require frequent updating to stay
current. There are numerous nursing and medical
 482 UNIT THREE  Putting It All Together for Evidence-Based Health Care
sources of systematic reviews, but it is important for
you to know the steps of the systematic review process
and to be able to appraise critically the quality of these
reviews.
Conducting Meta-Analyses to Synthesize
Research Evidence
A meta-analysis is conducted to pool statistically the
results from previous studies into a single quantitative
analysis that provides one of the highest levels of
evidence about the effectiveness of an intervention
(Andrel, Keith, & Leiby, 2009; Craig & Smyth, 2012;
Higgins & Green, 2008; Liberati et al., 2009). This
approach has objectivity because it includes analysis
techniques to determine the effect of an intervention
while examining the influences of variations in the
studies selected for the meta-analysis. The studies to
be included in the analysis need to be examined for
variations or heterogeneity in such areas as sample
characteristics, sample size, design, types of interven-
tion, and outcomes variables and measurement
methods (Higgins & Green, 2008). Heterogeneity in
the studies to be included in a meta-analysis can lead
to different types of biases, which are detailed in the
next section.
Statistically combining data from several studies
results in a large sample size with increased power to
determine the true effect of a specific intervention on
a particular outcome (see Chapter 15 for discussion of
power). The ultimate goal of a meta-analysis is to
determine if an intervention (1) significantly improves
outcomes, (2) has minimal or no effect on outcomes,
or (3) increases the risk of adverse events. Meta-
analysis is also an effective way to resolve conflicting
study findings and controversies that have arisen
related to a selected intervention. When conducting a
systematic review, authors might conduct a meta-
analysis on a group of similar studies to determine the
oasis-ebl|Rsalles|1476321133
effectiveness of an intervention (Higgins & Green,
2008).
Strong evidence for using an intervention in prac-
tice can be generated from a meta-analysis of multiple,
quality studies such as RCTs and quasi-experimental
studies. However, the conduct of a meta-analysis
depends on the accuracy, clarity, and completeness
of information presented in studies. Box 19-1 provides
a list of information that needs to be included in
a research report to facilitate the conduct of a
meta-analysis.
The steps for conducting a meta-analysis are similar
to the steps for conducting a systematic review that
were detailed in the previous section. The PRISMA
Statement introduced earlier provides clear directions
Box 19-1 Recommended Reporting
for Authors to Facilitate
Meta-Analysis
Demographic Variables Relevant to
Population Studied
Age
Gender
Marital status
Ethnicity
Education
Socioeconomic status
Methodological Characteristics
Sample size (experimental and control groups)
Type of sampling method
Sampling refusal rate and attrition rate
Sample characteristics
Research design
Groups included in study—experimental,
control, comparison, placebo groups
Intervention protocol and fidelity discussion
Data collection techniques
Outcome measurements
Reliability and validity of instruments
Precision and accuracy of physiological
measures
Data Analysis
Name of statistical tests
Sample size for each statistical test
Degrees of freedom for each statistical test
Exact value of each statistical test
Exact p value for each test statistic
One-tailed or two-tailed statistical test
Measures of central tendency (mean, median,
and mode)
Measures of dispersion (range, standard
deviation)
Post hoc test values for ANOVA (Analysis of
variance) test of three or more groups
for developing a report for either a systematic review
or a meta-analysis (see Table 19-3) (Liberati et al.,
2009). The following information is provided to
increase your ability to appraise critically meta-
analysis studies and to conduct a meta-analysis for
a selected intervention. The PRISMA Statement,
Cochrane Collaboration guidelines for meta-analysis
(Higgins & Green, 2008), and other resources (Andrel
et al., 2009; Conn & Rantz, 2003; Noordzij, Hooft,
 CHAPTER 19  Evidence Synthesis and Strategies for Evidence-Based Practice 483
Dekker, Zoccali, & Jager, 2009; Turlik, 2010) were
used to provide detail for conducting a meta-analysis.
Conn’s (2010) meta-analysis to determine the effect
of physical activity interventions on depressive
symptom outcomes in healthy adults is presented as
an example.
Clinical Question for Meta-Analysis
The clinical question developed for a meta-analysis is
usually clearly focused as: “What is the effectiveness
of a selected intervention?” The PICOS (participants,
intervention, comparative interventions, outcomes,
and study design) format discussed earlier might be
used to generate the clinical question (Higgins &
Green, 2008; Liberati et al., 2009; Moher et al., 2009).
Conn (2010) indicated that only one previous meta-
analysis had examined the effect of physical activities
(PA) on depressive symptoms among subjects without
clinical depression. Thus, she wanted to address the
following clinical question: “What is the effect of PA
on depressive symptoms in healthy adults?”
Purpose and Questions to Direct Meta-Analysis
Researchers need to identify clearly the purpose of
their meta-analysis and the questions or objectives
that guide the analysis. The Cochrane Collaboration
identified the following four basic questions to
guide a meta-analysis to determine the effect of an
intervention:
“1. What is the direction of effect?
“2. What is the size of effect?
“3. Is the effect consistent across studies?
“4. What is the strength of evidence for the effect?”
(Higgins & Green, 2008, p. 244)
Conn clearly identified the following purpose:
“This meta-analysis synthesized depressive symptom
outcomes of supervised and unsupervised PA inter-
ventions among healthy adults.… This meta-analysis
addressed the following research questions:
“(1) What are the overall effects of supervised PA
and unsupervised PA interventions on
depressive symptoms in healthy adults without
clinical depression?
“(2) Do interventions’ effects on depressive
symptom outcomes vary depending on
intervention, sample, and research design
characteristics?
“(3) What are the effects of interventions on
depressive symptoms among studies
comparing treatment subjects with before
versus after interventions?” (Conn, 2010,
pp. 128, 129)
Search Criteria and Strategies for Meta-Analyses
The methods for identifying search criteria and select-
ing search strategies are similar for meta-analyses and
systematic reviews. The search criteria are usually
narrowly focused for meta-analysis to identify selec-
tive studies examining the effect of a particular inter-
vention. The search needs to be rigorous and include
published sources identified through varied databases
and unpublished studies and grey literature identified
through other types of searches (see previous section).
Conn (2010) clearly identified her search strategies in
the following excerpt.
“Primary Study Search Strategies
“Multiple search strategies were used to ensure a
comprehensive search and thus limit bias while
moving beyond previous reviews. An expert refer-
ence librarian searched 11 computerized databases
(e.g., MEDLINE, PsychINFO, EMBASE) using broad
search terms.… Search terms for depressive symp-
toms were not used to narrow the search because
many PA interventions studies report depressive
symptom outcomes but do not consider these the
main outcomes of the study and thus papers are not
indexed by these terms. Several research registers
were examined including Computer Retrieval of
Information on Scientific Projects and mRCT, which
contains 14 active registers and 16 archived registers.
Computerized author searches were completed for
project principal investigators located from research
registers and for the first three authors on eligible
studies. Author searches were completed for disser-
tation authors to locate published papers. Ancestry
searches were conducted on eligible and review
papers. Hand searches were completed for 114
journals which frequently report PA intervention
research.” (Conn, 2010, p. 129)
484 UNIT THREE  Putting It All Together for Evidence-Based Health Care
Possible Biases for Meta-Analyses
and Systematic Reviews
Even with rigorous literature searches, authors of
meta-analyses and systematic reviews are often
limited to mainly published studies. The nature of the
sources can lead to biases and flawed or inaccurate
conclusions in the research syntheses. The common
biases that can occur in conducting and reporting
research syntheses include publication bias such as
time lag bias, location bias, duplicate publication
bias, citation bias, and language bias; bias from poor
study methodology; and outcome reporting bias.
Publication bias occurs because studies with positive
results are more likely to be published than studies
with negative or inconclusive results. Higgins and
Green (2008) found that the odds were four times
greater that positive study results would be published
versus negative results. Time-lag bias, a type of pub-
lication bias, occurs because studies with negative
results are usually published later, sometimes 2 to 3
years later, than studies with positive results. Some-
times studies with negative results are not published
at all, whereas studies with positive results might be
published more than once (duplicate publication
bias). Location bias can occur if studies are pub-
lished in lower impact journals and indexed in less-
searched databases. A citation bias occurs when
certain studies are cited more often than others and
are more likely to be identified in database searchers.
Language bias can occur if searches focus just on
studies in English and important studies exist in other
languages.
Biases in studies’ methodologies are often related
to design and data analysis problems. The strengths
and threats to design validity need to be examined
during critical appraisal of the studies for inclusion in
a meta-analysis or systematic review (see Chapter 10
for discussion of design validity). The analyses con-
ducted in studies need to be appropriate and complete
(see Chapters 20 through 25 on data analysis).
Outcome reporting bias occurs when study results
are not reported clearly and with complete accuracy.
For example, reporting bias occurs when researchers
selectively report positive results and not negative
results; or positive results might be addressed in detail
with limited discussion of negative results. Higgins
and Green (2008) provided a much more detailed dis-
cussion of potential biases in systematic reviews and
meta-analyses.
An analysis method called the funnel plot can be
used to assess for biases in a group of studies. Funnel
plots provide graphic representations of possible effect
sizes (ESs) or odds ratios (ORs) for interventions in
selected studies. The ES or strength of an intervention
in a study can be calculated by determining the differ-
ence between the experimental and control groups for
the outcome variable. The mean difference between
the experimental and control groups for several studies
is easily determined if the outcome variable is mea-
sured by the same scale or instrument in each study
(see Chapter 15 for calculation of ES). However, the
standardized mean difference (SMD) must be cal-
culated in a meta-analysis when the same outcome,
such as depression, is measured by different scales or
methods. More details are provided on SMD later in
this section. Figure 19-3 shows a funnel plot of the
SMDs from 13 individual studies. The SMDs from the
studies are fairly symmetrical or equally divided by
the line through the middle of the funnel in the graph.
A symmetrical funnel plot indicates limited or no pub-
lication bias. Asymmetry of the funnel plot is due to
publication bias, but Egger, Smith, Schneider, and
Minder (1997) also believed it reflects methodological
bias, reporting bias, heterogeneity in the studies’
sample size and interventions, and chance. In Figure
19-3, the studies with small sample sizes are toward
the bottom of the graph, and the studies with larger
samples are toward the top.
Figure 19-4 includes two example funnel plots with
the plot in Figure 19-4A showing no apparent publica-
tion bias. An unbiased sample of studies should appear
basically symmetrical in the funnel with the ORs of
the studies fairly equally divided on either side of the
line (see Chapter 24 for calculating OR). The funnel
plot shown in Figure 19-4B demonstrates publication
bias in favor of larger studies with positive results
when the studies having smaller effect and sample
sizes are removed. This collection of studies in a meta-
analysis could lead to the conclusion that a treatment
was effective when it might not be when looking at a
larger collection of studies with negative and positive
Study Size
0.0 0.1 0.2 0.3 0.4 0.5 0.6
Standardized Mean Difference (SMD) for RCTs
Figure 19-3  Funnel plot of standardized mean differences (SMDs) for
randomized controlled trials (RCTs) with limited bias.
 CHAPTER 19  Evidence Synthesis and Strategies for Evidence-Based Practice 485

Funnel plot with pseudo 95% confidence limits

0
Standard Error of Log Odds Ratio

.2

.4

.6

.8
–1 0 1 2
(A) Figure 19-4  A and B, Funnel plots examining
Log Odds Ratio
publication bias. (From Andrel, J. A., Keith, S. W.,
& Leiby, B. E. [2009]. Meta-analysis: A brief
introduction. Clinical & Translational Science, 2[5],
Funnel plot with pseudo 95% confidence limits
376.)
0
Standard Error of Log Odds Ratio

.2

.4

.6

.8
–1 0 1 2
(B) Log Odds Ratio

results as in the plot in Figure 19-4A. Conn (2010)
discussed her search results and risk of publication
bias in the following excerpt.
“Comprehensive searchers yielded 70 reports.…
The supervised PA two-group comparison included
1,598 subjects. The unsupervised PA two-group com-
parison included 1,081 subjects. The treatment
single-group comparisons included 1,639 supervised
PA and 3,420 unsupervised PA subjects.… Most
primary studies were published articles (s=54), and
the remainder were dissertations (s=14), book
chapter (s=1), and conference presentation materials
(s=1; s indicates the number of reports). Publication
bias was evident in the funnel plots for supervised and
unsupervised PA two-group outcome comparisons
and for treatment group, pre- vs. post-intervention
supervised PA and unsupervised PA comparisons. The
control group pre- and post-comparison distributions
on the funnel plots suggested less publication bias
than plots of treatment groups. Unless otherwise
specified, all results are from the treatment vs. control
comparisons.” (Conn, 2010, p. 131)
486 UNIT THREE  Putting It All Together for Evidence-Based Health Care
Results of Meta-Analysis for Continuous Outcomes
Many nursing studies examine continuous outcomes
or outcomes that are measured by methods that pro-
duced interval or ratio level data. Physiological mea-
sures to examine blood pressure produce ratio level
data. Likert scales such as the Center for Epidemio-
logic Studies Depression (CES-D) Scale produce
interval level data (see Figure 17-8 for a copy of
CES-D Scale). Blood pressure and depression are con-
tinuous outcomes. Meta-analysis includes a two-step
process: Step 1 is the calculation of a summary statis-
tic for each study to describe the intervention effect,
and step 2 is the summary (pooled) intervention effect
that is the weighted average of the interventions effects
estimated from the different studies. In step 1, to deter-
mine the effect of an intervention on continuous out-
comes, the mean difference between two groups is
calculated. The mean difference is a standard statistic
that is calculated to determine the absolute difference
between two groups. It is an estimate of the amount
of change caused by the intervention (e.g., physical
activity) on the outcome (e.g., depression) on average
compared with the control group. The mean difference
can be calculated to determine the effect of an inter-
vention only if the outcome is measured by the same
scale in all the studies (Higgins & Green, 2008).
A standardized mean difference (SMD), or d, is
used in studies as a summary statistic that is calculated
in a meta-analysis when the same outcome is mea-
sured by different scales or methods. The SMD is also
sometime referred to as the standardized mean effect
size. For example, in the meta-analysis by Conn
(2010), depression was commonly measured with
three different scales: Profile of Mood States, Beck
Depression Inventory, and CES-D Scale. Studies that
have differences in means in the same proportion to
the standard deviations have the same SMD (d) regard-
less of the scales used to measure the outcome vari-
able. The differences in the means and standard
deviations in the studies are assumed to be due to the
measurement scales and not variability in the outcome
(Higgins & Green, 2008). The SMD is calculated by
meta-analysis software, and the formula is provided
as follows:
difference in mean
outcome between groups
SMD ( d ) =
standard deviation of
outcome among participants
In step 2 of meta-analysis of summarizing the
effects of an intervention across studies, the pooled
intervention effect estimate is “calculated as a weighted
average of the intervention effects estimated in the
individual studies. A weighted average” is defined by
Higgins and Green (2008, p. 263) as:
sum of (estimate × weight)
Weighted average =
sum of weights
In combining intervention effect estimates across
studies, a random-effects meta-analysis model or
fixed-effect meta-analysis model can be used. The
assumption of using the random-effects model is that
all the studies are not estimating the same intervention
effect but related effects over studies that follow a
distribution across studies. When each study is esti-
mating the exact same quality, a fixed-effect model is
used. Meta-analysis results can be obtained using soft-
ware from SPSS and SAS statistical packages (see
Chapter 21). Cochrane Collaboration Review Manager
(RevMan) is software that can be used for conducting
meta-analyses. This chapter provides a very basic dis-
cussion of key ideas related to conducting meta-
analyses, and you are encouraged to review Higgins
and Green (2008) and other meta-analysis sources to
increase your understanding of this process (Andrel
et al., 2009; Fernandez & Tran, 2009; Turlik, 2010).
We also recommend the assistance of a statistician in
conducting these analyses.
Conn’s (2010) meta-analysis result identified a
standardized mean effect size of 0.372 between the
treatment and the control groups for the 38 supervised
PA studies and SMD of 0.522 among the 22 unsuper-
vised PA studies. This meta-analysis documented that
supervised and unsupervised PA reduced symptoms of
depression in healthy adults or adults without clinical
depression. Thus, a decrease in depression is another
important reason for encouraging patients to be
involved in physical activities.
Results of Meta-Analysis
for Dichotomous Outcomes
If the outcome data to be examined in a meta-analysis
are dichotomous, risk ratios, odds ratios, and risk dif-
ferences are usually calculated to determine the effect
of the intervention on the measured outcome. These
terms are introduced in this chapter but more informa-
tion is available in Craig and Smyth (2012), Higgins
and Green (2008), and Sackett et al. (2000). With
dichotomous data, every participant fits into one of
two categories, such as clinically improved or no clini-
cal improvement, effective screening device or inef-
fective screening device, or alive or dead. Risk ratio
(RR), also called relative risk, is the ratio of the risk
 CHAPTER 19  Evidence Synthesis and Strategies for Evidence-Based Practice 487
of subjects in the intervention group to the risk of
subjects in the control group for having a particular
health outcome. The health outcome is usually adverse,
such as the risk of a disease (e.g., cancer) or the risk
of complications or death (Higgins & Green, 2008).
The calculation for RR follows and is:.
risk of event in
experimental group
Relative risk ( RR) =
risk of event in control group
The odds ratio (OR) is defined as the ratio of the
odds of an event occurring in one group, such as the
treatment group, to the odds of it occurring in another
group, such as the standard care group. The OR is a
way of comparing whether the odds of a certain event
is the same for two groups (see Chapter 24). An
example is the odds of medication adherence or non-
adherence for an experimental group receiving an
intervention of education and specialized medication
packaging intervention versus a group receiving stan-
dard care. The calculation for OR is:
odds of event in
experimental group
Odds ratio (OR) =
odds of event in control
or comparison group
The risk difference (RD), also called the absolute
risk reduction, is the risk of an event in the experimen-
tal group minus the risk of the event in the control or
standard care group.
Risk difference ( RD) = risk of experimental group
− risk of control group
Meta-analysis results from studies with dichoto-
mous data are often presented using a forest plot.
Fernandez and Tran (2009) provided a format for pre-
senting a forest plot in a meta-analysis study (Figure
19-5). A forest plot usually includes the following
information: (1) author, year, and name of the study;
(2) raw data from the intervention and control groups
and total number in each group; (3) point estimate (OR
or RR) and confidence internal (CI) for each study
shown as a line and block on the graph; (4) numerical
values for point estimate (OR or RR) and CI for each
study; and (5) percent weights given to each study
(Fernandez & Tran, 2009; Higgins & Green, 2008). In
Figure 19-5, column 1 identifies each of the studies
using the clearest format for the studies being
analyzed. Column 2 includes the number of partici-
pants with the outcome (n) and total number of par-
ticipants in the intervention or experimental group,
expressed as n/N. Column 3 includes the number of
participants with the outcome and total number in the
control group. Column 4 graphically presents the OR
with a block and the 95% CI with a line. Column 5
provides the percent weights given to each of the three
studies in this example. Column 6 provides the numer-
ical values for the OR and 95% CI.
The bottom of the forest plot in Figure 19-5 pro-
vides a summary of results and significance including
total events for intervention and control groups, test
for heterogeneity, and test for overall effect. The large
diamond in the plot is the summary of the effect of
the studies included in the analysis. If the diamond is
left of the line that is positioned at 1, the results favor
the intervention or treatment, and the CI does not
include 1 if the results are statistically significant
(Fernandez & Tran, 2009). If the point estimates are
consistently more on one side of the vertical line, this
shows homogeneity of the studies. If the point esti-
mates are fairly equally distributed on both the left
and the right side of the veridical line, this shows
heterogeneity of the studies included in the meta-
analysis. The term heterogeneity was introduced
earlier; heterogeneity can exist in the sample size and
characteristics, types of intervention, designs, and
outcomes of the studies. Heterogeneity statistics for
random-effects meta-analyses include chi-square tests
(see Chapter 25), the I2, and a test for differences
across subgroups if they are appropriate (Higgins &
Green, 2008).
Magnus, Ping, Shen, Bourgeois, and Magnus
(2011) conducted a meta-analysis of the effectiveness
of mammography screening in reducing breast cancer
mortality in women 39 to 49 years old. Because mam-
mography screening is significant in reducing breast
cancer mortality of women older than 50 years and
early detection of breast cancer increases survival,
annual routine mammography screening has been rec-
ommended for all women 40 to 47 years old in the
United States. Thus, “the primary aim of the current
study was, after a quality assessment of identified ran-
domized controlled trials (RCTs), to conduct a meta-
analysis of the effectiveness of mammography
screening [intervention] in women aged 39-49 [popu-
lation] in reducing breast cancer mortality [dichoto-
mous outcome]. The second aim was to compare and
discuss the results of previously published meta-
analyses” (Magnus et al., 2011, p. 845). The following
excerpts describe the methods, results, and conclu-
sions of this meta-analysis.
488 UNIT THREE  Putting It All Together for Evidence-Based Health Care

Review: Intervention A for patients with diagnosis X

Details of the
Comparison: 01 Intervention A vs Control
review
Outcome: 01 Adverse events

Column 1 Column 2 Column 3 Column 4 Column 5 Column 6

Study Intervention Control OR (fixed) Weight OR (fixed)

Details and or sub-category n/N n/N 95% CI % 95% CI
results of the
Study 1 2006 18/168 26/107 63.33 0.37 [0.19, 0.72]
included studies
Study 2 2005 4/53 3/51 6.31 1.31 [0.28,6.15]
Study 3 2005 15/87 17/93 30.36 0.93 [0.43, 2.00]

Study authors

Line of no effect
and year n = Number of Graphical Proportion of Numerical
participants with display of weight of the display of
the outcome outcome study on the outcome
N = Total number of effect combined effect effect
participants

Total (95% CI) 308 251 100.00 0.60 [0.38, 0.96]

Total events: 37 (Intervention), 46 (Control)
Summary Test for heterogeneity: Chi2 = 4.22, df = 2(P = 0.12),
Summary effect
results and I2 = 52.6% Level of heterogeneity
Significance Test for overall effect:
Z = 2.12(P = 0.03) Statistical significance

0.1 0.2 0.5 1 2 5 10

Favors treatment Favors control

Figure 19-5  Meta-analysis graph for dichotomous data. CI, confidence interval; OR, odds ratio. (From Fernandez, R. S., & Tran, D. T. [2009]. The
meta-analysis graph: Clearing the haze. Clinical Nurse Specialist, 23[2], 58.)

“Methods: PubMed/MEDLINE, OVID, COCHRANE,
and Educational Resources Information Center (ERIC)
databases were searched, and extracted references
were reviewed. Dissertation abstracts and clinical
trials databases available online were assessed to
identify unpublished works. All assessments were
independently done by two reviewers. All trials
included were RCTs, published in English, included
data on women aged 39-49, and reported relative risk
(RR)/odds ratio (OR) or frequency data.
Results: Nine studies were identified.… The indi-
vidual trials were quality assessed, and the data were
extracted using predefined forms. Using the DerSi-
monian and Laird random effects model, the results
from the seven RCTs with the highest quality score
were combined, and a significant pooled RR estimate
of 0.83 (95% confidence interval [CI] 0.72-0.97) was
calculated.” (Magnus et al., 2011, p. 845)
The results of the study were graphically repre-
sented using a forest plot (Figure 19-6). The plot
clearly identifies the names of the seven studies
included in the meta-analysis on the left side of the
figure. The RR and CI for each study are identified
with a block and horizontal line. The numerical RR
and 95% CI values are identified on the right side of
the plot with the percent of weight given to each study.
Most of the studies show homogeneity with odds
ratios left of the vertical line except for the Stockholm
study. The forest plot would have been strengthened
by including the results from the test for heterogeneity
and test for overall effect. Magnus et al. (2011, p. 845)
concluded, “Mammography screenings were effective
and generate a 17% reduction in breast cancer mortal-
ity in women 39-49 years of age. The quality of the
trials varies, and providers should inform women in
this age group about the positive and negative aspects
of mammography screenings.”
 CHAPTER 19  Evidence Synthesis and Strategies for Evidence-Based Practice 489

Study Name Risk Ratio % Weight

(95% CI)

Stockhoim 1.52 (0.80, 2.88) 5.16

Gothenburg 0.55 (0.31, 0.95) 6.68

Maimo 0.73 (0.51, 1.04) 14.27

Age Trial 0.83 (0.66, 1.04) 26.89

Canada 0.97 (0.74, 1.27) 21.52

HIP 0.77 (0.53, 1.11) 13.45

Edinburgh 0.79 (0.53, 1.17) 12.03

Overall 0.83 (0.72, 0.97)

.25 .5 .75 1 1.25 1.5 1.75 2

Risk Ratio

Figure 19-6  Forest plot showing the individual randomized controlled trials and the overall pooled estimate from the seven original randomized
controlled trials with a high-quality score addressing the impact of mammography screening on breast cancer mortality in women 39 to 49 years old.
CI, Confidence interval. (From Magnus, M. C., Ping, M., Shen, M. M., Bourgeois, J., & Magnus, J. H. [2011]. Effectiveness of mammography screening in
reducing breast cancer mortality in women aged 39-49 years: A meta-analysis. Journal of Women’s Health, 20[6], 848.)

be used in practice and for policy development

Conducting Meta-Synthesis
of Qualitative Research
Qualitative research synthesis is the process and
product of systematically reviewing and formally inte-
grating the findings from qualitative studies (Sande-
lowski & Barroso, 2007). The process for conducting
a synthesis of qualitative research is still in the devel-
opmental phase. Various synthesis methods have
appeared in the literature, such as meta-synthesis,
meta-ethnography, meta-study, meta-narrative, quali-
tative metasummary, qualitative meta-analysis, and
aggregated analysis (Barnett-Page & Thomas, 2009;
Kent & Fineout-Overholt, 2008; Sandelowski &
Barroso, 2007; Walsh & Downe, 2005). Qualitative
researchers are not in agreement at the present
time about the method to use for synthesizing qualita-
tive research or if one method is possible to accom-
plish this process. Although the methodology is not
clearly developed for qualitative research synthesis,
researchers recognize the importance of summarizing
qualitative findings to determine knowledge that might
(Barnett-Page & Thomas, 2009; Finfgeld-Connett,
2010; Sandelowski & Barroso, 2007). The Cochrane
Collaboration recognizes the importance of synthesiz-
ing qualitative research, and the Cochrane Qualitative
Methods Group was developed as a forum for discus-
sion and development of methodology in this area
(Higgins & Green, 2008).
The qualitative research synthesis method that
seems to be gaining momentum in the nursing literature
is meta-synthesis. Methodological articles have been
published to describe meta-synthesis, but this method
is still in early phases of development (Finfgeld-
Connett, 2010; Kent & Fineout-Overholt, 2008; Walsh
& Downe, 2005). Meta-synthesis is defined as the
systematic compiling and integration of qualitative
study results to expand understanding and develop a
unique interpretation of study findings in a selected
area. The focus is on interpretation rather than the
combining of study results as with quantitative research
synthesis. Meta-synthesis involves the breaking down
of findings from different studies to discover essential
490 UNIT THREE  Putting It All Together for Evidence-Based Health Care
features and then the combining of these ideas into a
unique, transformed whole. Sandelowski and Barroso
(2007) identified metasummary as a step in conducting
meta-synthesis. Metasummary is the summarizing of
findings across qualitative reports to identify knowl-
edge in a selected area. A process for conducting a
meta-synthesis is described in the following section. A
meta-synthesis conducted by Denieffe and Gooney
(2011) of the symptom experience of women with
breast cancer is presented as an example.
Framing a Meta-Synthesis Exercise
Initially, researchers need to provide a frame for the
meta-synthesis to be conducted (Kent & Fineout-
Overholt, 2008; Walsh & Downe, 2005). Framing
involves identifying the focus and scope of the meta-
synthesis to be conducted. The focus of the meta-
synthesis is usually an important area of interest for
the individuals conducting it and a topic with an ade-
quate body of qualitative studies. The scope of a meta-
synthesis is an area of debate, with some qualitative
researchers recommending a narrow, precise approach
and others recommending a broader, more inclusive
approach. However, researchers recognize framing is
essential for making the synthesis process manageable
and the findings meaningful and potentially transfer-
able to practice. Framing the meta-synthesis is facili-
tated by the authors’ research and clinical expertise,
initial review of the relevant qualitative literature, and
discussion with expert qualitative researchers. Usually
a research question is developed to direct the meta-
synthesis process.
Denieffe and Gooney (2011) conducted their meta-
synthesis based on the stages developed by Sande-
lowski and Barroso (2007). These stages included
“identifying a research question, collecting relevant
data (qualitative studies), appraising the studies,
performing a metasummary and meta-synthesis”
(Denieffe and Gooney, 2011, p. 425). Denieffe and
Gooney developed the following question to direct
their meta-synthesis and provided a rationale for their
scope and focus.
“In this study the question was set as ‘What is the
symptom experience of women with breast cancer
from time of diagnosis to completion of treatment?’
The time frame selected from time of diagnosis to
completion of treatment, has been conceptualized …
as the ‘acute stage,’ encompassing initial diagnosis and
treatment in the first of a three-stage process of sur-
vivorship.” (Denieffe & Gooney, 2011, p. 425)
Searching the Literature and Selecting Sources
Most authors agree that a rigorous search of the litera-
ture needs to be conducted. The search needs to
include databases, books and book chapters, and full
reports of theses and dissertations. Special search
strategies that were identified earlier need to be
engaged to identify grey literature because qualitative
studies might be published in more obscure journals.
The search criteria need to identify the years of the
search, keywords to be searched, and language of
sources. Meta-syntheses are usually limited to qualita-
tive studies only and do not included mix-method
studies (Walsh & Downe, 2005). Also, qualitative
findings that have not been interpreted but are unana-
lyzed quotes, field notes, case histories, stories, or
poems are usually excluded (Finfgeld-Connett, 2010).
The search process is usually very fluid with the
conduct of additional computerized and hand searches
to identify more studies. Sandelowski and Barroso
(2007) identified a berry-picking process to search for
sources, which includes a dynamic process of modify-
ing search terms and methods to identify relevant
sources. However, it is important for researchers to
document systematically the strategies that they used
to search the literature and the sources found through
these different search strategies.
The final selection of studies to include in the meta-
synthesis depends on the focus and scope of the syn-
thesis. Some authors focus on one type of qualitative
research, such as ethnography, or one investigator in
a particular area. Others include studies with different
qualitative methodologies and investigators in a field
or related fields. The search criteria need to be consis-
tently implemented in determining the studies to be
included and excluded in the synthesis. A flow diagram
might be developed to identify the process for select-
ing the studies similar to the one identified for system-
atic reviews and meta-analyses (see Figure 19-1)
(Sandelowski & Barroso, 2007). Denieffe and Gooney
(2011) provided the following description of their lit-
erature search, search criteria, and selection of studies
for their meta-synthesis.
“Relevant qualitative research studies were located
and retrieved using computer searches in CINAHL,
PsychLIT, Academic Search Premier, Embase, and
MEDLINE. The research reports selected for this
synthesis met the following inclusion criteria: (1) the
study focused on women with breast cancer; (2)
there were explicit references to the use of
 CHAPTER 19  Evidence Synthesis and Strategies for Evidence-Based Practice 491
qualitative research methods; and (3) the study
focused on women’s perspectives and experiences of
symptoms with breast cancer. There were no restric-
tions related to the date the research was published.
Keywords used were breast cancer, experience,
symptom, and symptom experience.… The search
using electronic databases was supplemented by …
footnote chasing using reference lists, citation search-
ing, in addition to hand searching of journals, and
consultation with clinical colleagues and researchers
in the area. A total of 253 studies were identified as
being possibly relevant.… Only 31 studies were
found to be relevant to the research question and
included in the meta-synthesis. Reasons for this
reduction included papers that provided limited quali-
tative data, … did not address the research question,
… addressed post-treatment/survivor concerns, …
or data given may not have related to patients with
breast cancer.” (Denieffe & Gooney, 2011, pp.
425-426)
Appraisal of Studies and Analysis of Data
The critical appraisal process for qualitative research
varies among sources. We recommend that you use the
critical appraisal guidelines for qualitative research
presented in Chapter 18. These guidelines might be
used for examining the quality of individual studies
and a group of studies for a meta-synthesis. Usually a
table is developed as part of the appraisal process, but
this is also an area of debate. The table headings might
include (1) author and year of source, (2) aim or goal
of the study, (3) theoretical orientation, (4), method-
ological orientation, (5) type of findings, (6) sampling
plan, (7) sample size, and (8) other key content rele-
vant for comparison. This table provides a display of
relevant study elements so that a comparative appraisal
might be conducted (Sandelowski & Barroso, 2007;
Walsh & Downe, 2005). The comparative analysis of
studies involves examining methodology and findings
across studies for similarities and differences. The fre-
quency of similar findings might be recorded. The
differences or contradictions in studies need to be
resolved or explained (or both). Varied analysis tech-
niques often are used by the researchers to translate
the findings of the different studies into a new or
unique description.
Denieffe and Gooney (2011) developed a detailed
comparative analysis table of the 31 studies they
included in their meta-synthesis. Their table included
the headings mentioned in the previous paragraph and
the following: time frame from diagnosis, treatment,
age range, and ethnic origin. They indicated that the
“final stage of data analysis was the qualitative meta-
synthesis, interpreting the findings. Constant targeted
comparison within and between study findings was
undertaken, utilizing external literature to facilitate
interpretation of the emerging findings” (Denieffe &
Gooney, 2011, p. 426).
Discussion of Meta-Synthesis Findings
A meta-synthesis report might include findings pre-
sented in different formats based on the knowledge
developed and the perspective of the authors. A syn-
thesis of qualitative studies in an area might result in
the discovery of unique or more refined themes
explaining the area of synthesis. The findings from a
meta-synthesis might be presented in narrative format
or graphically presented in a model. The discussion of
findings also needs to include identification of the
limitations of the meta-synthesis. The report often
concludes with recommendations for further research
and possibly implications for practice or policy devel-
opment or both.
The synthesis by Denieffe and Gooney (2011) of
31 qualitative studies in the area of symptoms experi-
enced by women with breast cancer resulted in the
identification of four emerging themes: (1) breast
cancer and the impact on self, (2) self-image and
stigma, (3) self and self-control, and (4) more than just
a symptom. The researchers linked each of these
themes with the appropriate studies and presented this
information clearly in a table. Denieffe and Gooney
(2011) also developed a detailed model that linked the
themes about self to the diagnosis and treatment of the
women and the symptoms they experienced (Figure
19-7). The following excerpt provides the conclusions
from this meta-synthesis.
“The overarching idea emerging from this meta-
synthesis is that the symptoms experience for women
with breast cancer has effects on the very ‘self’ of the
individual. Emerging is women’s need to consider the
existential issues that they face while simultaneously
dealing with a multitude of physical and psychological
symptoms. This meta-synthesis develops a new, inte-
grated, and more complete interpretation of findings
on the symptom experience of women with breast
cancer. The results offer the clinician a greater under-
standing in depth and breadth than the findings from
individual studies on symptom experiences.” (Denieffe
& Gooney, 2011, p. 424)
492 UNIT THREE  Putting It All Together for Evidence-Based Health Care

Symptom Experience of Women with Breast Cancer

Pre-diagnosis Pain
Existential Issues
Fatigue
Time of Diagnosis

Time Event Line

Self-Image Anxiety
Surgery
Depression

Symptoms
Stigma Anticipations of
Hair loss
Treatment Chemotherapy
Skin changes
Loss of Control Initiation of
Self
and Autonomy Treatment Radiotherapy Weight Gain

Eating and Drinking

Leaving Active Hormone
Change in Roles Treatment therapy Nausea and Vomiting

Smell and Taste

Post Treatment Changes

Sexual Problems

Menopause / Fertility
Problems

Figure 19-7  Overall findings of meta-synthesis. (From Denieffe, S., & Gooney, M. [2011]. A meta-synthesis of women’s symptoms experience and
breast cancer. European Journal of Cancer Care, 20[4], 430.)

Mixed-Methods Systematic Reviews qualitative study designs are referred to as mixed-

In recent years, nurse researchers have been conduct- methods systematic reviews in this text. Mixed-
ing mixed-methods studies that include both quantita- methods systematic reviews have the potential to
tive and qualitative research methods (Creswell, 2009) contribute to Cochrane Interventions reviews for prac-
(see Chapter 10 for different types of mixed-methods tice and health policy in the following ways:
designs). Researchers recognize the importance of
synthesizing the findings of these studies to determine
important knowledge for practice and policy develop-
“1. In forming reviews by using evidence from
ment. For some synthesis areas, researchers need to
qualitative research to help define and refine a
combine the findings from both quantitative and quali-
question… .
tative studies to determine the current knowledge in
“2. Enhancing reviews by synthesizing evidence
that area. Harden and Thomas (2005) identified this
from qualitative research identified whilst
process of combining findings from quantitative and
looking for evidence … .
qualitative studies as mixed-methods synthesis.
“3. Extending reviews by undertaking a search and
Higgins and Green (2008) referred to this synthesis of
synthesis specially of evidence from qualitative
quantitative, qualitative, and mixed-methods studies
studies to address questions directly related to
as a mixed-method systematic review.
the effectiveness review.
The systematic reviews discussed earlier in this
“4. Supplementing reviews by synthesizing
chapter included only studies of a quantitative meth-
qualitative evidence to address questions on
odology, such as meta-analyses, RCTs, and quasi-
aspects other than effectiveness.” (Higgins &
experimental studies, to determine the effectiveness of
Green, 2008, p. 574)
an intervention. Mixed-methods systematic reviews
might include various study designs, such as qualita-
tive research and quasi-experimental, correlational,
and descriptive studies (Bettany-Saltikov, 2010b; Conducting mixed-methods systematic reviews
Higgins & Green, 2008; Liberati et al., 2009). Reviews involves implementing a complex synthesis process
that include syntheses of various quantitative and that includes expertise in synthesizing knowledge
 CHAPTER 19  Evidence Synthesis and Strategies for Evidence-Based Practice 493
from quantitative, qualitative, and mixed-methods
studies. Higgins and Green (2008) recommended two
types of approaches to integrate the findings from
quantitative, qualitative, and mixed-methods studies:
(1) multilevel syntheses and (2) parallel syntheses.
Multilevel synthesis involves synthesizing the find-
ings from quantitative studies separate from qualita-
tive studies and integrating the findings from these two
syntheses in the final report. Parallel synthesis
involves the separate synthesis of quantitative and
qualitative studies, but the findings from the qualita-
tive synthesis are used in interpreting the synthesized
quantitative studies.
Further work is needed to develop the methodology
for conducting a mixed-methods systematic review.
The steps overlap with the systematic review and
meta-synthesis processes that have been previously
described. The process might best be implemented
with a team of researchers with expertise in conducting
different types of studies and research syntheses. The
basic structure for the mixed-methods systematic
review might include the following: (1) identify
purpose and questions or aims of review; (2) develop
the review protocol that includes search strategies for
quantitative, qualitative, and mixed-methods studies;
(3) identify search criteria for quantitative studies; (4)
identify search criteria for qualitative and mixed-
methods studies; (5) conduct a rigorous search of the
literature; (6) select relevant quantitative, qualitative,
and mixed-method studies for synthesis; (7) construct
a table of information of studies to allow comparative
appraisal of the studies; (8) conduct critical appraisals
of the quality of quantitative and qualitative studies;
(9) synthesize study findings; and (10) develop a report
that integrates the results of syntheses for both quan-
titative, qualitative, and mixed-method studies. The
reader is encouraged to refer to the steps in systematic
review and meta-analysis for conducting quantitative
research syntheses and to the meta-synthesis discus-
sion for synthesizing qualitative studies.
Wulff, Cummings, Marck, and Yurtseven (2011)
conducted a mixed-methods systematic review to
examine the association of medication administration
technologies and patient safety. This review included
12 studies with the following designs: five preinter-
vention and postintervention studies, five correlational
studies, and two qualitative studies. The major focus
of this review was the synthesis of the 10 quantitative
studies that identified the benefits of implementing
medication administration technologies to improve
patient safety. However, the problem identified by
both the quantitative and the qualitative studies was
that nurses develop workarounds in implementing
different types of medication administration technolo-
gies that could compromise patient safety.
Models to Promote Evidence-
Based Practice in Nursing
Two models commonly used to facilitate EBP in
nursing are the Stetler Model of Research Utilization
to Facilitate EBP (Stetler, 2001) and the Iowa Model
of Evidence-Based Practice to Promote Quality of
Care (Titler et al., 2001). This section introduces these
two models that might be used to implement evidence-
based protocols, algorithms, and guidelines in clinical
agencies.
Stetler Model of Research Utilization
to Facilitate Evidence-Based Practice
An initial model for research utilization in nursing
was developed by Stetler and Marram in 1976 and
expanded and refined by Stetler in 1994 and 2001 to
promote EBP for nursing. The Stetler model (2001)
(see Figure 19-8) provides a comprehensive frame-
work to enhance the use of research evidence by
nurses to facilitate EBP. The research evidence can be
used at the institutional or individual level. At the
institutional level, synthesized research knowledge is
used to develop or update protocols, algorithms, poli-
cies, procedures, or other formal programs imple-
mented in the institution. Individual nurses, including
practitioners, educators, and policy makers, summa-
rize research and use the knowledge to influence edu-
cational programs, make practice decisions, and
impact political decision making. Stetler’s model is
included in this text to guide individual nurses and
healthcare institutions in using research evidence in
practice. The following sections briefly describe the
five phases of the Stetler model: (I) preparation, (II)
validation, (III) comparative evaluation and decision
making, (IV) translation and application, and (V)
evaluation (see Figure 19-8).
Phase I: Preparation
The intent of the Stetler model (2001) is to make using
research evidence in practice a conscious, critical
thinking process that is initiated by the user. The first
phase (preparation) involves determining the purpose,
focus, and potential outcomes of making an evidence-
based change in a clinical agency. The agency’s priori-
ties and other external and internal factors that could
be influenced by or could influence the proposed prac-
tice change need to be examined. After the purpose of
the evidence-based project has been identified and
 494

PHASE I: PHASE II: PHASE III: COMPARATIVE PHASE IV: PHASE V:

PREPARATION VALIDATION EVALUATION/DECISION MAKING TRANSLATION/APPLICATION EVALUATION

Evaluate dynamically:
1. Identify goal for
Perform utilization A. Confirm type, level, each “use”
focused critique and and method of application, 2. Obtain evidence re:
synopsis: per details in part II change process and
Search, sort, Identify and, if goal-related progress,
and select applicable, record as well as any
sources of key study details results/outcomes
research and qualifiers 3. Use iterative evidence
evidence Fit of Feasibility to achieve goals
setting (r,r,r) B. Use: Review
Accept operational details
Consider influential Reject
factors Synthesize Substan- Current • Informally: Use in practice
findings and tiating practice
Affirm priority evaluate per evidence • Formally: Identify design
criteria evidence-based document/s;
Define purpose Stop Evaluate as part of
package for dissemination;
and outcomes per as needed, develop routine practice
issue/catalyst E-B change plan,
including evaluation

OR

State decision/s B´. Consider use:

re: use of findings, • Informally: Obtain
per strength of evidence: targeted practice
A. Not use = Stop information; evaluate
• Formally: Do formal details
UNIT THREE  Putting It All Together for Evidence-Based Health Care

B. Use now as in B; plan/implement

a pilot "use" project,
B′. Consider use
with evaluation
• Per results, accept and
extend, with or
without modification,
OR
if reject = Stop

Figure 19-8  Stetler Model, part I: Steps of research utilization to facilitate EBP.
Stetler, C. B. (2001). Updating the Stetler Model of Research Utilization to facilitate evidence-based practice. Nursing Outlook, 42(6), 276.
 CHAPTER 19  Evidence Synthesis and Strategies for Evidence-Based Practice 495
approved by the agency, a detailed search of the litera-
ture is conducted to determine the strength of the
evidence available for use in practice. The research
literature might be reviewed to solve a difficult clini-
cal, managerial, or educational problem; to provide
the basis for a policy, standard, algorithm, or protocol;
or to prepare for an in-service program or other type
of professional presentation.
Phase II: Validation
In the validation phase, research reports are critically
appraised to determine their scientific soundness. If
the studies are limited in number or are weak or both,
the findings and conclusions are considered inade-
quate for use in practice, and the process stops. The
quality of the research evidence is greatly strength-
ened if a systematic review or meta-analysis has been
conducted in the area where you want to make an
evidence-based change. If the research knowledge
base is strong in the selected area, a decision must be
made regarding the priority of using the evidence in
practice by the clinical agency.
Phase III: Comparative Evaluation
and Decision Making
Comparative evaluation includes four parts: (1) sub-
stantiation of the evidence, (2) fit of the evidence with
the healthcare setting, (3) feasibility of using research
findings, and (4) concerns with current practice (Figure
19-8). Substantiating evidence is produced by replica-
tion, in which consistent, credible findings are obtained
from several studies in similar practice settings. The
studies generating the strongest research evidence
are RCTs and meta-analyses of RCTs and quasi-
experimental studies, which provide extremely strong
evidence about the effectiveness of nursing interven-
tions. To determine the fit of the evidence in the clini-
cal agency, the characteristics of the setting are
examined to determine the forces that would facilitate
or inhibit the evidence-based change. Stetler (2001)
believed the feasibility of using research evidence in
practice involved examining the three R’s related to
making changes in practice: (1) potential risks, (2)
resources needed, and (3) readiness of the people
involved. The final comparison involves determining
whether the research information provides credible,
empirical evidence for making changes in the current
practice. The research evidence needs to document
that an intervention increased the quality in current
practice by solving practice problems and improving
patient outcomes. By conducting phase III, the overall
benefits and risks of using the research evidence in
a practice setting can be assessed. If the benefits
(improved patient, provider, or agency outcomes) are
much greater than the risks (complications, morbidity,
mortality, or increased costs) for the organization, the
individual nurse, or both, using the research-based
intervention in practice is feasible.
Three types of decisions (decision making) are
possible during this phase: (1) to use the research
evidence, (2) to consider using the evidence, and (3)
not to use the research evidence. The decision to use
research knowledge in practice is determined mainly
by the strength of the evidence. Depending on the
research knowledge to be used in practice, the indi-
vidual practitioner, hospital unit, or agency might
make this decision. Another decision might be to con-
sider using the available research evidence in practice.
When a change is complex and involves multiple dis-
ciplines, the individuals involved often need additional
time to determine how the evidence might be used and
what measures will be taken to coordinate the involve-
ment of different health professionals in the change.
A final option might be not to make a change in prac-
tice because of the poor quality of the research evi-
dence, costs, and other potential problems.
Phase IV: Translation and Application
The translation and application phase involves
planning for and using the research evidence in prac-
tice. The translation phase involves determining
exactly what knowledge will be used and how that
knowledge will be applied to practice. The use of the
research evidence can be cognitive, instrumental, or
symbolic. Cognitive application is a more informal
use of the research knowledge to modify one’s way of
thinking or appreciation of an issue (Stetler, 2001).
Cognitive application may improve the nurse’s under-
standing of a situation, allow analysis of practice
dynamics, or improve problem-solving skills for clini-
cal problems. Instrumental and symbolic applications
are formal ways to make changes in practice. Instru-
ment application involves using research evidence to
support the need for change in nursing interventions
or practice protocols, algorithms, and guidelines.
Symbolic or political use occurs when information is
used to support or change an agency policy. The appli-
cation phase includes the following steps for planned
change: (1) assess the situation to be changed, (2)
develop a plan for change, and (3) implement the plan.
During the application phase, the protocols, policies,
procedures, or algorithms developed with research
knowledge are implemented in practice (Stetler, 2001).
A pilot project on a single hospital unit might be con-
ducted to implement the change in practice, and the
results of this project could be evaluated to determine
496 UNIT THREE  Putting It All Together for Evidence-Based Health Care
if the change should be extended throughout the
healthcare agency or corporation.
Phase V: Evaluation
The final stage is to evaluate the effect of the evidence-
based change on selected agency, personnel, or patient
outcomes. The evaluation process can include both
formal and informal activities that are conducted by
administrators, nurse clinicians, and other health pro-
fessionals (see Figure 19-8). Informal evaluations
might include self-monitoring or discussions with
patients, families, peers, and other professionals.
Formal evaluations can include case studies, audits,
quality assurance, and outcomes research projects.
The goal of the Stetler model (2001) is to increase the
use of research evidence in nursing to facilitate EBP.
This model provides detailed steps to encourage
nurses to become change agents and make the neces-
sary improvements in practice based on the best
current research evidence.
Iowa Model of Evidence-Based Practice
Nurses are actively involved in conducting research,
synthesizing research evidence, and developing
evidence-based guidelines for practice. Nurses have a
strong commitment to EBP and can benefit from the
direction provided by the Iowa model to expand their
research-based practice. The Iowa Model of Evidence-
Based Practice provides direction for the develop-
ment of EBP in a clinical agency. Titler et al. initially
developed this EBP model in 1994 and revised it in
2001. In a healthcare agency, triggers initiate the need
for change, and the focus should always be to make
changes based on best research evidence. These trig-
gers can be problem-focused and evolve from risk
management data, process improvement data, bench-
marking data, financial data, and clinical problems
(see Figure 19-9). The triggers can also be knowledge-
focused, such as new research findings, changes in
national agencies or organizational standards and
guidelines, an expanded philosophy of care, or ques-
tions from the institutional standards committee. The
triggers are evaluated and prioritized based on the
needs of the clinical agency. If a trigger is considered
an agency priority, a group is formed to search for the
best evidence to manage the clinical concern (Titler
et al., 2001).
In some situations, the research evidence is inad-
equate to make changes in practice, and additional
studies are needed to strengthen the knowledge base.
Sometimes the research evidence can be combined
with other sources of knowledge (theories, scientific
principles, expert opinion, and case reports) to provide
fairly strong evidence for developing research-based
protocols for practice. The strongest evidence is gen-
erated from meta-analyses of several RCTs, system-
atic reviews that usually include meta-analyses, and
individual studies. Systematic reviews provide the
best research evidence for developing evidence-based
guidelines. The research-based protocols or evidence-
based guidelines are pilot-tested on a particular unit
and then evaluated to determine the impact on patient
care (see Figure 19-9). If the outcomes are favorable
from the pilot test, the change is made in practice and
monitored over time to determine its impact on the
agency environment, staff, costs, and the patient and
family (Titler et al., 2001). An agency can promote
EBP by using the Iowa model to identify triggers for
change, implement patient care based on the best
research evidence, and monitor changes in practice to
ensure quality care.
Implementing Evidence-Based
Guidelines in Practice
EBP of nursing and medicine has expanded exten-
sively since the 1990s. Research knowledge is gener-
ated every day that needs to be critically appraised
and synthesized to determine the best evidence for
use in practice (Craig & Smyth, 2012; Higgins &
Green, 2008; Melnyk & Fineout-Overholt, 2011). This
section discusses the development of EBP guidelines
and provides a model for using these guidelines
in practice. Chobanian et al. (2003) conducted an
excellent systematic review to determine the best
research evidence available for assessing, diagnosing,
and managing hypertension. This systematic review,
which included several meta-analyses and integrative
reviews, was used to develop the JNC 7 evidence-
based guideline for hypertension (U.S. Department of
Health and Human Services, National Institutes of
Health, National Heart, Lung, and Blood Institute,
2003). The JNC 7 evidence-based guideline is pre-
sented later in this chapter. JNC-8 is being developed
by a national panel of expert researchers and clinicians
with an expected availability to the public for review
and comment by 2012 and publication to follow (see
the status of the guideline at http://www.nhlbi.nih.gov/
guidelines/hypertension/jnc8/index.htm). This section
focuses on the development and use of evidence-based
guidelines in practice.
Development of Evidence-Based Guidelines
Once a significant health topic or condition has
been selected, guidelines are developed to promote
 CHAPTER 19  Evidence Synthesis and Strategies for Evidence-Based Practice 497

Problem-focused triggers Knowledge-focused triggers

1. Risk management data
2. Process improvement data
3. Internal/external benchmarking data
4. Financial data
5. Identification of clinical problem
1. New research or other literature
2. National agencies or organizational
standards and guidelines
3. Philosophies of care
4. Observation from institutional standards committee

Is this topic
a priority
for the
organization?

No Yes

Consider Form a team

other
triggers
Assemble relevant research and related literature

Critique and synthesize research for use in practice

Is there
a sufficient
Yes research No
base?

Pilot the change in practice Base practice on other types of evidence: Conduct
1. Select outcomes to be achieved 1. Case reports research
2. Collect baseline data 2. Expert opinion
3. Design evidence-based 3. Scientific principles
practice (EBP) guideline(s) 4. Theory
4. Implement EBP on pilot units
5. Evaluate process and outcomes
6. Modify the practice guideline

Is change
appropriate for
No adoption in Yes
practice?

Continue to evaluate quality Institute the change in practice

of care and new knowledge

Monitor and analyze structure,

process, and outcome data
• Environment
• Staff
• Cost
• Patient and family

Disseminate results

Figure 19-9  Iowa Model of Evidence-Based Practice to Promote Quality Care.

Titler, M. G., Kleiber, C., Steelman, V. J., Rakel, B. A., Budreau, G., Everett, L. Q., et al. (2001). The Iowa Model of Evidence-Based Practice to promote
quality care. Critical Care Nursing Clinics of North America, 13(4), 500.
498 UNIT THREE  Putting It All Together for Evidence-Based Health Care
effective management of this health condition. Since
the 1980s, the Agency for Healthcare Research and
Quality (AHRQ) has had a major role in identifying
health topics and developing evidence-based guide-
lines for these topics (http://www.ahrq.gov). In the late
1980s and early 1990s, a panel or team of experts was
often charged with developing guidelines for the
AHRQ. The AHRQ solicited the members of the
panel, who usually included nationally recognized
researchers in the topic area; expert clinicians, such as
physicians, nurses, pharmacists, and social workers;
healthcare administrators; policy developers; econo-
mists; government representatives; and consumers.
The group designated the scope of the guidelines and
conducted extensive reviews of the literature includ-
ing relevant systematic reviews, meta-analyses, quali-
tative research syntheses, mixed-methods systematic
reviews, individual studies, and theories.
The best research evidence available was synthe-
sized to develop recommendations for practice. Most
of the evidence-based guidelines included systematic
reviews, meta-analyses, and multiple individual
studies. The guidelines were examined for their use-
fulness in clinical practice, their impact on health
policy, and their cost-effectiveness. Consultants, other
researchers, and additional expert clinicians often
were asked to review the guidelines and provide input.
Based on the experts’ critique, the AHRQ revised and
packaged the guidelines for distribution to healthcare
professionals. Some of the first guidelines focused on
the following healthcare problems: (1) acute pain
management in infants, children, and adolescents; (2)
prediction and prevention of pressure ulcers in adults;
(3) urinary incontinence in adults; (4) management of
functional impairments with cataracts; (5) detection,
diagnosis, and treatment of depression; (6) screening,
diagnosis, management, and counseling about sickle
cell disease; (7) management of cancer pain; (8) diag-
nosis and treatment of heart failure; (9) low back prob-
lems; and (10) otitis media diagnosis and management
in children.
At the present time, standardized guideline devel-
opment ranges from a structured process such as the
one just discussed to a less structured process in which
a guideline might be developed by a healthcare orga-
nization, healthcare plan, or professional organization.
The AHRQ initiated the National Guideline Clearing-
house (NGC, 2012b) in 1998 to store the evidence-
based practice guidelines. Initially, the NGC had 200
guidelines, but now the collection has expanded to
thousands of clinical practice guidelines. The NGC is
a publicly available database of evidence-based clini-
cal practice guidelines and related documents. Free
Internet access to guidelines is available at http://
www.guideline.gov. The NGC is updated weekly with
new content that the AHRQ produces in partnership
with the American Medical Association and the Amer-
ican Association of Health Plans (now America’s
Health Insurance Plans). The key components of the
NGC and its user-friendly resources can be found
on the AHRQ website at http://www.guideline.gov/
index.aspx. Some of the critical information on the
NGC is provided here to show you what is available
and how to access the NGC resources:
• “Structured abstracts (summaries) about the
guideline and its development.
• Links to full-text guidelines, where available, and/
or ordering information for print copies.
• Downloads of the Complete NGC Summary for
all guidelines represented in the database.
• A Guideline Comparison utility that gives users
the ability to generate side-by-side comparisons
for any combination of two or more guidelines.
• Unique guideline comparisons called Guideline
Syntheses prepared by NGC staff, compare
guidelines covering similar topics, highlighting
areas of similarity and difference. NGC Guideline
Syntheses often provide a comparison of
guidelines developed in different countries,
providing insight into commonalities and
differences in international health practices.
• An electronic forum, NGC-L for exchanging
information on clinical practice guidelines, their
development, implementation, and use.
• An Annotated Bibliography database where users
can search for citations for publications and
resources about guidelines, including guideline
development and methodology, structure,
evaluation, and implementation.
Other features include the following:
• What’s New enables users to see what guidelines
have been added each week and includes an
index of all guidelines in NGC.
• NGC Update Service is a weekly electronic
mailing of new and updated guidelines posted to
the NGC Web site.
• Detailed Search enables users to create very
specific search queries based on the various
attributes found in the NGC Classification
Scheme.
• NGC Browse permits users to scan for guidelines
available on the NGC site by disease/condition,
treatment/intervention, or developing
organization.
 CHAPTER 19  Evidence Synthesis and Strategies for Evidence-Based Practice 499
• Full-text guidelines and/or companion documents
available through the guideline developer that can
be downloaded.
• Glossary provides definitions of terms used in Neonatal Nurses: http://awhonn.org
the standardized abstracts (summaries).” (NGC,
2012a, http://www.guideline.gov/browse/
by-topic.aspx)
Criteria for submitting clinical practice guidelines
and the application process are provided online. Fol-
lowing are the criteria that an evidence-based guide-
line must meet to be submitted to the NGC:
• “The guideline must contain systematically
developed recommendations, strategies, or other
information to assist healthcare decision-making
in specific clinical circumstances.
• The guideline must have been produced under
the auspices of a relevant professional
organization (e.g., medical specialty society,
government agency, healthcare organization, or
health plan).
• The guideline development process must have
included a verifiable, systematic literature search
and review of existing evidence published in
peer-reviewed journals.
• The guideline must be current and the most
recent version (i.e., developed, reviewed, or
revised within the last 5 years).” (NGC, 2012b,
http://www.guideline.gov/)
The NGC provides varied audiences with an easy-
to-use mechanism for obtaining objective, detailed
information on clinical practice guidelines. In addi-
tion, the NGC (2012a) provides a list of the guidelines
that are in the process of being developed (http://
www.guideline.gov/browse/by-topic.aspx).
In addition to the evidence-based guidelines, the
AHRQ has developed many tools to assess the quality
of care that is provided by the evidence-based guide-
lines. You can search the AHRQ (2012a, 2012c)
website (http://www.qualitymeasures.ahrq.gov/) for
an appropriate tool to measure a variable in a research
project or to evaluate outcomes of care in a clinical
agency.
Numerous professional organizations, healthcare
agencies, universities, and other groups provide
evidence-based guidelines for practice. Websites are
as follows:
• Academic Center for Evidence-Based Nursing:
http://www.acestar.uthscsa.edu
• Association of Women’s Health, Obstetric, and
• Centers for Health Evidence.net: http://www
.cche.net
• CMA InfoBase: http://mdm.ca/cpgsnew/cpgs/
index.asp
• Guidelines Advisory Committee: http://www
.gacguidelines.ca
• Guidelines International Network: http://
www.g-i-n.net/
• HerbMed: Evidence-Based Herbal Database,
1998, Alternative Medicine Foundation: http://
www.herbmed.org/
• MD Consult: http://www.mdconsult.com/php/
286943359-1063/homepage
• National Association of Neonatal Nurses: http://
www.nann.org/
• National Institute for Clinical Excellence (NICE):
http://www.nice.org.uk/catcg2.asp?c=20034
• Oncology Nursing Society: http://www.ons.org/
• PIER—the Physicians’ Information and Education
Resource (authoritative, evidence-based guidance
to improve clinical care; ACP-ASIM members
only): http://pier.acponline.org/index.html
• Primary Care Clinical Practice Guidelines: http://
www.medscape.com/pages/editorial/public/pguide
lines/index-primarycare
• U.S. Preventive Services Task Force: http://www
.uspreventiveservicestaskforce.org/about.htm
Implementing the Joint National Committee
on Prevention, Detection, Evaluation,
and Treatment of High Blood Pressure
Evidence-Based Guideline in Practice
Evidence-based guidelines have become the standards
for providing care to patients in the United States and
other nations. A few nurses have participated in com-
mittees that have developed these evidence-based
guidelines, and many APNs are using these guidelines
in their practices. An evidence-based guideline for the
assessment, diagnosis, and management of high blood
pressure is provided as an example. This guideline
was developed from JNC 7 and was published in the
Journal of the American Medical Association (Choba-
nian et al., 2003). The NIH Department of Health and
Human Services National Heart, Lung, and Blood
Institute developed educational materials to commu-
nicate the specifics of this guideline to promote its use
by healthcare providers. This guideline is presented in
500 UNIT THREE  Putting It All Together for Evidence-Based Health Care

EVALUATION TREATMENT
Classification of blood pressure (BP)* Principles of hypertension treatment
Category SBP mm Hg DBP mm Hg • Treat to BP <140/90 mmHg or BP <130/80 mm Hg
Normal <120 and <80 in patients with diabetes or chronic kidney disease.
• Majority of patients will require two medications
Prehypertension 120–139 or 80–89
to reach goal.
Hypertension, Stage 1 140–159 or 90–99
Hypertension, Stage 2 ≥160 or ≥100 Algorithm for treatment of hypertension
*See Blood Pressure Measurement Techniques (reverse side)
Key: SBP = systolic blood pressure Lifestyle modifications
DBP = diastolic blood pressure
Diagnostic workup of hypertension Not at goal blood pressure (<140/90 mm Hg)
• Assess risk factors and comorbidities. (<130/80 mm Hg for patients with diabetes
• Reveal identifiable causes of hypertension. or chronic kidney disease)
• Assess presence of target organ damage. See Strategies for Improving Adherence to Therapy
• Conduct history and physical examination.
• Obtain laboratory tests: urinalysis, blood glucose, hematocrit.
and lipid panel, serum potassium, creatinine, and calcium. Initial drug choices
Optional: urinary albumin/creatinine ratio.
• Obtain electrocardiogram.
Assess for major cardiovascular disease (CVD) Without compelling With compelling
risk factors indications indications
• Hypertension • Physical inactivity
• Obesity • Microalbuminuria, estimated
(body mass index ≥30 kg/m )
2 glomerular filtration rate <60 mL/min Stage 1 Stage 2 Drug(s) for the
• Dyslipidemia • Age (>55 for men, >65 for women) hypertension hypertension compelling
• Diabetes mellitus • Family history of premature CVD (SBP 140–159 or (SBP ≥160 or indications
(men age <55, women age <65) DBP 90–99 mm Hg) See Compelling
• Cigarette smoking DBP ≥100 mm Hg) Indications for
Thiazide-type diuretics 2-drug combination Individual Drug Classes
Assess for identifiable causes of hypertension for most. May consider for most (usually Other
ACEI, ARB, BB, CCB, thiazide-type diuretic antihypertensive
• Sleep apnea • Cushing’s syndrome or or combination. and ACEI, or ARB, drugs (diuretics, ACEI,
• Drug induced/related steroid therapy
or BB, or CCB). ARB, BB, CCB)
• Chronic kidney disease • Pheochromocytoma
• Primary aldosteronism • Coarctation of aorta as needed.
• Renovascular disease • Thyroid/parathyroid disease

Not at goal blood pressure

Optimize dosages or add additional drugs until goal

blood pressure is achieved. Consider consultation with
hypertension specialist.
See Strategies for Improving Adherence to Therapy

Figure 19-10  Reference card from the seventh report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High
Blood Pressure (JNC 7). U.S. Department of Health and Human Services, National Institutes of Health, National Heart, Lung, and Blood Institute.
(2003). Reference card from the Seventh report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood
Pressure (JNC 7). Bethesda, MD: NIH Publication No. 03-5231. Retrieved from www.nhlbi.nih.gov/guidelines/hypertension/jnc7card.htm.

Figure 19-10 and provides clinicians with direction for
the following: (1) classification of blood pressure as
normal, prehypertension, hypertension stage 1, and
hypertension stage 2; (2) conduct a diagnostic workup
of hypertension; (3) assessment of the major cardio-
vascular disease risk factors; (4) assessment of the
identification of causes of hypertension; and (5) treat-
ment of hypertension. An algorithm provides direction
for the selection of the most appropriate treatment
methods for each patient diagnosed with hypertension
(U.S. Department of Health and Human Services,
National Institutes of Health, National Heart, Lung,
and Blood Institute, 2003).
APNs and RNs need to assess the usefulness and
quality of each evidence-based guideline before they
implement it in their practice. Figure 19-11 presents
the Grove Model for Implementing Evidence-Based
Guidelines in Practice. In this model, nurses identify
a practice problem, search for the best research evi-
dence to manage the problem in their practice, and
note that an evidence-based guideline has been devel-
oped. The quality and usefulness of the guideline must
be assessed by the healthcare provider before it is used
in practice, and that involves examining the following:
(1) the authors of the guideline, (2) the significance of
the healthcare problem, (3) the strength of the research
evidence, (4) the link to national standards, and (5) the
cost-effectiveness of using the guideline in practice.
The quality of the JNC 7 guideline is examined as an
example using the four criteria identified in the Grove
 CHAPTER 19  Evidence Synthesis and Strategies for Evidence-Based Practice 501

Practice problem

Search for best evidence yields evidence-based guideline

Quality of evidence-based guideline

Authors of guidelines Significance of Strength of Link to national Cost

healthcare problem research evidence standards effectiveness
analysis

Integration of evidence-based guideline with clinical expertise

Use of evidence-based guideline in practice by

Patient, provider, healthcare agency

Monitor outcomes from use of evidence-based guideline by

Patient, provider, healthcare agency

Refinement of evidence-based guideline

Evidence-based practice

Figure 19-11  Grove Model for Implementing Evidence-Based Guidelines in Practice.

model (see Figure 19-11). The authors of the JNC 7

guideline were expert researchers, clinicians (physi-
cians), policy developers, healthcare administrators,
and the National High Blood Pressure Education
Program Coordinating Committee. These authors have
the expertise to develop an evidence-based guideline
for the significant health problem of hypertension.
35 million office visits as the primary diagnosis.…
Recent clinical trials have demonstrated that effective
BP [blood pressure] control can be achieved in most
patients with hypertension, but the majority will
require 2 or more antihypertensive drugs.” (Choba-
nian et al., 2003, p. 2562)

“Hypertension is a significant healthcare problem
because it affects approximately 50 million individuals
in the United States and approximately 1 billion indi-
viduals worldwide.… Hypertension is the most
common primary diagnosis in the United States with
The research evidence for the development of the
JNC 7 guideline was extremely strong. The JNC 7
report included 81 references; 9 (11%) of the refer-
ences were meta-analyses, and 35 (43%) were RCTs
(experimental studies); 44 (54%) sources are consid-
ered extremely strong research evidence. The other
502 UNIT THREE  Putting It All Together for Evidence-Based Health Care
references were strong and included retrospective
analyses or case-controlled studies, prospective or
cohort studies, cross-sectional surveys or prevalence
studies, and clinical intervention studies (nonrandom)
(Chobanian et al., 2003). The JNC 7 provides the
national standard for the assessment, diagnosis, and
treatment of hypertension. The recommendations
from the JNC 7 are supported by the Department of
Health and Human Services and disseminated through
NIH publication no. 03-5231. Use of the JNC 7 guide-
line in practice is cost-effective because the clinical
trials have shown that “antihypertensive therapy has
been associated with 35% to 40% mean reductions in
stroke incidence; 20% to 25% in myocardial infarction
[MI]; and more than 50% in HF [heart failure]” (Cho-
banian et al., 2003, p. 2562).
The next step is for APNs and physicians to use the
JNC 7 guideline in their practice (see Figure 19-11).
Healthcare providers can assess the adequacy of the
guideline for their practice and modify the hyperten-
sion treatments based on the individual health needs
and values of their patients. The outcomes for the
patient, provider, and healthcare agency need to be
examined. The outcomes would be recorded in the
patients’ charts and possibly in a database owing to
the electronic medical records and would include the
following: (1) blood pressure readings for patients; (2)
incidence of diagnosis of hypertension based on the
JNC 7 guidelines; (3) appropriateness of the treat-
ments implemented to manage hypertension; and (4)
incidence of stroke, MI, and HF over 5, 10, 15, and
20 years. The healthcare agency outcomes include the
access to care by patients with hypertension, patient
satisfaction with care, and the cost related to diagnosis
and treatment of hypertension and the complications
of stroke, MI, and HF. This EBP guideline will be
refined in the future based on clinical outcomes,
outcome studies, and new controlled clinical trials (see
previous discussion of JNC 8). The use of this
evidence-based guideline and additional guidelines
promote an EBP for APNs (see Figure 19-10).
Evidence-Based Practice Centers
In 1997, the AHRQ launched its initiative to promote
EBP by establishing 12 evidence-based practice
centers (EPCs) in the United States and Canada.
“[The EPCs develop evidence reports and technology
assessments on] topics relevant to clinical, social
science/behavioral, economic, and other health care
organization and delivery issues—specifically those
that are common, expensive, and/or significant for
the Medicare and Medicaid populations. With this
program, AHRQ became a ‘science partner’ with
private and public organizations in their efforts to
improve the quality, effectiveness, and appropriate-
ness of health care by synthesizing the evidence and
facilitating the translation of evidence-based research
findings. Topics are nominated by non-federal part-
ners such as professional societies, health plans, insur-
ers, employers, and patient groups.” (AHRQ, 2012b,
http://www.ahrq.gov/clinic/epc/)
Under the EPC Program, the AHRQ awards 5-year
contracts to institutions to serve as EPCs. EPCs review
all relevant scientific literature on clinical, behavioral,
organizational, and financial topics to produce evi-
dence reports and technology assessments. These
reports are used to inform and develop coverage deci-
sions, quality measures, educational materials, tools,
guidelines, and research agendas. The EPCs also
conduct research on methodology of systematic
reviews. The AHRQ developed the following criteria
as the basis for selecting a topic to be managed by an
EPC:
• “High incidence or prevalence in the general
population and in special populations, including
women, racial and ethnic minorities, pediatric
and elderly populations, and those of low
socioeconomic status.
• Significance for the needs of the Medicare,
Medicaid, and other Federal health programs.
• High costs associated with a condition,
procedure, treatment, or technology, whether
due to the number of people needing care, high
unit cost of care, or high indirect costs.
• Controversy or uncertainty about the
effectiveness or relative effectiveness of available
clinical strategies or technologies.
• Impact potential for informing and improving
patient or provider decision making.
• Impact potential for reducing clinically significant
variations in the prevention, diagnosis, treatment,
or management of a disease or condition; in the
use of a procedure or technology; or in the
health outcomes achieved.
• Availability of scientific data to support the
systematic review and analysis of the topic.
 CHAPTER 19  Evidence Synthesis and Strategies for Evidence-Based Practice 503
• Submission of the nominating organization’s plan
to incorporate the report into its managerial or
policy decision making, as defined above.
• Submission of the nominating organization’s plan
to disseminate derivative products to its
members and plan to measure members’ use of
these products, and the resultant impact of such
use on clinical practice.” (AHRQ, 2012b)
The AHRQ (2012b) website (http://www.ahrq.gov/
clinic/epc) provides the names of the EPCs and the
focus of each center. This site also provides a link to
the evidence-based reports produced by these centers.
These EPCs have had an important role in the devel-
opment of evidence-based guidelines since the 1990s
and will continue to make significant contributions to
EBP in the future. They are also involved in the devel-
opment of measurement tools to examine the out-
comes from EBP.
Introduction to
Translational Research
Some of the barriers to EBP have resulted in the
development of a new type of research to improve the
translation of research knowledge to practice. This
new research methodology is called transitional
research and is being supported by the NIH (2012).
Translational research is an evolving concept that
is defined by the NIH as the translation of basic sci-
entific discoveries into practical applications. Basic
research discoveries from the laboratory setting need
to be tested in studies with humans. In addition, the
outcomes from human clinical trials need to be
adopted and maintained in clinical practice. Transla-
tional research is being encouraged by both medicine
and nursing to increase the implementation of
evidence-based interventions in practice and to deter-
mine if these interventions are effective in producing
the outcomes desired in clinical practice (Chesla,
2008; NIH, 2012). Translational research was origi-
nally part of the National Center for Research
Resources. However, in December 2011, the National
Center for Advancing Translation Sciences (NCATS)
was developed as part of the NIH Institutes and
Centers (NIH, 2012).
The NIH wanted to encourage researchers to
conduct translational research and developed the Clin-
ical and Translational Science Awards (CTSA)
Consortium in October 2006. The consortium started
with 12 centers located throughout the United States
and expanded to 39 centers in April 2009. The program
is projected to be fully implemented in 2012 with
about 60 institutions involved in clinical and transla-
tional science. A website has been developed to
enhance communication and encourage sharing of
information related to translational research projects
at http://www.ctsaweb.org/.
The CTSA Consortium is mainly focused on
expanding the translation of medical research to prac-
tice. Titler (2004, p. S1) defined transitional research
for the nursing profession as the: “Scientific investiga-
tion of methods, interventions, and variables that
influence adoption of evidence-based practices (EBPs)
by individuals and organizations to improve clinical
and operational decision making in health care. This
includes testing the effect of interventions on promot-
ing and sustaining the adoption of EBPs.” Baumbusch
et al. (2008) developed a collaborative model for
knowledge translation between research and practice
in clinical settings. As you search the literature for
relevant research syntheses and studies, you will note
that translation studies are being published in nursing.
Whittemore, Melkus, Wagner, Dziura, Northrup, and
Grey (2009) conducted a translation study to promote
the transfer of a Diabetic Prevention Program to
primary care. These types of studies will assist you in
translating research findings to your practice and
determining the impact of EBP on patients’ health.
However, national funding is needed to expand the
conduct of transitional research in nursing.
We hope that the content in this chapter increases
your understanding of EBP, the conduct of research
syntheses, the application of EBP models, and the
implementation of EBP guidelines. We encourage you
to take an active role in moving nursing toward EBP
that improves outcomes for patients, healthcare pro-
fessionals, and healthcare agencies.
KEY POINTS
• Evidence-based practice (EBP) is the conscientious
integration of best research evidence with clinical
expertise and patient values and needs in the deliv-
ery of quality, cost-effective health care. Best
research evidence is produced by the conduct and
synthesis of numerous, high-quality studies in a
health-related area.
• There are benefits and barriers associated with
EBP. The benefits of EBP are that the standards for
hospital accreditation by the Joint Commission
504 UNIT THREE  Putting It All Together for Evidence-Based Health Care
support EBP as does the Magnet Hospital Program®
managed by the American Nurses’ Credentialing
Center.
• Guidelines are provided for conducting the research
synthesis processes of systematic review, meta-
analysis, meta-synthesis, and mixed-methods sys-
tematic review. These synthesis processes are used
to determine the best research evidence in a selected
area and the quality of the research evidence avail-
able for practice.
• A systematic review is a structured, comprehensive
synthesis of the research literature to determine the
best research evidence available to address a
healthcare question. A systematic review involves
identifying, locating, appraising, and synthesizing
quality research evidence for expert clinicians to
use to promote EBP.
• A meta-analysis is conducted to pool statistically
the results from previous studies into a single quan-
titative analysis that provides one of the highest
levels of evidence about the effectiveness of an
intervention.
• Meta-synthesis is defined as the systematic compil-
ing and integration of qualitative study results to
expand understanding and develop a unique inter-
pretation of study findings in a selected area. The
focus is on interpretation rather than the combining
of study results as with quantitative research
synthesis.
• Reviews that include syntheses of various quantita-
tive, qualitative, and mixed-methods studies are
referred to as mixed-methods systematic reviews in
this text.
• Two models have been developed to promote EBP
in nursing: the Stetler Model of Research Utiliza-
tion to Facilitate EBP (Stetler, 2001) and the Iowa
Model of Evidence-Based Practice to Promote
Quality of Care (Titler et al., 2001).
• The phases of the revised Stetler model are (I)
preparation, (II) validation, (III) comparative eval-
uation and decision making, (IV) translation and
application, and (V) evaluation.
• The Iowa model provides guidelines for imple-
menting patient care based on the best research
evidence and monitoring changes in practice to
ensure quality care.
• The process for developing evidence-based guide-
lines is described, and an example of the guideline
for assessment, diagnosis, and treatment of hyper-
tension is provided.
• The Grove Model for Implementing Evidence-
Based Guidelines in Practice is provided to assist
nurses in determining the quality of evidence-based
guidelines and the steps for using these guidelines
in practice.
• An excellent source for evidence-based guidelines
is the National Guideline Clearinghouse (NGC)
that was initiated by the AHRQ in 1998.
• Evidence-based practice centers (EPCs), created by
the AHRQ in 1997, have had an important role in
the conduct of research, development of systematic
reviews, and formulation of evidence-based guide-
lines in selected practice areas.
• Translational research is an evolving concept that
is defined by the NIH as the translation of basic
scientific discoveries into practical applications.
REFERENCES
Agency for Healthcare Research and Quality (AHRQ). (2012a).
AHRQ tools and resources for better health care. Retrieved from
http://www.ahrq.gov/qual/tools/toolsria.htm.
Agency for Healthcare Research and Quality (AHRQ). (2012b).
Evidence-based practice centers: Synthesizing scientific evidence
to improve quality and effectiveness in health care. Retrieved from
http://www.ahrq.gov/clinic/epc/.
Agency for Healthcare Research and Quality (AHRQ). (2012c).
National Quality Measures Clearinghouse (NQMC). Retrieved
October 7, 2011, from http://qualitymeasures.ahrq.gov/.
American Nurses Credentialing Center (ANCC). (2012). Magnet
program overview. Retrieved from http://www.nursecredentialing
.org/Magnet/ProgramOverview.aspx.
Andrel, J. A., Keith, S. W., & Leiby, B. E. (2009). Meta-analysis:
A brief introduction. Clinical & Translational Science, 2(5),
374–378.
Barnett-Page, E., & Thomas, J. (2009). Methods for the synthesis
of qualitative research: A critical review. BMC Medical Research
Methodology, 9, 59. DOI: 10.1186/147-2288-9-59.
Baumbusch, J. L., Kirkham, S. R., Khan, K. B., McDonald, H.,
Semeniuk, P., Tan, E., et al. (2008). Pursuing common agendas:
A collaborative model for knowledge translation between research
and practice in clinical settings. Research in Nursing & Health,
31(2), 130–140.
Benzies, K. M., Premji, S., Hayden, K. A., & Serrett, K. (2006).
State-of-the-evidence reviews: Advantages and challenges of
including grey literature. Worldviews on Evidence-Based Nursing,
3(2), 55–61.
Bettany-Saltikov, J. (2010a). Learning how to undertake a system-
atic review: Part 1. Nursing Standard, 24(50), 47–56.
Bettany-Saltikov, J. (2010b). Learning how to undertake a system-
atic review: Part 2. Nursing Standard, 24(51), 47–58.
Biswas, R., Umakanth, S., Strumberg, J., Martin, C. M., Hande, M.,
& Nagra, J. S. (2007). The process of evidence-based medicine
and the search for meaning. Journal of Evaluation in Clinical
Practice, 13(4), 529–532.
Bolton, L. B., Donaldson, N. E., Rutledge, D. N., Bennett, C., &
Brown, D. S. (2007). The impact of nursing interventions: Over-
view of effective interventions, outcomes, measures, and priorities
for future research. Medical Care Research & Review, 64(Suppl
2), 123S-143S.
 CHAPTER 19  Evidence Synthesis and Strategies for Evidence-Based Practice 505
Brown, S. J. (2009). Evidence-based nursing: The research-practice
connection. Sudbury, MA: Jones & Bartlett Publishers.
Butler, K. D. (2011). Nurse practitioners and evidence-based
nursing practice. Clinical Scholars Review, 4(1), 53–57.
Chesla, C. A. (2008). Translational research: Essential contributions
from interpretive nursing science. Research in Nursing & Health,
31(4), 381–390.
Chobanian, A. V., Bakris, G. L., Black, H. R., Cushman, W. C.,
Green, L. A., Izzo, J. L., et al. (2003). The seventh report of
the Joint National Committee on Prevention, Detection, Evalua-
tion, and Treatment of high blood pressure: The JNC 7 report.
Journal of the American Medical Association, 289(19), 2560–
2572.
Cochrane Collaboration. (2012). Cochrane reviews. Retrieved from
http://www.cochrane.org/cochrane-reviews.
Conn, V. S. (2010). Depressive symptom outcomes of physical
activity interventions: Meta-analysis findings. Annals of Behav-
ioral Medicine, 39(2), 128–138.
Conn, V. S., & Rantz, M. J. (2003). Research methods: Managing
primary study quality in meta-analyses. Research in Nursing &
Health, 26(4), 322–333.
Conn, V. S., Valentine, J. C., Cooper, H. M., & Rantz, M. J. (2003).
Methods: Grey literature in meta-analyses. Nursing Research,
52(4), 256–261.
Craig, J. V., & Smyth, R. L. (2012). The evidence-based practice
manual for nurses (3rd ed.). Edinburgh, UK: Churchill
Livingstone.
Creswell, J. W. (2009). Research design: Qualitative, quantitative,
and mixed-methods approaches. Los Angeles, CA: Sage.
Denieffe, S., & Gooney, M. (2011). A meta-synthesis of women’s
symptoms experience and breast cancer. European Journal of
Cancer Care, 20(4), 424–435.
Doran, D. (2011). Nursing-sensitive outcomes: The state of the
science (2nd ed.). Sudbury, MA: Jones & Bartlett Learning.
Egger, M., Smith, G. D., Schneider, M., & Minder, C. (1997). Bias
in meta-analysis detected by a simple graphical test. British
Medical Journal, 315, 629–634.
Eizenberg, M. M. (2010). Implementation of evidence-based
nursing practice: Nurses’ personal and professional factors?
Journal of Advanced Nursing, 67(1), 33–42.
Fernandez, R. S., & Tran, D. T. (2009). The meta-analysis
graph: Clearing the haze. Clinical Nurse Specialist, 23(2),
57–60.
Finfgeld-Connett, D. (2010). Generalizability and transferability of
meta-synthesis research findings. Journal of Advanced Nursing,
66(2), 246–254.
Gerrish, K., Guillaume, L., Kirshbaum, M., McDonnell, A., Tod,
A., & Nolan, M. (2011). Factors influencing the contribution of
advanced practice nurses to promoting evidence-based practice
among front-line nurses: Findings from a cross-sectional survey.
Journal of Advanced Nursing, 67(5), 1079–1090.
Goulding, L., Furze, G., & Birks, Y. (2010). Randomized controlled
trials of interventions to change maladaptive illness beliefs in
people with coronary heart disease: Systematic review. Journal of
Advanced Nursing, 66(5), 946–961.
Harden, A., & Thomas, J. (2005). Methodological issues in combin-
ing diverse study types in systematic reviews. International
Journal of Social Research Methodology, 8(3), 257–271.
Higgins, J. P. T., & Green, S. (2008). Cochrane handbook for sys-
tematic reviews of interventions. West Sussex, UK: Wiley-
Blackwell & The Cochrane Collaboration.
Horstman, P., & Fanning, M. (2010). Tips for writing magnet evi-
dence. Journal of Nursing Administration, 40(1), 4–6.
Institute of Medicine. (2001). Crossing the quality chasm: A new
health system for the 21st century. Washington, DC: National
Academy Press.
Joanna Briggs Institute. (2012). Search the Joanna Briggs Institute.
Retrieved from http://www.joannabriggs.edu.au/Search.aspx.
Kent, B., & Fineout-Overholt, E. (2008). Using meta-synthesis to
facilitate evidence-based practice. Worldviews on Evidence-Based
Nursing, 5(3), 160–162.
Liberati, A., Altman, D. G., Tetzlaff, J., Mulrow, C., Gotzsche,
P. C., Ioannidis, J. P., et al. (2009). The PRISMA Statement for
reporting systematic reviews and meta-analyses of studies that
evaluate healthcare interventions: Explanation and elaboration.
Annals of Internal Medicine, 151(4), W-65-W-94.
Magnus, M. C., Ping, M., Shen, M. M., Bourgeois, J., & Magnus,
J. H. (2011). Effectiveness of mammography screening in reduc-
ing breast cancer mortality in women aged 39-49 years: A meta-
analysis. Journal of Women’s Health, 20(6), 845–852.
Mantzoukas, S. (2009). The research evidence published in high
impact nursing journals between 2000 and 2006: A quantitative
content analysis. International Journal of Nursing, 46(4), 479–
489.
Melnyk, B. M., & Fineout-Overholt, E. (2011). Evidence-based
practice in nursing & healthcare: A guide to best practice (2nd
ed.). Philadelphia, PA: Lippincott Williams & Wilkins.
Moher, D., Cook, D. J., Eastwook, S., Olkin, I., Rennie, D., &
Stroup, E. F. (1999). Improving the quality of reports of meta-
analyses of randomized controlled trials: The QUOROM state-
ment. The Lancet, 354, 1896–1900.
Moher, D., Liberati, A., Tetzlaff, J., Altman, D. G., & PRISMA
Group. (2009). Preferred Reporting Items for Systematic Reviews
and Meta-Analyses: The PRISMA Statement. Retrieved from
http://www.prisma-statement.org.
National Guideline Clearinghouse (NGC). (2012a). National
Guideline Clearinghouse: Guidelines by topics. Agency for
Healthcare Research and Quality. Retrieved from http://www
.guideline.gov/browse/by-topic.aspx.
National Guideline Clearinghouse (NGC). (2012b). National
Guideline Clearinghouse: Home. Agency for Healthcare Research
and Quality. Retrieved from http://www.guideline.gov/.
National Institutes of Health (NIH). (2012). NIH: National Center
for Translational Science. Bethesda, MD: Author. Retrieved from
http://www.ncrr.nih.gov/clinical_research_resources/clinical_
and_translational_science_awards/index.asp.
Noordzij, M., Hooft, L., Dekker, F. W., Zoccali, C., & Jager, K. J.
(2009). Systematic reviews and meta-analyses: When they are
useful and when to be careful. Kidney International, 76(11),
1130–1136.
Nurse Executive Center. (2005). Evidence-based nursing practice:
Instilling rigor into clinical practice. Washington, DC: The Advi-
sory Board Company.
Rew, L. (2011). The systematic review of literature: Synthesizing
evidence for practice. Journal for Specialists in Pediatric Nursing,
16(1), 64–69.
506 UNIT THREE  Putting It All Together for Evidence-Based Health Care
Sackett, D. L., Straus, S. E., Richardson, W. S., Rosenberg, W., &
Haynes, R. B. (2000). Evidence-based medicine: How to practice
& teach EBM (2nd ed.). London, UK: Churchill Livingstone.
Sandelowski, M., & Barroso, J. (2007). Handbook for synthesizing
qualitative research. New York, NY: Springer.
Shojania, K. G., Sampson, M., Ansari, M. T., Ji, J., Doucette, S., &
Moher, D. (2007). How quickly do systematic reviews go out of
date? Survival analysis. Annals of Internal Medicine, 147(4),
224–234.
Stetler, C. B. (1994). Refinement of the Stetler/Marram model for
application of research findings to practice. Nursing Outlook,
42(1), 15–25.
Stetler, C. B. (2001). Updating the Stetler Model of Research Uti-
lization to facilitate evidence-based practice. Nursing Outlook,
49(6), 272–279.
Stetler, C. B., & Marram, G. (1976). Evaluating research findings
for applicability in practice. Nursing Outlook, 24(9), 559–563.
The Joint Commission. (2012). About our standards. Retrieved
from http://www.jointcommission.org/standards_information/
standards.aspx.
Titler, M. G. (2004). Overview of the U.S. invitational conference
“Advancing Quality Care Through Translation Research.” World-
views on Evidence-Based Nursing, 1(1), S1–S5.
Titler, M. G., Kleiber, C., Steelman, V. J., Rakel, B. A., Budreau,
G., Everett, L. Q., et al. (1994). Research-based practice to
promote the quality of care. Nursing Research, 43(5), 307–313.
Titler, M. G., Kleiber, C., Steelman, V. J., Rakel, B. A., Budreau,
G., Everett, L. Q., et al. (2001). The Iowa Model of Evidence-
Based Practice to promote quality care. Critical Care Nursing
Clinics of North America, 13(4), 497–509.
Turlik, M. (2010). Evaluating the results of a systematic review/
meta-analysis. Podiatry Management. Retrieved from www
.podiatrym.com.
U.S. Department of Health and Human Services, National Institutes
of Health, National Heart, Lung, and Blood Institute. (2003).
Reference card from the Seventh report of the Joint National Com-
mittee on Prevention, Detection, Evaluation, and Treatment of
High Blood Pressure (JNC 7). Bethesda, MD: NIH Publication
No. 03-5231. Retrieved from www.nhlbi.nih.gov/guidelines/
hypertension/jnc7card.htm.
Walsh, D., & Downe, S. (2005). Meta-synthesis method for qualita-
tive research: A literature review. Journal of Advanced Nursing,
50(2), 204–211.
Whittemore, R., Melkus, G., Wagner, J., Dziura, J., Northrup, V., &
Grey, M. (2009). Translating the diabetes prevention program to
primary care: A pilot study. Nursing Research, 58(1), 2–12.
Wulff, K., Cummings, G. G., Marck, P., & Yurtseven, O. (2011).
Medication administration technologies and patient safety:
A mixed-method systematic review. Journal of Advanced Nursing,
67(10), 2080–2085.
UNIT FOUR
Analyzing Data, Determining Outcomes, and Disseminating Research
Collecting and Managing Data
http://evolve.elsevier.com/Grove/practice/
D
ata collection is one of the most exciting parts
of research. After all the planning, writing, and
negotiating, you should be eager and well pre-
pared for this active part of research. The passion that
comes from wanting to know the answer to your
research question brings a sense of excitement and
eagerness to start collecting your data. However,
before you leap into data collection, you need to spend
some time carefully planning this adventure and pilot
test each step. Planning data collection begins with
identifying all the data to be collected. The data to be
collected are determined by the research questions,
objectives, or hypotheses of the proposed study. As
you develop the data collection plan, be sure that you
gather all the data needed to answer the research ques-
tions, achieve the study objectives, or test the hypoth-
eses. Chapter 16 includes detailed information about
measurement, so the focus in this chapter is on the
logistical and pragmatic aspects of quantitative data
collection. Data collection strategies for qualitative
studies are described in Chapter 12.
To start planning the data collection process, you
need to determine the best mode by which the data can
be collected. Factors that influence the plan to collect
and enter data into a database for analysis include cost,
time, the availability of assistance, and the need for
consistency. The development of the data collection
plan is followed by developing data collection forms
and a codebook for data entry. Conducting a pilot test
with a small group of subjects is the next recommended
step. The pilot test may result in modifications of the
plan, and then the actual data collection can begin.
During data collection, various problems may arise.
Potential situations are described in this chapter along
with problem-solving strategies. The chapter concludes
with the discussion of data entry and management.
20
CHAPTER
Data Collection Modes
Data can be collected by interview (face-to-face
or telephone); observations; focus groups; self-
administered questionnaires (online or hard copy); or
extraction from existing documents such as patient
medical records, motor vehicle department accident
records, or state birth records (Figure 20-1). Many
factors need to be considered when a researcher is
deciding on the mode for collecting data. Harwood
and Hutchinson (2009) describe four factors that need
to be part of your decision-making process: (1) purpose
and complexity of the study, (2) availability of finan-
cial and physical resources, (3) characteristics of study
participants and how best to gain access to them from
the population, and (4) your skills and preferences as
a researcher.
Researcher-Administered or Participant-
Administered Instruments
If you need a subject’s accurate blood pressure or
height and weight, a self-report measure may be
neither valid nor reliable for the purpose of your study.
However, if the purpose of your study can be accom-
plished with a self-report survey method, you must
decide whether the format will be researcher-
administered or self-administered. It may be best for
the researcher to administer self-report paper-and-
pencil instruments if the potential subjects have
minimal language or literacy ability, whereas it may
be best to consider electronic data collection or
medical record extraction if the subjects are likely to
have hearing impairments, transportation problems, or
physical difficulties.
If the researcher is administering the survey, will it
be in person or by telephone? If self-administered, will
507
 508 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
Researcher
Administered
Healthcare Self-Report
Equipment Survey
Self-Report
Extract from
Existing Records Interview
and databases
Figure 20-1  Data collection modes.
the participant complete a pencil-and-paper copy or an
online electronic copy? Internet survey centers spe-
cialize in this mode of data collection and have expert
help or tutorials for assessing the best mode for your
study purpose. For example, in deciding on a tele-
phone survey, how many times will you try to reach a
potential subject before you give up, what days of the
week or hours of the day will you call and how might
that bias your sample or their responses, and how will
you accurately determine the response rate (Harwood,
2009)? If you decide on a mailed paper-and-pencil
survey, what will you do with undelivered or incom-
plete returns? Will you search for correct mailing
addresses and try again? Will you send a reminder if
the survey is not received within a particular time
frame, and, if so, what time frame will you give a
respondent, and how many reminders will you send
(Harwood, 2009)?
Electronic Data Collection
When you are using an existing instrument, you may
oasis-ebl|Rsalles|1476321157
need permission to convert the questions into an
online format, a special type of form that allows the
data to be scanned into a database, or into an applica-
tion for a phone or other electronic device. Each of
these modes of data collection may require special
hardware and software. Universities, schools of
nursing, and funded researchers are purchasing these
sometimes expensive products because the costs of
acquiring the hardware and software are considerably
less than the costs of entering data manually.
Scannable Forms
Other software allows the preparation of special data
collection forms that rely on optical character recogni-
tion (OCR), which requires exact placement on the
Participant
Administered
Web-Based
Survey Survey
Electronic
Diary
Device
page for each potential response. To maintain the
precise location of each response on print copies of
these instruments, careful attention must be given to
printing or copying these forms. The complete form is
scanned, and the answers (data) are automatically
recorded in a database. Additional features include
data accuracy verification, selective data extraction
and analysis, auditing and tracking, and flexible export
interfaces. Figure 20-2 shows the scannable version of
the Parents and Newborn Screening Survey developed
by Patricia Newcomb, PhD, RN, CPNP, and Barbara
True, MSN, CNS. Subjects completing the survey fill
in the circle that corresponds to the appropriate option
for each question.
Online Data Collection
Computer software packages developed by a variety
of companies (e.g., Zoomerang and SurveyMonkey)
enable researchers to provide an online copy of instru-
ments and other data collection forms. These types of
software programs have unique features that allow the
researcher to develop point-and-click automated forms
that can be distributed electronically. The following
questions need to be considered with use of these
programs. For an online survey, is it a secure site for
the purposes of confidentiality and anonymity? How
will you ensure that only eligible participants com-
plete the survey? Will potential subjects receive a per-
sonalized email from you with a link to a website?
How will you obtain the email addresses? Can you
offer help if the subjects have any questions about
your study?
Online services can be easy to use for both the
researcher and study participants but may be costly
and require specific assurances about confidentiality
Text continued on p. 513
 CHAPTER 20  Collecting and Managing Data 509

ID
1 0 6 1

Parents and Newborn Screening Survey

The University of Texas at Arlington College of Nursing
Andrews Women’s Hospital

Instructions: Please use a BLACK PEN for Please shade circles like this:
completing the survey. Do not use pencil. Not like this:

1. My age is:

Less than 18 18-24 25-30 31-35 36+

2. My race/ethnic group is:

African-American Asian Caucasian Hispanic Other

3. My highest level of education is:

Less than high school High school diploma Some college Bachelor’s degree Master’s degree or more

4. I work in the healthcare field:

Yes No

5. I started pregnancy care:

In the first 3 months of my pregnancy
In the second 3 months of my pregnancy
In the last 3 months of my pregnancy
I did not get medical care during my pregnancy

6. My care for this pregnancy and birth was paid for by:
Medicaid
Private insurance
I paid for it by myself
I don’t know how it will be paid for

7. I learned about newborn screening from (may circle more than one):
I never heard of newborn screening before this
My doctor or midwife
My child’s doctor or nurse practitioner
A book, video, or brochure
My hospital nurse
My doctor’s nurse
Internet
Friend or family member
Other

109
Parents and Newborn Screening Survey
08/22/2011 Please Turn Over
Page 1 0534001092

Figure 20-2  Scannable form: Parents and Newborn Screening Survey. (Developed by Patricia Newcomb, PhD, RN, CPNPN, and Barbara True, MSN,
CNS; Teleform designed by Denise Cauble and Whitney Mildren, Graduate Research Assistants and PhD students, College of Nursing, The University of
Texas at Arlington.) Continued
510 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research

ID

Instructions: Please use a BLACK PEN for Please shade circles like this:
completing the survey. Do not use pencil. Not like this:

8. I have other children.

Yes No
9. I learned about newborn screening before this baby was born.
Yes No
10. The language I speak most of the time is:
English Spanish Other__________

Please put a check in the box that shows how much you agree or disagree with the statement.

Strongly Agree Agree Not sure Disagree Strongly Disagree

11. I understand what I need to know about

newborn screening.

12. I know when my baby should have another

newborn screening.

13. If my baby has a disease that shows up on

newborn screening, serious problems can be
prevented if my baby gets treatment right away.

14. I will get the results of the newborn screening

tests by mail.

15. My baby’s doctor will get the results of the

newborn screening tests by mail.

16. Doing the newborn screening is worth the

discomfort the baby feels.

17. I know what genetic testing is.

18. I know where to take my baby for the second

newborn screening test.

19. Newborn screening can identify babies with

certain serious inherited diseases.

20. I understand what DNA does.

21. If my baby’s newborn test is abnormal, my

baby’s father might have something wrong with his
DNA.

109
Parents and Newborn Screening Survey
08/22/2011 Please Go To Next Page
Page 2 8572001092

Figure 20-2, cont’d

 CHAPTER 20  Collecting and Managing Data 511

ID

Instructions: Please use a BLACK PEN for

completing the survey. Do not use pencil.

Strongly Agree Agree Not sure Disagree Strongly Disagree

22. All babies should have genetic testing when

they are born.

23. If my baby’s newborn test is abnormal, I might

have something wrong in my DNA.

24. Newborn screening will test for some, but not

all, serious diseases that run in families.

25. Babies with serious disorders may look healthy

when they are born.

26. I am scared that the newborn test might find

something wrong with my baby.

27. I know what newborn screening bloodspots

are.

28. The state of Texas will keep my baby’s

bloodspots unless I mail them a form telling them
not to.

29. I wish I had more information about newborn

screening.

30. Some of the tests in the newborn screening are

genetic tests.

31. There is DNA in my baby’s blood spots.

32. It would be OK to use my baby’s bloodspots for

research to find treatments for serious diseases.

33. It is OK for the state to keep my baby’s

bloodspots for research without getting special
permission from me.

34. It would be OK to keep my baby’s DNA for

future study if my baby’s name or other private
information is not connected to the DNA sample.

109
Parents and Newborn Screening Survey
08/22/2011 Please Turnover
Page 3 6735001091

Figure 20-2, cont’d Continued

512 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research

ID

Instructions: Please use a BLACK PEN for Please shade circles like this:
completing the survey. Do not use pencil. Not like this:

35.It would be OK with me for the state to share

my baby’s bloodspots with researchers all over the
nation.

36. It would be OK to let the state share my baby’s

bloodspots with researchers if an ethics committee
reviews the research first.

37. If future research is done on my baby’s

bloodspots I want to know.

38. I do not want the state or any other group to

keep the bloodspots from my baby’s testing for any
reason.

39. I want my baby to have newborn screening.

40. My baby will have another newborn screening,

but I do not really want it.

This is the End. Thank you!

Please share any comments that would help us improve the newborn screening process:

If you would like to learn more about newborn screening, please pick up some of our
information about newborn screening at the desk or from your nurse when you leave today.

109
Parents and Newborn Screening Survey
08/22/2011
Page 4 5991001094

Figure 20-2, cont’d


of data and anonymity of subjects. The National
Institutes of Health (NIH) supports a secure Internet
environment for building online data surveys and
data management packages (Harris et al., 2009). This
service, developed by experts at Vanderbilt University,
is called REDCap (Research Electronic Data Capture)
and may be available at your university research site
(http://project-redcap.org/).
Im et al. (2007) conducted a survey in the United
States of gender and ethnic differences in the experi-
ence of cancer pain. These researchers administered
their questionnaire over the Internet and through a
paper-and-pencil format based on subject preference.
The following excerpt describes the data collection
procedure for their study:
“To administer the Internet questionnaire, a Web site
conforming to the Health Insurance Portability and
Accountability Act standards, the System Administra-
tion, Networking, and Security Institute Federal
Bureaus of Investigation recommendations, and the
Institutional Review Board [IRB] policy of the institu-
tion where the researchers were affiliated was devel-
oped and published on an independent, dedicated
Web site server. When potential participants visited
the project Web site, informed consent was obtained
by asking them to click a button labeled I agree to
participate. After this, questions on specific diagnoses,
cancer therapies, and medications were asked, and
the appropriateness of answers was checked auto-
matically through a server-side program; participants
were connected automatically to the Internet survey
web page if the answers were appropriate.
“Upon request, pen-and-pencil questionnaires
were provided by mail to the community consultants,
who distributed the questionnaires in person only to
those who were identified as cancer patients. These
oasis-ebl|Rsalles|1476461039
questionnaires accompanied hard copies of the same
informed consent form included in the Internet format
of the questionnaire, and the pen-and-pencil ques-
tionnaire included a sentence ‘Filling out this question-
naire means that you are aged over 18 years old and
giving your consent to participate in this survey.’ After
the self-administered questionnaires were completed,
community consultants retrieved all except five (these
were mailed directly to the research team by the
participants) in person at the community settings and
mailed them to the research team. Supplementing
pen-and-pencil questionnaires was essential to recruit
the target number of ethnic minority cancer patients
across the nation who did not have access to the
CHAPTER 20  Collecting and Managing Data 513
Internet but were interested in participating in the
study. Among the 276 participants who were recruited
through community settings, 246 … used the pen-
and-pencil questionnaires. … There were no statisti-
cally significant differences in psychometric properties
between the Internet format and the pen-and-pencil
format of the questionnaire. … It took an average of
30-40 minutes for the participants to complete either
the Internet format or the pen-and-pencil format of
the questionnaire.” (Im et al., 2007, pp. 299-300)
Im et al. (2007) maximized their sample size and
obtained a more representative sample by giving par-
ticipants an option to complete their questionnaire on
the Internet or using paper-and-pencil format. The
researchers took steps to ensure that the data collected
by the two formats were comparable by testing for
significant differences and finding none. The time to
complete the Internet and paper-and-pencil question-
naires did not vary. Im et al. (2007) also ensured that
an ethical study was conducted and subjects’ rights
were protected.
The additional advantage of Internet data collection
is that responses can be time/date stamped. For
example, if subjects are instructed to complete the
questionnaire before bedtime, the time can be verified.
If subjects are instructed to complete a daily diary, date
of entry would be documented, and subjects would be
discouraged from entering all diary days on the last
day just before returning the diary to the researcher
(Fukuoka, Kamitani, Dracup, & Jong, 2011).
Computer-Based Data Collection
With the advent of laptop and tablet computers, data
collectors can code data directly into an electronic file
at the data collection site. If a computer is used for
data collection, a program must be written for enter-
ing, cleaning, and storing data. A computer enables
users to collect large amounts of data with few errors
that can be readily analyzed with a variety of statisti-
cal software packages. In addition to researchers
using technology at the point of data collection to
record data, technology has made it possible to inter-
face physiological monitoring systems with comput-
ers for data collection. An advantage of using
computers for the acquisition and storage of physio-
logical data is the increased accuracy and precision
that can be achieved by reducing errors associated
with manually recording or transcribing physiological
data from a monitor. Another advantage is that more
data points can be recorded electronically than could
514 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
be recorded manually. Computers linked to physio-
logical monitoring systems can store multiple data for
multiple indicators, such as blood pressures, oxygen
saturation levels, and sleep stages. Because data can
be electronically recorded, data collection is less
labor intensive, and the data are ready to analyze
more quickly. The initial cost of equipment may be
high, but it is reasonable when the cost of hiring and
training human data collectors is considered.
There are some concerns with the use of computer-
ized data acquisition systems, but physiological data
are usually best gathered and stored directly into a
computer database to ensure accurate, complete data
collection. Physiological data typically require large
computer storage space. The computer-equipment
interface may require more space in an already
crowded clinical setting; when possible, existing
equipment should be used to collect data. Purchasing
the equipment, setting it up, and installing the software
can be time-consuming and expensive at the start of
your project. Thus, initial studies usually require sub-
stantial funding. Another concern is that the nurse
researcher may focus on the machine and technology
and neglect observing and interacting with the subject.
The most serious disadvantage of computerized
data collection is the possibility of measurement error
that can occur with equipment malfunctions and soft-
ware errors. Regular maintenance and calibrations, or
reliability checks of the equipment and software,
reduce this problem. The benefits of collecting repeated
measures over time may outweigh the risk of missing
data because of poor compliance. For example, collect-
ing continuous rectal temperature data from a subject
is easier and less burdensome than asking the subject
to measure an oral temperature every 1 to 2 hours.
Savian, Paratz, and Davies (2006) conducted a
single-blind randomized, crossover study with 14
mechanically ventilated intensive care unit patients.
“[The purpose of the study was to determine the
effectiveness of] manual hyperinflation (MHI) and ven-
tilator hyperinflation (VHI) on respiratory mechanics
(static compliance [Cst]), oxygenation (arterial oxygen
tension [PaO2]/fraction of inspired oxygen [FIO2]
ratio), and secretion removal (wet weight of sputum
and peak expiratory flow rate [PEFR]) at different
levels of PEEP [positive end-expiratory pressure] …
a secondary aim was to investigate the hemodynamics
heart rate [HR], mean arterial pressure [MAP] and
metabolic response (carbon dioxide output [VCO2])
during MHI and VHI.” (Savian et al., 2006, p. 335)
The computerized systems used to collect and
record data in the study by Savian et al. (2006) are
detailed in the following excerpt:
“PEFR and CO2 [carbon dioxide] production were
measured using a flow and CO2 sensor connected to
the patient’s airways and to the CO2SMO [carbon
dioxide] respiratory mechanics monitor (CO2SMO
Plus Model 8000, Novametrix Medical Systems Inc.,
Wallingford, CT). All information from the CO2SMO
monitor was simultaneously recorded in the Analysis
Plus computer program.
“Static lung compliance was recorded by the static
measures function device on the Bennett 7200 venti-
lator where a plateau pressure was obtained by
including an inspiratory pause of 2 seconds into the
mandatory breath. …
“PaO2/FIO2 ratio was calculated from arterial blood
samples taken immediately before and immediately
after MHI and VHI. Four milliliters of arterial blood
were drawn into a syringe containing heparin and
analyzed by a blood gas machine (Bayer Australian
Limited 865, Pymble, NSW, CAN 000128 714). This
procedure was standardized across subjects.
“HR and MAP were read directly from the moni-
toring system (Merlin pressure module M1006A
Hewlett Packard, Palo Alto, CA) and recorded every
minute before, during, and for 5 minutes after MHI
and VHI.” (Savian et al., 2006, p. 336)
The use of computerized data collection by Savian
et al. (2006) enabled them to collect repeated mea-
sures on several physiological variables in an accurate
and precise way. The data were collected by sensors
and stored in the computer to reduce error and facili-
tate data analysis.
Phones and Other Electronic Devices
Software applications for mobile phones have evolved
from personal digital assistants (PDAs) that allow the
researcher to collect and download data directly into
the computer from observations as they occur. Health-
care providers load applications that facilitate accurate
assessment, diagnosis, and pharmacological and non-
pharmacological management of patients. PDAs are
also used to store deidentified data from office com-
puters in a form that is easily transportable. PDA soft-
ware is currently available that may help nurse
practitioners collect data for research. Multiple nurse
practitioners involved in a research project could

forward data electronically from PDAs to a central
research site for analysis. Encrypted electronic devices
are needed to protect the confidentiality of data during
transmission. These electronic devices can be mis-
placed or stolen, threatening confidentiality. Research-
ers need to protect the data with a security code to
ensure that no one but themselves can access data in
these formats.
Mobile phones and computers are becoming more
similar with the increased sophistication of applica-
tions for mobile phones. Some of these applications
can be used to collect various data. Other electronic
devices include pill containers that record when pills
are accessed and watches with timers to remind
participants to take certain health-related actions.
However, the use of these devices for research may
require considerable preparation. You may need to
hire programmers with the needed expertise, and you
may need to purchase, rent, or borrow the needed
number of devices or monitors.
Factors Influencing
Data Collection
When planning data collection, cost, time, the avail-
ability of assistance, and the need for consistency are
critical factors to consider. The researcher balances
these factors with the need to maintain the reliability
and validity of the study in the development of the
data collection plan.
Cost Factors
Cost is a major consideration when planning a study.
Measurement tools, such as continuous electrocardio-
gram monitors (Holter monitor), wrist activity moni-
tors (accelerometers), spirometers, pulse oximeters, or
glucometers, used in physiological studies may need
to be rented, purchased, or loaned from the manufac-
turer or other company. You may need to pay a fee to
use instruments or questionnaires. Some instruments
and questionnaires are available only if a copy is pur-
chased for each participant. Data collection forms may
need to be formatted or developed for electronic use.
In some cases, printing costs for materials such as
teaching materials or questionnaires that will be used
during the study must be considered. Providing the
required copy of the signed consent form doubles the
expense of consent forms. Small payments to partici-
pants in the form of cash or gift cards should be con-
sidered as compensation for a subject’s time and effort
in providing the data. Sometimes childcare may need
to be provided for parents and other caregivers who
CHAPTER 20  Collecting and Managing Data 515
would not otherwise be able to participate in your
study. In some studies, postage is an additional
expense. There may be costs involved in coding the
data for entry into the computer and for conducting
data analyses. Consultation with a statistician early in
the development of a research project and during data
analysis must also be budgeted. You may need to hire
someone who can remain blinded for data entry or
analysis or someone who can type the final report,
develop graphics or presentations, or type and edit
manuscripts for publication.
In addition to the above-described direct costs of a
research project, there are costs associated with the
researcher’s time and travel to and from the study
site. You also must estimate the expense of presenting
the research findings at conferences and include
those expenses in the budget. To prevent unexpected
expenses from delaying the study, examine all costs in
an organized manner. A budget is best developed early
in the planning process and revised as plans are modi-
fied. Seeking funding for at least part of the study costs
can facilitate the conduct of a study.
Time Factors
Researchers often underestimate the time required for
participants to complete data collection forms and for
the research team to recruit and enroll subjects for a
study. The first aspect of time—the participant’s time
commitment—must be determined early in the process
because the time needed for participant involvement
must be included in the informed consent process and
document. While conducting your pilot study, make
note of the time required to collect data from a subject.
You may need to revise your timeline and consent
form to reflect the expected time commitment
accurately.
The second aspect of time—the time needed to
complete data collection—is especially challenging
to predict because events during the data collection
period sometimes are not under the researcher’s
control. For example, a sudden heavy staff workload
may make data collection temporarily difficult or
impossible, or the number of potential subjects might
be reduced for a period. In some situations, research-
ers must obtain permission from each subject’s physi-
cian before they are permitted to collect data on that
subject. Activities required for this stipulation, such
as contacting physicians, explaining the study, and
obtaining permission, require extensive time. In some
cases, potential subjects are lost before the researcher
can obtain the mandatory permission, extending the
time required to obtain the necessary number of
subjects.
516 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
How long will it take to identify potential subjects,
explain the study, and obtain consent? How much
time will be needed for activities such as completing
questionnaires or obtaining physiological measures?
Novice researchers have difficulty making reasonable
estimates of time and costs related to a study. Vali­
dating the time and cost estimates with an experi-
enced researcher can be very informative. Experienced
researchers know the challenges of data collection
and have learned that data collection may take two to
three times longer than predicted. If the cost and time
factors are prohibitive, you may need to simplify your
study so that fewer variables are measured, fewer
instruments are used, or fewer subjects are needed.
Make the design less complex, and use fewer data
collectors. A blinded intervention study involves more
research staff and is generally not feasible for a novice
researcher. These are serious modifications, however,
with implications for the validity of the findings, so
you and your team should thoroughly examine the
consequences before making such revisions. If pre-
liminary time or cost estimates go beyond expecta-
tions, you can revise the time schedules and budget
with new projections for completing the study.
Consistency
Consistency in data collection across subjects is criti-
cal. What time of year will data be collected? For
example, if you collect data during holiday seasons,
data about sleeping, eating, or exercising may vary.
Pediatric patients with asthma may experience more
symptoms during the winter months than during
summer. Planning data collection for a study of
symptom management with this population would
need to take this possibility into consideration.
The specific days and hours of data collection may
influence the consistency of the data collected and
must be carefully considered. For example, the energy
level and state of mind of subjects from whom data
are gathered in the morning may differ from that of
subjects from whom data are gathered in the evening.
With hospitalized study participants, visitors are more
likely to be present at certain times of day and may
interfere with data collection or influence participant
responses. Patient care routines vary with the time of
day. In some studies, the care recently received or the
care currently being provided may alter the data you
gather. The subjects you approach on Saturday to par-
ticipate in the study may differ from the subjects you
approach on weekday mornings. Subjects seeking care
on Saturday may have a full-time job, whereas sub-
jects seeking care on weekday mornings may be either
unemployed or too ill to work.
Who will collect the data? If you decide to use data
collectors, they must be trained in responsible conduct
of research and issues of informed consent, ethics, and
confidentiality and anonymity (see Chapter 9). They
must be informed about the research project, familiar
with the instruments to be used, and have equivalent
training in the data collection process. In addition to
training, data collectors need written guidelines or
protocols that indicate which instruments to use, the
order in which to introduce the instruments, how to
administer the instruments, and a time frame for the
data collection process (Harwood, 2009; Kang, Davis,
Habermann, Rice, & Broome, 2005).
If more than one person is collecting the data, con-
sistency among data collectors (interrater reliability)
must be ensured through testing (see Chapter 16). The
training needs to continue until interrater reliability
estimates are at least 85% to 90% agreement between
the expert and the trainee or trainees. Waltz, Strick-
land, and Lenz (2010) suggest that a minimum of
10% of the data needs to be compared across raters
before interrater reliability can be adequately reported.
The trained data collector’s interrater reliability with
the expert trainer should be assessed intermittently
throughout data collection to ensure consistency from
the first to the last participant in the study. Data col-
lectors also must be encouraged to identify and record
any problems or variations in the environment that
affect the data collection process. The description of
the training of the data collectors is usually reported
in the methods section of an article so that others
can assess the data collection process (Harwood &
Hutchinson, 2009).
Availability of Assistance
Who is going to help you with the study? If you are a
student, will your mentor or supervising faculty
member participate? Does your mentor or supervising
faculty member have research assistants who could
assist in your study? Will nurses, physicians, and other
health professionals assist with recruitment? Do they
have time to do this? Are they willing to help?
Will the researcher collect all the data, or will data
collectors be employed for this purpose? Can data
collectors be nurses working in the area? Data collec-
tion may be delayed when nurses providing patient
care are also expected to be data collectors. Even when
a nurse agrees to help you with subject recruitment or
data collection, patient care takes priority over data
collection and increases the risk for missing data or
missing the opportunity to enroll eligible subjects.
If clinicians are going to recruit subjects or collect
data, the clinicians need to complete training for

protection of human subjects during research. An
IRB requires documentation of this training for each
person involved in recruitment and data collection. If
you are going to be doing all the data collection
yourself, will you be available every day of the
week? What hours will you be available? If others
will be involved in collecting data, allow time for
training on data collection procedures. You need to
be available by telephone or other means for ques-
tions and emergencies when others are collecting
data for your study. Keeping these factors in mind,
you are now ready to plan the data collection process
for your study.
Data Collection and Coding Plan
The factors of cost, time, availability of assistance,
and need for consistency shape the data collection plan
that you develop. A data collection plan details how
you will implement your study. The plan for collecting
data is specific to the study being conducted and
requires that you consider some common elements of
research. You need to map out procedures you will use
to collect data, anticipate the time and cost of data
collection, develop data collection forms that ease data
entry, and prepare a codebook that will help you to
code the variables to be entered in a database. This
extensive planning increases the accuracy of the data
collected and the validity of the study findings. The
validity and strength of the findings from several
carefully planned studies increase the quality of
the research evidence that is then available for imple-
menting into clinical practice (Melnyk & Fineout-
Overholt, 2010).
Identifying data include variables such as patient
record number, home address, and date of birth (see
Chapter 9). Avoid collecting these data unless they are
essential to answer the research question. For example,
collect a patient’s age instead of date of birth. Review
regulations by the Health Insurance Portability and
Accountability Act about the participant’s private
health information (www.hhs.gov/ocr/hipaa).
The methodology of a study may include contacting
subjects later for additional data collection. In this
case, you will need to obtain the subject’s address and
telephone number and protect the information appro-
priately. Names and phone numbers of family members
or friends may also be useful if subjects are likely to
move or may be difficult to contact. This information
can be obtained only with subjects’ permission as part
of their informed consent. Consider the importance of
each piece of data and the subject’s time required to
collect it. If the data can be obtained from patient
CHAPTER 20  Collecting and Managing Data 517
records or any other written sources, you do not need
to ask the subject to provide this information. To
collect data from a patient’s records, make sure to
include permission to do this in the consent form, and
ensure that the IRB has authorized your team to do this.
Data Collection Forms
Before data collection begins, you may need to develop
or modify forms on which to record data. These forms
can be used to record demographic data, information
from the patient record, observations, or values from
physiological measures. The demographic variables
commonly collected in nursing studies include age,
gender, race, education, income, employment status,
diagnosis, and marital status. You may want to collect
additional demographic data if researchers have iden-
tified participant characteristics that affect the study
variables. You also might need to collect other data
that may be extraneous or confounding variables, such
as the subject’s physician, stage of illness, length of
illness or hospitalization, complications, date of data
collection, time of day and day of week of data col-
lection, and any untoward events that occur during
the data collection period. If there are only women in
your sample, the subject’s age and reproductive
status, parity, and number of children in the home may
be confounding variables. In a study of patients with
ventilator-associated pneumonia, the researcher needs
to record the length of time between when the patient
was intubated and when ventilator-associated pneu-
monia was diagnosed. The researcher for this study
also needs to record whether the patient had a preexist-
ing pulmonary disease.
Data collection forms must be designed so that the
data are easily recorded, coded, and entered into the
computer. You need to decide whether data will be
collected in raw form or coded at the time of collec-
tion. Coding in quantitative studies is the process of
transforming data into numerical symbols that can be
entered easily into the computer. For example, vari-
ables such as race, gender, ethnicity, and diagnoses
can be categorized and given numerical labels. For
gender, the male category could be identified by a “1”
and the female category by a “2.” You may also want
to include an “other” category (coded “3”) for partici-
pants who are transgendered or transsexual. To be able
to compare your sample with samples in federally
funded studies, you may need to separate the questions
about ethnicity and race. In 2003, the Office of Man-
agement and Budget of the U.S. government directed
researchers and others collecting data for federal pur-
poses or at federal expense to separate the questions
of race and ethnicity (Office of Minority Health,
 518 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
2010). At the same time, the Office of Management
and Budget specified the categories for each. The fol-
lowing questions are correct according to these federal
guidelines. How would a subject who is biracial or
multiracial complete the form? You may want to word
the question to ask the participant’s primary race or
allow multiple responses.
Ethnicity
(1)  Hispanic or Latino
(2)  Non-Hispanic or Latino
Race
(1)  American Indian or Alaskan Native
(2)  Asian
(3)  Black or African American
(4)  Native Hawaiian or Other Pacific Islander
(5)  White
The coding categories developed for a study not
only must be mutually exclusive but also exhaustive,
which means that the value for a specific variable fits
into only one category, and each observation must fit
into a category. For example, a subject is highly unlikely
to want to reveal his or her exact income but would be
more willing to indicate that the income is in a particu-
lar range. The income ranges would not be mutually
exclusive or exhaustive if they were categorized in the
following way on a demographic questionnaire:
Income range (please check the range that most accu-
rately reflects your income)
___ (1)  $30,000 to $40,000
___ (2)  $40,000 to $50,000
___ (3)  $50,000 to $59,000
___ (4)  $60,000 to $70,000
___ (5)  $70,000 or more
These categories are not exclusive because they
overlap, and a subject with a $40,000 income could
mark category 1 or category 2 or both. The categories
are not exhaustive because a subject may have an
income of either $25,000 or $59,500, yet the ques­
oasis-ebl|Rsalles|1476461046
tionnaire does not contain categories that include
each of these incomes. How much detail do you
need on income? Do you want to know if the par­
ticipant’s household income is below poverty level?
To determine poverty level, you must know not
only the household income but also how many people
live in the household and compare this information
with federal poverty guidelines (http://aspe.hhs.gov/
poverty/09poverty.shtml).
The following income ranges are both exclusive
and exhaustive and would be appropriate for collect-
ing demographic data from subjects:
Income range (please check the range that most accu-
rately reflects your family’s income for a year,
before taxes)
___ (1)  Less than $30,000
___ (2)  $30,000 to $49,999
___ (3)  $50,000 to $69,999
___ (4)  $70,000 or greater
Data collection forms offer many response styles.
The person completing the form (subject or data col-
lector) might be asked to check a blank space before
or after the words “male,” “female,” or “other” or to
circle one of the words. If code numbers or variable
name codes are used, the meaning of the codes should
be clearly indicated on the collection forms so that the
individual completing the form understands them and
is not confused or misled by the code. Developing a
codebook for your data collection forms and data entry
is discussed later.
Placement of the data on the forms is important
because careful placement makes it easier for subjects
to complete the form without missing an item and for
data entry staff to locate responses for computer entry.
Placement of blanks on the left side of the page seems
to be most efficient for data entry, but this layout may
prove problematic when subjects are completing the
forms. The least effective arrangement is when the
data are positioned irregularly on the form because
the risk of data being missed during data entry is high.
Subjects’ names should not be on the data collection
forms; only the subject’s identification number should
appear. The researcher may keep a master list of sub-
jects and their code numbers, which is stored in a
separate location and either encrypted in an electronic
file or data repository or locked in a file drawer to
ensure the subjects’ privacy. Often this master list of
subjects and codes is kept with the subjects’ consent
forms in a locked file drawer. This master list is
required if collecting data later or recontacting the
subject is a necessary component of your study.
You should always organize your data collection
forms and instruments to begin with less personal
types of questions about age and education before
delving into more personal questions about contracep-
tives or feelings and attitudes. Also, you would not
want to save your most important items for the last
page of the questionnaire and risk missing data if a
participant becomes too fatigued or bored to finish the
questions. Different types of questions require more
or less time to complete, a factor that needs to be
considered. Also, questions may ask for a response
related to different time frames. For example, if one
questionnaire asks about the past week and two other
questionnaires ask about the past month, these should
be organized so that the subject is not confused by
going back and forth between time frames. If you have
several instruments or forms, you may want to put
 CHAPTER 20  Collecting and Managing Data 519

them together in a booklet to minimize the likelihood DATA COLLECTION FORM

that a questionnaire or form will be missed. Demographics
Figure 20-3 provides a sample data collection form. ____Subject Identification Number
It includes four items that could be problematic for ____Age
coding, data analysis, or both. The blank used to enter ____Gender
1. Male
“Surgical Procedure Performed” would lead to prob-
2. Female
lems when it is time to enter the data into a computer- ____Weight (in pounds)
ized data set. Because multiple surgical procedures ____Height (in inches)
could have been performed, developing codes for the __________Surgical Procedure Performed
various surgical procedures would be difficult and __/__/__Surgery Date (Month/Day/Year)
time-consuming. In addition, different words might be __/__/__Surgery Time (Hour/Minute/AM or PM)
used to record the same surgical procedure. It may be Narcotics Ordered After Surgery___________________
necessary to tally the surgical procedures manually. ___________________

Unless this degree of specification of procedures is Narcotic Administration

important to the study, an alternative would be to Date Time Narcotic Dose
develop larger categories of procedures before data 1.
collection and place the categories on the data collec- 2.
tion form. A category of “Other” might be useful for 3.
4.
less commonly performed surgical procedures. This
5.
method would require the data collector to make a
Instruction on Use of Pain Scale
judgment regarding which category was appropriate __/__/__Date (Month/Day/Year)
for a particular surgical procedure. Another option __/__/__Time (Hour/Minute/AM or PM)
would be to write in the category code number for a Comments:
particular surgical procedure after the data collection
form is completed but before data entry. If the specific __Treatment Group
surgical procedure is important to your study, you may 1. TENS
want to record the code the facility uses to bill for the 2. Placebo-TENS
procedure. Similar problems occur with the items 3. No-Treatment Control
“Narcotics Ordered after Surgery” and “Narcotic
Treatment Implemented
Administration.” Unless these data are to be used in __/__/__Date (Month/Day/Year)
statistical analyses, it might be better to categorize this __/__/__Time (Hour/Minute/AM or PM)
information manually for descriptive purposes. If Comments:
these items are needed for planned statistical proce-
dures, use care to develop appropriate coding. You
Dressing Change
may need detailed information if you want to know __/__/__Date (Month/Day/Year)
the appropriateness of the narcotic doses given. The __/__/__Time (Hour/Minute/AM or PM)
researcher might be interested in determining differ- _____Hours since surgery
ences in the amount of narcotics administered in a Comments:
given period in relation to weight and height. For
blinded studies, you do not want to record the treat- Measurement of Pain
ment group assignments on the data collection form. _____Score on Visual Analogue Pain Scale
Placing the treatment group code on the data collec- __/__/__Date Pain Measured
(Month/Day/Year)
tion form would be problematic because the informa-
__/__/__Time Pain Measured
tion is no longer blinded and could influence the data (Hour/Minute/AM or PM)
recorded by the data collectors. _____Hours since surgery
Comments:
Data Collection Detailed Plan
To ensure consistency in the data collection process, _____Data Collector Code
you need to develop a detailed plan. Envision the Comments:
overall activities that will be occurring during data
collection. Write each step and develop the forms,
training, and equipment needed for that step. Focus on
who, what, when, where, why, and how. How will you Figure 20-3  Data collection form.
520 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
recruit subjects? At what point is a subject assigned to
a group? It is optimal to assign subjects randomly to
an intervention group or control group after baseline
data are collected but before introducing the interven-
tion. In this way, all subjects demonstrate the ability
to complete the questions and measures and have the
opportunity to decline further participation before
group assignment. When and how will you implement
the intervention? Will data be collected from more than
one subject at a time, or is it necessary to focus atten-
tion on one subject at a time? How much time is needed
to collect data from each subject? The length of time
per subject is determined by study design, setting, and
available space. In addition, if you plan for three sub-
jects in the morning and three in the afternoon, what
are the contingencies for subjects who arrive late or
need additional time? Some subjects may be available
only during their lunch break or in the evening.
You might develop a data collection flow diagram
to illustrate the process for collecting data in your
study. An example is shown in Figure 20-4. The Con-
solidated Standards of Reporting Trials (CONSORT)
guidelines for reporting randomized trials in publica-
tions (Bennett, 2006) would also be very useful for
depicting the design and flow of your data collection
process and expected enrollment numbers, expected
attrition rates, and final sample sizes for each group
(see Chapter 19).
Decision Points
Decision points that occur during data collection must
be identified, and all options must be considered. One
decision may pertain to whether too few potential
subjects are meeting the sampling inclusion criteria.
Will you review study progress every week or every
month? If too few subjects from your potential pool
are eligible, at what point will you consider changing
exclusion criteria? For example, if you are recruiting
only first-time mothers older than 30 years of age, and
you are turning away willing participants because they
are too young, you and your research team need to
reconsider the rationale for that criterion and perhaps
decide either to lower the age range or to seek different
recruitment sites.
Other decisions include whether the subject under-
stands the information needed to give informed
consent, whether the subject comprehends instructions
related to providing data, and whether the subject has
provided all the data needed. As you look through the
completed data forms, are all responses completed? If
the subject skips a page, you will need to return that
page to the subject for completion. If the question
about income is not completed, how will you handle
that missing response? Your data collection flow chart
should indicate how much missing data will be allowed
per subject. At what point will you decide to exclude
a participant from your study?
Developing a Codebook for Data Definitions
We advise that you develop your codebook before
initiating data collection or during the pilot study. A
codebook identifies and defines each variable in your
study and includes an abbreviated variable name
(income), a descriptive variable label (gross household
annual income), and the range of possible numerical
values for every variable entered in a computer file (0
= none; 1 = <$30,000; 6 = >$100,000). Some code-
books also identify the source of each datum, linking
your codebook with your data collection forms and
scales. The codebook keeps you in control and pro-
vides a safety net for when you access the data later.
Some computer programs, such as SPSS for Windows,
allow you to print out your data definitions after
setting up a database. Figure 20-5 is an example of
data definitions from SPSS for Windows. Another
example of coding is presented in Figure 20-6.
Developing a logical method of abbreviating vari-
able names can be challenging. For example, you
might use a quality-of-life (QOL) questionnaire in
your study. It will be necessary for you to develop an
abbreviated variable name for each item in the ques-
tionnaire. For example, the fourth item on a QOL
questionnaire might be given the abbreviated variable
name QOL4. A question asking the last time a home
health nurse visited might be abbreviated HHN Lst-
visit. Although abbreviated variable names usually
seem logical at the time the name is created, it is easy
to confuse or forget these names unless they are clearly
documented with a variable label.
During the piloting phase of your research with the
first few pilot subjects, you can easily refine your vari-
able names and labels and request your research team
or a statistician to review the variables. This practice
encourages you to identify places in your forms that
might prove to be a problem during data entry because
of lack of clarity. Also, you may find that a single
question contains not one but five variables. For
example, an item might ask whether the subject
received support from her or his mother, father, sister,
brother, or other relatives and ask the subject to circle
the number that represents those who provided support.
You might think that you could code mother as “1,”
father as “2,” sister as “3,” brother as “4,” and other
as “5.” However, because the individual can circle
more than one, each relative must be coded separately.
Thus, mother is one variable and would be a
 CHAPTER 20  Collecting and Managing Data 521

ENROLLMENT AND SURVEY ADMINISTRATION PROCEDURES

Begin patient enrollment

Explain surveys
to patients

No
Does the patient
agree to participate?
Yes
Create patient file
Record reason for
patient refusal
Stop

Begin survey Assign unique

administration ID number

Does the patient need Yes Is a family member No Is staff member

assistance to complete or friend available
available to assist?
the surveys? to help?
No Yes Yes No
Give patient the survey Give patient and helper Assist Record reason
packet and pencil the survey packet patient for nonenrollment
and pencil

Stop
Surveys completed

Check data

Are all questions answered?

Yes No
Were any questions Yes Ask patient for answers
answered twice? or clarification
No
Enter data Adjust answers based
on patient clarification

Data quality check

and reports

Create point-of-service Schedule follow-up Send data to project-

reports visit survey management site for
aggregated comparison reports

Figure 20-4  Data collection flow chart.

 Q1 Value
Standard Position 2
Attributes Label I was motivated to migrate from my
country because my pay was too low.
Type Numeric
Format F8
Measurement Ordinal
Role Input
Valid Values 1 Strongly Disagree
2 Disagree
3 Neutral
4 Agree
5 Strongly Agree
Missing Values System

Q 39 Value
Standard Position 40
Attributes Label What is your gender?
Type Numeric
Figure 20-5  Example of data definitions from SPSS for Format F8
Windows. (Source: Nurse International Relocation Measurement Nominal
Questionnaire 2 [Gray & Johnson, 2009].)
Role Input
Valid Values 1 Male
2 Female
Missing Values System

Q50 Value
Standard Position 54
Attributes Label Which of the following best describes
your current employment situation?
Type Numeric
Format F8
Measurement Nominal
Role Input
Valid Values 1 Employed and working full time
2 Employed and working part time
3 Employed, currently on leave
4 Self-Employed
5 Unemployed
6 Other
Missing Values System

Variable Variable Source Value Levels Valid Missing Comments

Name Label Range Data

A1 to A5 Family Apgar Q2Family Apgar 1=never 1–5 9 Code as is (CAI)

2=hardly ever
3=some of the time
4=almost always
5=always

MF3 Mother's feeling, Tuesday diary, 1=poor 1–6 9 Code 1 to 6

Day 3 mother 6=good left to right

Figure 20-6  Example of coding.


dichotomized value, coded “1” if circled and “0” if not
circled. The father would be coded similarly as a
second dichotomous variable, and so on. Identifying
these items before data collection may allow you to
restructure the item on the questionnaire or data col-
lection form to simplify computer entry.
Give the codebook with its data definitions to the
individual or individuals who will enter your data into
the computer before initiating data collection. Deci-
sion rules for data entry should also be finalized. For
example, if a subject selects two responses for a single
item, will the variable be coded as missing, or does
convention dictate that the lowest or highest value
should be entered? Individuals who are entering data
need to have clarity on the distinctions between coding
a missing value or nonapplicable value as blank rather
than entering a “0” value.
In addition, provide the following information to
the person entering the data:
• Dates for data collection initiation and completion
• Estimated number of subjects to be included in the
study and how often batches of subjects’ data will
be entered
• Plan for documenting refusal rate, sample size, and
attrition during the study
• Copies of all scales, questionnaires, and data col-
lection forms to be used in the study
• Location of every variable on scales, question-
naires, or data collection forms
• Statistical package to be used for analysis of the
data
• Statistical analyses to be conducted to describe the
sample and to address the research purpose and the
objectives, questions, or hypotheses
• Contact information for the statistician or project
director who is available to consult about data entry
questions or data analysis
• Computer directory location of the database in
which the data will be entered and copied for
backup
• Timeline for receiving the data—for example,
whether you will deliver the data in batches or wait
until all the data have been gathered before deliver-
ing it
With this information, the assistant can develop
the database in preparation for receiving the data.
The time needed to prepare the database varies
depending on the number of variables and the com-
plexity of the response categories. Approximate dates
for completion of the data entry, analyses, or both
must be negotiated before beginning data collection.
If you have a deadline for completing the study
or presenting your results, you should share this
CHAPTER 20  Collecting and Managing Data 523
information with the people performing data entry
and analysis.
Pilot Study
Completing a pilot study may save you difficulty later
when you implement the final steps of the research
process. Pilot testing helps you to identify problems
you might encounter while collecting data and helps
you develop strategies for addressing potential prob-
lems. Chapter 3 provides reasons to conduct a pilot
study. Following approval of the study by your IRB,
use your research plan to recruit three to five subjects
who meet your eligibility criteria. Use the data collec-
tion methods that you have selected and prepared. Pay
attention to how long it takes to recruit a subject,
obtain informed consent, and collect the data. Ask the
participant to identify questions or aspects of the
process that were unclear or confusing. Based on the
pilot study and feedback of the first subjects, modify
your data collection forms and methods of data col-
lection to ensure the feasibility, validity, and reliability
of the study.
Collecting Data
Data collection is the process of selecting subjects
and gathering data from these subjects. The actual
steps of collecting the data are specific to each study
and depend on the research design and measurement
methods. Data may be collected on subjects by observ-
ing, testing, measuring, questioning, recording, or any
combination of these methods. The researcher is
actively involved in this process either by collecting
data or by supervising data collectors. You will apply
ethical principles, people-management strategies, and
problem-solving skills constantly as data collection
tasks are implemented. Even after pilot testing, snags
in the research plan can occur, and support systems
are needed for data collectors who encounter situa-
tions in the home or clinic that require reporting to
legal authorities. For example, during a home visit, a
data collector may find that family members are
neglecting a subject in the study who cannot get out
of bed. Frequent interactions with data collectors
on your team are also essential for assessing any
minor or major risks and reporting adverse effects to
your IRB.
Data Collection Tasks
In both quantitative and qualitative research, the
investigator performs four tasks during the process of
data collection. These tasks are interrelated and occur
524 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
concurrently rather than in sequence. The tasks are (1)
selecting subjects, (2) collecting data in a consistent
way, (3) maintaining research controls as indicated in
the study design, and (4) solving problems that threaten
to disrupt the study. Selecting subjects is discussed in
Chapter 15. Collecting data may involve administer-
ing Internet or paper-and-pencil scales; asking sub-
jects to complete data collection forms in person or
online; or recording data from observations, patient
medical records, or monitoring equipment (Chapters
16 and 17 provide information on measurement strate-
gies). Data collection tasks for qualitative studies are
discussed in more detail in Chapter 12.
Maintaining Control
Maintaining control and consistency of the design and
methods during subject selection and data collection
protects the integrity or validity of the study. Research-
ers build controls into the design to minimize the
influence of intervening forces on the study findings.
Maintenance of these controls is essential. For
example, a study to describe the changes in sleep
stages during puberty may require controlling the
environment of the bedroom to such an extent that a
sleep laboratory is the only setting in which integrity
can be maintained. Control is not always realistic in a
natural field setting, and, in some cases, these controls
can fail without the researcher realizing it. Often the
researcher must opt for a randomized controlled study
to address potential control failures so that they are
equally likely to occur in either group.
In addition to maintaining controls identified in the
research plan, you must continually watch for previ-
ously unidentified extraneous variables that might
have an impact on the data being collected. These
variables are often specific to a study and tend to
become apparent during the data collection period.
The extraneous variables identified during data collec-
tion must be considered during data analysis and inter-
pretation. These variables also must be noted in the
research report to allow future researchers to control
them. For example, Lee and Gay (2011) studied sleep
quality in new mothers and asked about the infant’s
sleep location, but the location at the beginning of the
night was often not the same by morning and could
not be controlled in the home setting.
Problem Solving
Little has been written about the problems encoun-
tered by nurse researchers. Research reports often read
as though everything went smoothly. The implication
is that if you are a good researcher, you will have no
problems, which is not true. Research journals
generally do not provide enough space for the
researcher to describe the problems encountered, and
inexperienced researchers may get a false impression.
Some of the problems are hinted at in a published
paper in either the limitations section or in a discus-
sion of areas for future research. A more realistic sense
of the problems encountered by a researcher can be
obtained through personal discussions with the
primary author about his or her process of data collec-
tion for a particular sample or using a particular
method or instrument. Some common problems expe-
rienced by researchers are discussed in the following
section.
A problem can be perceived either as a frustration
or as a challenge. The fact that the problem occurred
is not as important as successfully resolving it. The
final and perhaps most important task during the data
collection period may be debriefing with your research
team in weekly meetings for problem resolution.
Data Collection Problems
Murphy’s law (if anything can go wrong, it will, and
at the worst possible time) seems to prevail in research,
just as in other dimensions of life. For example, data
collection frequently requires more time than was
anticipated, and collecting the data is often more dif-
ficult than expected. Even following a pilot study, you
may encounter challenges during the data collection
process. Sometimes changes must be made in the way
the data are collected, in the specific data collected, or
in the timing of data collection. People react to the
study in unpredictable ways. Institutional changes
may force modifications in the research plan, or
unusual or unexpected events may occur. You must be
as consistent as possible during the data collection
process, but you must also be flexible in dealing with
unforeseen problems. Sometimes, sticking with the
original plan at all costs is a mistake. Skills in finding
ways to resolve problems that protect the integrity of
the study can be critical.
In preparation for data collection, possible prob-
lems must be anticipated, and solutions for these prob-
lems must be explored. The following discussion
describes some common problems and concerns and
presents possible solutions. Problems that tend to
occur with some regularity in studies have been cat-
egorized as people problems, researcher problems,
institutional problems, and event problems.
People Problems
Nurses cannot place a subject in a laboratory test
tube, instill one drop of the independent variable,
and then measure the effect. Nursing studies are

often conducted by examining subjects as they inter-
act with their environments. Many aspects of the
environment can be controlled by using a laboratory
setting, but other studies require the natural setting
to have external validity. When research involves
people, nothing is completely predictable. People, in
their complexity and wholeness, have an impact on
all aspects of nursing studies. Researchers, potential
subjects, family members of subjects, healthcare pro-
fessionals, institutional staff members, and others
(“innocent bystanders”) interact within the study sit-
uation. You will need to observe closely and evalu-
ate these interactions to determine their impact on
your study.
Problems Selecting a Sample
The first step in initiating data collection—selecting a
sample—may be the beginning of people problems.
You may find that few people are available who fit
your inclusion criteria or that many people you
approach refuse to participate in the study even though
the request seems reasonable. Appropriate subjects,
who were numerous a month earlier, seem to have
disappeared. Institutional procedures may change,
which might make many potential subjects ineligible
for participation in the study. You may have to evalu-
ate the inclusion and exclusion criteria or seek addi-
tional sources for potential subjects. In research
institutions that care for the indigent, patients tend to
be reluctant to participate in research. This lack of
participation might arise because these patients are
frequently exposed to studies, feel manipulated, or
misunderstand the research. Patients may feel that
they are being used or fear that they will be harmed
in some way. For example, recruiting Spanish-
speaking women for a study of stress and acculturation
may be met with high refusal rates if these women are
worried about revealing their legal status in the United
States. Recruiting women who are planning a preg-
nancy in the next 6 months may not yield participants
because they are fearful that others (work colleagues
or friends) will find out about their plan.
Subject Attrition
After you have selected a sample, certain problems
might cause subject attrition (a loss of subjects from
the study over time). For example, some subjects may
agree to participate but then fail to follow through.
Some may not complete needed forms and question-
naires or may fill them out incorrectly. To reduce these
problems, a research team member can be available to
subjects while they complete essential questions.
Some subjects may not return for a second interview
CHAPTER 20  Collecting and Managing Data 525
or may not be home for a scheduled visit. Although
you have invested time to collect data from these
subjects, their data may have to be excluded from
analysis because of incompleteness. Generally, the
more data collection time points you require as part of
your design, the higher the risk for attrition. Attrition
can occur because of subject burden accumulating
over time, because healthy adults relocate for employ-
ment or family reasons, or because of death in a more
critically ill population.
Sometimes subjects must be dropped from the
study by the research team because of changes in
health status. For example, a patient may be trans-
ferred out of the intensive care unit where the study is
being conducted. Another possibility might be that the
patient’s condition may worsen and the patient no
longer meets the inclusion criteria. The limits of third-
party reimbursement may force the healthcare pro-
vider to discontinue the services you are studying.
Subject attrition occurs to some extent in all longi-
tudinal studies. One way for you to deal with this
problem is to anticipate the attrition rate and increase
the planned number of subjects to ensure that a mini-
mally desired number will complete the full study.
Review similar studies to anticipate the attrition rate.
For example, Lim, Chiu, Dohrmann, and Tan (2010)
reported a 31% attrition rate in their quasi-experimental
study of the knowledge of registered nurses employed
in long-term care. The investigators collected pretest
data from 58 subjects and 4 weeks later collected post-
test data from 40 subjects. If subject attrition is higher
than expected, consider additional small incentives
along the way or a bonus for completing the final
assessment to achieve an adequate final sample size.
Attrition is usually higher in the placebo or control
group, but a well-designed, attention-control group
should have an attrition rate that is similar to the attri-
tion seen in your intervention group. Sometimes a
study might end with a smaller than expected sample
size. If so, the effect of a smaller sample on the power
of planned statistical analyses must be considered
because this smaller sample may be inadequate to test
the hypotheses.
Researchers should report information about sub-
jects’ acceptance to participate in a study and attrition
during the study to determine if the sample is
representative of the study target population. Journal
editors often require that manuscripts include a flow
chart indicating the number of subjects meeting
sample criteria, the numbers refusing to participate,
and the reasons for refusal. If data are collected over
time (repeated measures) or the study intervention is
implemented over time, subjects often drop out of a
526 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research

Assessed for eligibility

(n  165)

Excluded (n  68)

Not meeting inclusion criteria

(n  25)
Enrollment
Refused to participate
(n  18)
Is it Randomized? YES Other reasons: Did not speak
English (n  25)

Allocated to interventions
Telephone Counseling
(n  38 women and 38 partners)
Exercise Allocated to control intervention
(n  23 women and 23 partners) (n  37 women and 37 partners)
Received allocated intervention Received allocated intervention
Telephone Counseling Allocation (n  36)
(n  38 women and 38 partners) Did not receive allocated intervention
Exercise (n  1)
(n  23 women and 23 partners) Give reasons-Not interested
Did not receive allocated intervention
(n  0)

Lost to follow-up
Lost to follow-up (n  3 women, 7 partners by T3)
(n  2 women, 4 partners lost in Give reasons: Not interested
exercise condition by T3) Follow-Up Discontinued intervention
Give reasons-not interested (n  2)
Discontinued intervention Give reasons-Not interested
(n  0)

Analyzed Analyzed
(n  38 women, and partners for (n  33 women, 30 partners)
counseling; 21 women and 19 partners Excluded from analysis
Analysis
for exercise) (n  3 women, 7 partners)
Excluded from analysis Give reasons-No T3 data
(n  2 women and 4 partners in collected
exercise)
Give reasons-No T3 data collected

Figure 20-7  Sample selection and allocation.

T3 = Tumor more than 5cm across. (From Badger, T., Segrin, C., Dorros, S., Meek, P., & Lopez, A. M. [2007]. Depression and anxiety in women with
breast cancer and their partners. Nursing Research, 56[1], 44–53.)

study, and it is important to document when and how
much attrition occurred. Using the CONSORT guide-
lines published by Bennett (2006), Badger, Segrin,
Dorros, Meek, and Lopez (2007) provided a flow chart
to document participants’ selection, refusal rate,
assignment to group, and attrition over the weeks
of data collection (see Figure 20-7). The flow chart
clearly identifies important aspects of the sampling
process and reasons for attrition. This information
enables researchers and clinicians to evaluate the rep-
resentativeness of their sample for external validity
and for any potential bias in interpreting the results.
Subject as an Object
The quality of interactions between the researcher and
subjects during the study is a critical dimension for

maintaining subject participation. When researchers
are under pressure to complete a study, people can be
treated as objects rather than as subjects, particularly
if electronic data collection is used. In addition to being
unethical, such impersonal treatment alters interac-
tions, diminishes subject satisfaction, and increases
likelihood for missing data and subject attrition. Sub-
jects are scarce resources and must be treated with
care. Treating the subject as an object can affect
another researcher’s ability to recruit from this popula-
tion in the future. Treating the subject as an object can
be minimized by building strategies into the consent
process, such as offering them a personal copy of their
results, recognizing their valuable participation with
small gifts as tokens of appreciation, or providing
monetary reimbursement for their time and effort.
External Influences on Subject Responses
People interacting with the subject, the researcher, or
both can have an important impact on the data collec-
tion process. Family members may not agree to the
subject’s participation in the study or may not under-
stand the study process. These individuals often influ-
ence the subject’s decision to participate. Researchers
benefit from taking the time to explain the study and
seeking cooperation of family members. Family coop-
eration is essential when the potential subject is criti-
cally ill and unable to give informed consent.
Family members or other patients may also influ-
ence the subject’s responses to scales or interview
questions. In some cases, subjects may ask family
members, friends, or other patients to complete study
forms for them. The subject may discuss questions on
the forms with other people who happen to be in the
room, and therefore the data recorded do not reflect
the subject’s real feelings. If interviews are conducted
while others are in the room, the subject’s responses
may depend on his or her need to meet the expecta-
tions of the other persons. Sometimes a family member
may answer questions addressed verbally to the
patient. The setting in which a questionnaire is com-
pleted or an interview is conducted may determine the
extent to which the answers are a true reflection of a
subject’s feelings. If the privacy afforded by the setting
varies from one subject to another, the subjects’
responses may also vary and threaten both the internal
and the external validity of the findings.
Usually, the most desirable setting for an interview
is a private area away from distractions. If it is not
possible to arrange for such a setting, the researcher
can be present at the time the questionnaire is com-
pleted to decrease the influence of others. If the ques-
tionnaire is to be completed later or taken home and
CHAPTER 20  Collecting and Managing Data 527
returned at a later time, the probability of influence by
others increases, and return of the questionnaire packet
becomes less likely, even if the subject is provided
with a stamped return envelope. The impact of this
problem on the integrity of the data depends on the
nature of the questionnaire items. For example, a
marital relationship questionnaire may have different
responses if the subject is allowed to complete it alone
and return it immediately to the researcher or have to
complete it aloud with the spouse in attendance.
Passive Resistance
Healthcare professionals and institutional staff mem­
bers working with the study participants in clinical
settings may affect the data collection process. Some
professionals verbalize strong support for the study
and yet passively interfere with data collection. For
example, nurses providing care may fail to follow the
guidelines agreed on for providing the specific care
activities being studied, or information needed for the
study may be left off patient records. The researcher
may not be informed when a potential subject has been
admitted, and a physician who has agreed that his or
her patients can be participants may decide as each
patient is admitted that this one is not quite right for
the study. In addition, when the permission of the
physician or nurse practitioner is required, the pro-
vider might be unavailable to the researcher.
Nonprofessional staff members may not realize the
impact of the data collection process on their work
patterns until the process begins. The data collection
process may violate their beliefs about how care
should be provided (or has been provided). If ignored,
their resistance can completely undo a carefully
designed study. For example, research on skin care
may disrupt a nursing aide’s bathing routine, so he or
she may continue the normal routine regardless of the
study protocol and invalidate the study findings. When
there is funding to support subject recruitment and
data collection, funds can be used to reimburse the
clinic or hospital staff members for their time, to
create a raffle for one substantial gift, to offer a gift
certificate to buy something needed for the clinic, or
to send a nurse to a continuing education course.
When funding is limited, staff members’ enthusiasm
for your study may be enhanced if they are able to
participate in the research as authors or presenters in
dissemination of the research findings.
Because of the potential impact of these problems,
the researcher must maintain open communication and
nurture positive relationships with other professionals
and staff members during data collection. Problems
that you and your team recognize early and deal with
528 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
promptly have fewer serious consequences than prob-
lems you try to ignore. However, not all problems can
be resolved. Sometimes you may need to seek creative
ways to work around an individual or to counteract
the harmful consequences of passive resistance.
Researcher Problems
Some problems are a consequence of the researcher’s
interaction with the study situation or lack of skill in
data collection techniques. These problems are often
difficult to identify because of the researcher’s per-
sonal involvement. However, their effect on the study
can be serious.
Researcher Interactions
Researcher interactions can interfere with data collec-
tion in interview situations. To gain the cooperation of
the subject, the researcher needs to develop rapport
with the subject. One way to do this is to select data
collectors who resemble the types of subjects being
recruited as much as possible. Rapport may suffer if
a young man collects data from female caregivers of
elderly adults about their experience with end-of-life
care. Similarly, a white middle-aged woman collecting
data from young African American men or Hispanic
teens is likely to be more disadvantaged than a data
collector who is more similar to the population of
interest.
A balance is needed between rapport and over-
involvement. The researcher can become so involved
in interactions with a study participant that data col-
lection on that particular subject is not completed. If
you are collecting data from patient records while you
are surrounded by professionals with whom you inter-
act socially and professionally, it is sometimes diffi-
cult to focus completely on the study situation. This
lack of attention usually leads to loss of data.
Lack of Skill in Data Collection Techniques
The researcher’s skill in using a particular data collec-
tion technique can affect the quality of the data col-
lected. A researcher who is unskilled at the beginning
of data collection might practice the data collection
techniques with the assistance of an experienced
researcher. A pilot study to test data collection tech-
niques is always helpful. If data collectors are being
used, they also need opportunities to practice data
collection techniques before the study is initiated.
Sometimes a skill is developed during the course of a
study; if this is the case, as one’s skill increases, the
data being collected may change and confound the
study findings and threaten the validity of the study.
If more than one data collector is used, the degree to
which skills improve may vary across time and data
collectors. The consistency of data collectors must be
evaluated during the study to detect any changes in
their data collection techniques.
Researcher Role Conflict
As a researcher, one is observing and recording events.
Nurses who conduct clinical research often experience
a conflict between their researcher role and their clini-
cian role during data collection. In some cases, the
researcher’s involvement in the event, such as provid-
ing physical or emotional care to a patient during data
collection, could alter the event and bias the results. It
would be difficult to generalize the findings to other
situations in which the researcher was not present to
intervene. However, the needs of patients must take
precedence over the needs of the study.
The dilemma is to determine when the needs of
patients are great enough to warrant researcher inter-
vention. Some patient questions or situations are life-
threatening, such as respiratory distress and changes
in cardiac function, and require immediate action by
anyone present. Other patient needs are simple, can be
addressed by any nurse available, and can be answered
if the response is not likely to alter the results of your
study. Examples of these interventions include giving
the patient a bedpan, informing the nurse of the
patient’s need for pain medication, or helping the
patient to open food containers. These situations
seldom cause a dilemma.
Solutions to other situations are not as easy. For
example, suppose that your study involves examining
the emotional responses of patients’ family members
during and immediately after the patient’s surgery.
Your study includes an experimental group that
receives one 30-minute family support session before
and during the patient’s surgery and a control group
that receives no support session. Both sets of families
are being monitored for 1 week after surgery to
measure level of anxiety and coping strategies. You
are currently collecting data on the control group. The
data consist of demographic information and scales
measuring anxiety and coping. One of the family
members is in great distress. After completing the
demographic information, she verbally expresses her
fears and the lack of support she has received from the
nursing staff. Two other subjects from different fami-
lies hear the expressed distress and concur; they move
closer to the conversation and look to you for informa-
tion and support.
In this situation, a supportive response from you is
likely to modify the results of the study because these
responses are part of the treatment to be provided to

the experimental group only. This interaction is likely
to narrow the difference between the two groups and
decrease the possibility that your results will show a
significant difference between the two groups. How
should you respond? Are you obligated to provide
support? To some extent, almost any response would
be supportive. One alternative is to provide the needed
support and not include these family members in the
control group. Another alternative is to recruit the help
of a nonprofessional to collect the data from the
control group. However, most people would provide
some degree of support in the described situation,
even though their skills in supportive techniques may
vary.
Other dilemmas include witnessing unethical
behavior that interferes with patient care or witnessing
subjects’ unethical or illegal behavior (Humphreys
et al., 2011). Consent forms are often required to stipu-
late that any member of the research team is legally
required to report illegal behaviors, such as neglect or
abuse of children and elderly adults. Try to anticipate
these dilemmas before data collection whenever pos-
sible and include this information in the consent form
(Wong, Tiwari, Fong, Humphreys, & Bullock, 2011).
Pilot studies can help you to identify dilemmas likely
to occur in a study, and you can build strategies into
the design to minimize or avoid them. However, some
dilemmas cannot be anticipated, and you must respond
to these problems spontaneously. There is no pre-
scribed way to handle difficult dilemmas; each case
must be dealt with individually. You should discuss any
unethical and illegal behavior with members of your
IRB or ethics committee or with legal advisors. These
situations must be reported to the IRB, and experts
there can advise you on the next step or course of
action. After you have resolved the dilemma, it is wise
to reexamine the situation for its effect on study results
and consider options in case the situation arises again.
Another type of conflict arises when a subject
makes inaccurate statements or asks a question about
health practices or treatment. Rather than offering pro-
fessional advice or responding to the question, the
research nurse should acknowledge that it is a good
question, but that the research protocol does not allow
for a response during data collection. When data
collection is complete, the research nurse can help
formulate the question for the subject’s healthcare
provider or provide a readily available pamphlet or
website for more information.
Maintaining Perspective
Data collection includes both joys and frustrations.
Researchers must be able to maintain some degree of
CHAPTER 20  Collecting and Managing Data 529
objectivity during the process and yet not take them-
selves too seriously. A sense of humor is invaluable.
You must be able to experience the emotions and then
become the rational problem solver. Management
skills and mental health are as invaluable to a research
career as being obsessive about data collection and
data management.
Institutional Problems
Institutions are in a constant state of change. They will
not stop changing for the period of a study, and these
changes often affect data collection. A nurse who has
been most helpful in your study may be promoted or
transferred. The unit on which your study is conducted
may be reorganized, moved, or closed during data
collection. An area used for subject interviews may be
transformed into an office or a storeroom. Patient
record forms may be revised, omitting data that you
and your team are collecting. The medical record per-
sonnel may be reorganizing files and temporarily
unable to provide needed records.
These problems for the most part are completely
outside of the researcher’s control. Pay attention to
the internal communication network of the institution
for advanced warning of impending changes. Con-
tacts within the institution’s administrative decision
makers could warn you about the impact of proposed
changes on an ongoing study. In many cases, the IRB
in the local hospital will have a nurse representative
who can provide the needed consultation. However,
in many cases, data collection strategies might have
to be modified to meet the newly emerging situation.
Balancing flexibility with maintaining the integrity
of the study may be the key to successful data col-
lection. As a data collection site, the subject’s home
setting may be more desirable and convenient for a
subject than a complex facility or institution, and
response rates may improve. The disadvantage is
that home visits are time intensive for the researcher,
and the subject may not be home at the agreed
appointment time despite confirmed appointments
and reminder calls.
Event Problems
Unpredictable events can be a source of frustration
during a study. Research tools ordered from a testing
company may be lost in the mail. The printer may
break down just before 500 data collection forms are
to be printed, or a machine to be used in data collec-
tion may break down and require 6 weeks for repair.
A computer ordered for data collection may not arrive
when promised or may malfunction. Data collection
forms may be misplaced, misfiled, or lost.
530 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
Local, national, or world events can also influence
a subject’s response to a questionnaire or ability to
enroll in a study. For example, a researcher conducting
a study of the effects of fatigue on the health of air
traffic controllers encounters a rash of national media
reports about controllers falling asleep while on duty.
This event could be expected to modify subjects’
responses. In attempting to deal with the impact of the
event on the study, the researcher could obtain IRB
approval for a modification that would allow for con-
tinued data collection from the intended sample but to
examine the impact of news such as this on subjects’
responses rather than the original purpose. However,
the emotional climate of the airports participating in
the study may not be conducive to this option. The
researcher may choose to wait 3 months before col-
lecting additional data and examine the data before
and after the event for statistically significant differ-
ences in responses. If no differences are found, the
researcher could justify using all the data for
analysis.
Other, less dramatic events can also have an impact
on data collection. If data collection for the entire
sample is planned for a single time, a snowstorm or a
flood may require that the researcher cancel the
session. Weather may decrease attendance far below
that expected at a support group or series of teaching
sessions. A bus strike can disrupt transportation
systems to such an extent that subjects can no longer
get to the data collection site. A new health agency
may open in the city, which may decrease demand for
the care activities being studied. Conversely, an exter-
nal event can also increase attendance at clinics to
such an extent that existing resources are stretched and
data collection is no longer possible. These events are
also outside the researcher’s control and are impossi-
ble to anticipate. In most cases, however, restructuring
the data collection period can salvage the study. To do
so, it is necessary to examine all possible alternatives
for collecting the study data. In some cases, data col-
lection can simply be rescheduled; in other situations,
the changes may need to be more complex. For
example, recruiting women to participate in a study
that requires an hour or longer of their time may
necessitate that the researcher provide childcare. Pro-
viding childcare would be more costly and add com-
plexity to the process, but it may be the best alternative
for increasing participation.
Serendipity
Serendipity is the accidental discovery of something
useful or valuable. During the data collection phase of
studies, researchers often become aware of elements
or relationships that they had not previously identified.
These aspects may be closely related to the study
being conducted or have little connection with it. They
come from increased awareness and close observation
of the study situation. Because the researcher is
focused on close observation, other elements in the
situation can come into clearer focus and take on new
meaning. Serendipitous findings are important to the
development of new insights in nursing theory. They
can be important for understanding the totality of the
phenomenon being examined. Additionally, they lead
to areas of research that generate new knowledge. A
relatively easy way to capture these insights as they
occur is to keep a research journal. These events must
be carefully recorded, even if their impact or meaning
is not understood at the time.
Serendipitous findings can also lead the researcher
astray. Sometimes researchers forget the original plan
and redirect their attention to the newly discovered
dimensions. Although modifying data collection to
include data related to the new discovery may be
valid, there has not been time to plan carefully a study
related to the new findings. The study’s approval by
the IRB covers only information included in the sub-
mitted study proposal. Examination of the new data
should be an offshoot only of the initial study and
would require seeking additional IRB approval.
Having Access to
Support Systems
The researcher must have access to individuals or
groups who can provide mentorship, support, and con-
sultation during the data collection period. Support
can usually be obtained from academic committees,
from IRB staff, and from colleagues on your research
team.
Support of Academic Committees
Although thesis and dissertation committees are basi-
cally seen as stern keepers of the sanctity of the
research process, they also serve as support systems
for novice researchers. Committee members must be
selected from faculty who are willing and able to
provide the needed expertise and support. Experi-
enced academic researchers are usually more knowl-
edgeable about the types of support needed. Because
they are directly involved in research, they tend to be
sensitive to the needs of a novice researcher and more
realistic about what can be accomplished in the des-
ignated time frame.

Institutional Support
A support system within the institution where the
study is being conducted is also important. Support
might come from people serving on the institutional
research committee or from nurses working on the
unit where the study is conducted. These people
may have knowledge of how the institution functions,
and their closeness to the study can increase their
understanding of the problems experienced by the
researcher and subjects. Do not overlook their ability
to provide useful suggestions and assistance. Your
ability to resolve some of the problems encountered
during data collection may depend on having someone
within the power structure of the institution who can
intervene.
Colleague Support
In addition to professional support, having at least one
peer in your research world with whom to share the
joys, frustrations, and current problems of data collec-
tion is important. This colleague can often serve as a
mirror to allow you to see the situation clearly and
perhaps more objectively. With this type of support,
the researcher can share and release feelings and gain
some distance from the data collection situation.
Alternatives for resolving the problem can be dis-
cussed in a less emotional context. Data collection is
demanding but rewarding. With time, confidence and
expertise of the novice researcher increase.
Data Safety and Monitoring Board as Source
of Support
If you are conducting an intervention study that is
deemed to be of low risk to the patient, such as a
behavioral intervention to improve sleep quality, a
data safety and monitoring plan will suffice. In these
situations, less support for you as a researcher is
needed. This plan is deemed adequate when it con-
oasis-ebl|Rsalles|1476461052
forms to the IRB requirements for reporting any
adverse event and includes annual progress reports. It
requires that the researcher explicitly states the plan
to review the data from each set number of subjects
or from each 3-month or 6-month batch of recruited
subjects, depending on the extent of the study.
If the study involves an intervention protocol that
is higher risk to patient safety, a data safety and moni-
toring board is required. This board includes members
who are not directly involved in the study and who
can be objective about the findings to date. This board
should meet at regular intervals and discuss whether
the study should continue or be stopped based on the
data collected to date. The board should consist of
very experienced researchers and clinical experts. See
CHAPTER 20  Collecting and Managing Data 531
Chapter 14 for more information on conducting inter-
vention studies.
Managing Data
Once data collection begins, you have to be prepared
to handle large quantities of data. To avoid a state of
total confusion, make careful plans before data collec-
tion begins. Plans are needed to keep all data from a
single subject together until analysis is initiated. Write
the subject code number on each form, and check the
forms for each subject to ensure that they all are
present. Researchers have been known to sort their
data by form, such as putting all the scales of one kind
together, only to realize afterward that they had failed
to code the forms with subject identification numbers
first. They then had no idea which scale belonged to
which subject, and valuable data were lost.
Allot space as needed for storing forms. Purchase
file folders, and design a labeling method to allow
easy access to data; color coding is often useful. If you
are using multiple forms, the subject’s demographic
sheet could be one light color, with a different pastel
color for the pain questionnaire and a contrasting light
color for all the physiological data sheets used to
record blood pressure, pulse, and respiration readings.
Use envelopes to hold small pieces of paper or note
cards that might fall out of a file folder. Plan to code
data and enter them into the computer as soon as pos-
sible after data collection to reduce the loss or disor-
ganization of data. If data are collected on a computer,
make sure the data are backed up and stored in a sepa-
rate space so that they are not lost if the computer fails.
Preparing Data for Computer Entry
Data must be carefully checked and problems cor-
rected before you initiate the data entry phase. The
data entry process should be essentially automatic and
require no decisions regarding the data. Anything that
alters the rhythm of data entry increases errors. For
example, the subject’s entry should be coded as it
appears, and any reverse coding that may be needed
should be done at a later time by computer manipula-
tion in a consistent manner rather than trying to have
the data entry person recode during data entry.
Such simplicity in data entry reduces the number
of data entry errors and markedly decreases the time
required for entry. It is not sufficient to establish
general rules for individuals entering data such as “in
this case always do X.” This action still requires the
person who is entering data to recognize a problem,
refer to a general rule, and correct the data before
entry. Correcting the data requires using a different
532 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
color ink from the subject’s mark, and the correction
should be initialed by the researcher making the
correction.
1. Missing data. Provide the data if possible or deter-
mine the impact of the missing data on your analy-
sis. In some cases, the subject must be excluded
from at least some of the analyses, so you must
determine what data are essential.
2. Items in which the subject provided two responses
when only one was requested. For example, if the
question asked the subject to mark the most impor-
tant item in a list of 10 items and the subject
selected 2 items, you must decide how to resolve
this problem; do not leave the decision to an assis-
tant who is entering the data. In the codebook and
on the form itself, indicate how that particular
datum is to be coded and entered.
3. Items in which the subject has marked a response
between two options. This problem commonly
occurs with Likert-type scales, particularly scales
using forced choice options. Given four options,
the subject places a mark on the line between
response 2 and response 3. In the codebook and on
the form, indicate how the datum is to be coded.
This is often best coded as a missing value, but
coding rules should be consistent. A rationale can
be made to take the highest value, the lowest value,
or code toward the center value.
4. Items that ask the subject to write in some informa-
tion such as occupation or diagnosis. Such items
are a data enterer’s nightmare. Develop a list of
codes for entering such data. Rather than leaving it
up to the assistant to determine which code matches
the subject’s written response, the researcher
should enter this code in a different color and initial
that change before turning the data over for entry.
After the data have been checked and needed codes
written in, it is prudent to make a copy rather than
turning over the only set of your data to an
assistant.
Data Entry Period
If you are entering your own data, develop a rhythm
to your data entry process. Avoid distractions while
entering data, and limit your data entry periods to
2-hour intervals to reduce fatigue, errors, and repeti-
tive wrist strain or injury. Backup the database after
each data entry period, and store it on an encrypted
flash drive, on a secure website, or in a fireproof safe.
It is possible for the computer to crash and lose all of
your data. If an assistant is entering your data, make
yourself as available as possible to respond to ques-
tions and address problems. After entry, the data
should be randomly checked for accuracy. Data check-
ing is discussed in Chapter 21.
Storage and Retrieval of Data
In this time of flash drives and thumb drives, it is
relatively easy to store data. The original data forms
and database must be stored for a specified number of
years dictated by the funding source or by the journal
publisher. There are several reasons to store data. The
data can be used for secondary analyses. For example,
researchers participating in a project related to a par-
ticular research focus may pool data from various
studies for access by all members of the group. Data
should be available to document the validity of your
analyses and the published results of your study.
Because of nationally publicized incidents of scien-
tific misconduct, where researchers fabricated data
and published multiple manuscripts, you would be
wise to preserve documentation that your data were
obtained as you claim. Issues that have been raised
include how long data should be stored, the need
for institutional policy regarding data storage, and
whether graduate students who conduct a study
should leave a copy of their data at the university.
Some researchers store their data for 5 years after
publication, whereas others store their data until they
retire from a research career. Researchers should
check with their funding sponsors and publishers for
guidelines on how long to keep the data. Most
researchers store data in their office or laboratory;
others archive their data in a central location with
storage fees or retrieval fees. Graduate students do
have a responsibility to keep and securely store data
from their studies.
KEY POINTS
• Careful planning is needed before collecting and
managing data.
• The researcher may need to develop data collection
forms and format these forms to promote accuracy
and ease of data entry.
• The researcher must determine exactly how and in
what sequence data will be collected and the timing
of the process. Information about the procedures to
be used must be described in the subject’s informed
consent.
• The researcher must decide who will collect the
data.
• If data collectors are used, they must be provided
information about the research project, the instru-
ments, and data collection protocol.

• Consistency in data collection across subjects is
critical, and training is required to promote consis-
tency among data collectors if more than one data
collector is used.
• After training, data collectors must be evaluated
periodically and randomly to determine their
consistency.
• Decision points that occur during data collection
must be identified, and all options must be
considered.
• Data collection also involves maintaining research
controls and solving problems that threaten to
disrupt the study.
• Problems that arise during data collection involve
recruitment and attrition issues, treatment of the
subject as an object, external influences on subject
responses, passive resistance from staff members
or family, researcher interactions, lack of skill in
data collection techniques, and researcher role
conflicts.
• A successful study requires support that is often
obtained from academic committees, healthcare
agencies, and work colleagues.
• Data collected during a study must be accurately
entered in an encrypted computer and safely stored
in either a data repository or on an encrypted flash
drive.
REFERENCES
Badger, T., Segrin, C., Dorros, S., Meek, P., & Lopez, A. M. (2007).
Depression and anxiety in women with breast cancer and their
partners. Nursing Research, 56(1), 44–53.
Bennett, J. M. (2006). The consolidated standards of reporting trials
(CONSORT): Guidelines for reporting randomized trials. Nursing
Research, 54(2), 128–132.
Fukuoka, Y., Kamitani, E., Dracup, K., & Jong, S. S. (2011). New
insights into compliance with a mobile phone diary and pedometer
use in sedentary women. Journal of Physical Activity & Health,
8(3), 398–403.
Gray, J. & Johnson, (2009). Nurse International Relocation
Questionnaire 2. Unpublished research tool. Available from
jgray@uta.edu.
CHAPTER 20  Collecting and Managing Data 533
Harris, P. A., Taylor, R., Thielke, R., Payne, J., Gonzalez, N., &
Conde, J. G. (2009). Research electronic data capture (REDCap)
—A metadata-driven methodology and workflow process for pro-
viding translational research informatics support. Journal of
Biomedical Informatics, 42(2), 377–381.
Harwood, E. M. (2009). Data collection methods series: Part 3:
Developing protocols for collecting data. Journal of Wound
Ostomy Continence Nursing, 36(3), 246–250.
Harwood, E. M. & Hutchinson E. (2009). Data collection methods
series: Part 2: Select the most feasible data collection mode.
Journal of Wound Ostomy Continence Nursing, 36(2), 129–135.
Humphreys, J., Epel, E. S., Cooper, B. A., Lin, J., Blackburn, E. H.,
& Lee, K. A. (2011, March 8). Telomere shortening in formerly
abused and never abused women. Biological Research for
Nursing. Available from http://brn.sagepub.com/content/early
/2011/03/07/1099800411398479.
Im, E., Chee, W., Guevara, E., Liu, Y., Lim, H., Tsai, H., et al.
(2007). Gender and ethnic differences in cancer pain experience:
A multiethnic survey in the United States. Nursing Research,
56(5), 296–306.
Kang, D. H., Davis, L., Habermann, B., Rice, M., & Broome, M.
(2005). Hiring the right people and management of research staff.
Western Journal of Nursing Research, 27(8), 1059–1066.
Lee, K. A. & Gay, C. L. (2011). Can modifications to the bedroom
environment improve the sleep of new parents? Two randomized
controlled trials. Research in Nursing & Health, 34(1), 7–19.
Lim, L. M., Chiu, L. H., Dohrmann, J., & Tan, K. (2010). Registered
nurses’ medication management of the elderly in aged care facili-
ties. International Nursing Review, 57(1), 98–106.
Melnyk, B. M. & Fineout-Overholt, E. (2010). Evidence-based
practice in nursing and healthcare: A guide to best practice (2nd
ed.). Philadelphia, PA: Lippincott Williams & Wilkins.
Office of Minority Health. (2010). OMB standards for data on race
and ethnicity. Retrieved from http://minorityhealth.hhs.gov/.
Savian, C., Paratz, J., & Davies, A. (2006). Comparison of the
effectiveness of manual and ventilator hyperinflation at different
levels of positive end-expiratory pressure in artificially ventilated
and intubated intensive care patients. Heart & Lung, 35(5),
334–341.
Waltz, C. F., Strickland, O. L., & Lenz, E. R. (2010). Measurement
in nursing and health research (4th ed.). New York, NY: Springer
Publishing Company.
Wong, J. Y., Tiwari, A., Fong, D. Y., Humphreys, J., & Bullock, L.
(2011). Depression among women experiencing intimate partner
violence in a Chinese community. Nursing Research, 60(1),
58–65.

21
CHAPTER
Introduction to Statistical Analysis
D
ata analysis is often considered one of the most
exciting steps of the research process. During
this phase, you will finally obtain answers to
the questions that led to the development of your
study. Nevertheless, nurses probably experience
greater anxiety about this phase of the research process
than any other, as they question issues that range from
their knowledge about critically appraising published
studies to their ability to conduct research. Critical
appraisal of the results section of a quantitative study
requires you to be able to (1) identify the statistical
procedures used; (2) judge whether these statistical
procedures were appropriate for the hypotheses, ques-
tions, or objectives of the study and for the data avail-
able for analysis; (3) comprehend the discussion of
data analysis results; (4) judge whether the author’s
interpretation of the results is appropriate; and (5)
evaluate the clinical importance of the findings (see
Chapter 18 for more details on critical appraisal).
As a neophyte researcher performing a quantitative
study, you are confronted with many critical decisions
related to data analysis that require statistical knowl-
edge. To perform statistical analysis of data from a
quantitative study, you need to be able to (1) determine
the necessary sample size to power your study ade-
quately; (2) prepare the data for analysis; (3) describe
the sample; (4) test the reliability of measures used in
the study; (5) perform exploratory analyses of the
data; (6) perform analyses guided by the study objec-
tives, questions, or hypotheses; and (7) interpret the
results of statistical procedures. We recommend con-
sulting with a statistician or expert researcher early in
the research process to help you develop a plan for
accomplishing these seven tasks. A statistician is also
invaluable in conducting data analysis for a study and
interpreting the results.
Critical appraisal of the results of studies and sta-
tistical analyses both require an understanding of the
statistical theory underlying the process of analysis.
534
  http://evolve.elsevier.com/Grove/practice/
This chapter and the following four chapters provide
you with the information needed for critical appraisal
of the results sections of published studies and for
performance of statistical procedures to analyze data
in studies and in clinical practice. This chapter intro-
duces the concepts of statistical theory and discusses
some of the more pragmatic aspects of quantitative
data analysis: the purposes of statistical analysis, the
process of performing data analysis, the method for
choosing appropriate statistical analysis techniques
for a study, and resources for conducting statistical
analysis procedures. Chapter 22 explains the use of
statistics for descriptive purposes, such as describing
the study sample or variables. Chapter 23 focuses on
the use of statistics to examine proposed relationships
among study variables, such as the relationships
among the variables dyspnea, anxiety, and quality of
life. Chapter 24 explores the use of statistics for pre-
diction, such as using independent variables of age,
gender, cholesterol values, and history of hypertension
to predict the dependent variable of cardiac risk level.
Chapter 25 guides you in using statistics to determine
differences between groups, such as determining the
difference in muscle strength and falls (dependent
variables) between an experimental or intervention
group receiving a strength training program (indepen-
dent variable) and a comparison group receiving stan-
dard care.
Concepts of Statistical Theory
One reason nurses tend to avoid statistics is that many
were taught the mathematical mechanics of calculat-
ing statistical formulas and were given little or no
explanation of the logic behind the analysis procedure
or the meaning of the results (Grove, 2007). This
mathematical process is usually performed by com-
puter, and information about it offers little assistance
to the individuals making statistical decisions or
 CHAPTER 21  Introduction to Statistical Analysis 535
explaining results. We approach data analysis from the
perspective of enhancing your understanding of the
meaning underlying statistical analysis. You can use
this understanding either for critical appraisal of
studies or for conducting data analyses.
The ensuing discussion explains some of the con-
cepts commonly used in statistical theory. The logic
of statistical theory is embedded within the explana-
tions of these concepts. The concepts presented in this
chapter include probability theory, classical hypothe-
sis testing, Type I and Type II errors, statistical power,
statistical significance versus clinical importance,
inference, samples and populations, descriptive and
inferential statistical techniques, measures of central
tendency, the normal curve, sampling distributions,
symmetry, skewness, modality, kurtosis, variation,
confidence intervals, and parametric and nonparamet-
ric types of inferential statistical analyses.
Probability Theory
Probability theory addresses statistical analysis as
the likelihood of accurately predicting an event or the
extent of an effect. Nurse researchers might be inter-
ested in the probability of a particular nursing outcome
in a particular patient care situation. For example,
what is the probability of patients older than 75 years
of age with cardiac conditions falling when hospital-
ized? With probability theory, you could determine
how much of the variation in your data could be
explained by using a particular statistical analysis. In
probability theory, the researcher interprets the
meaning of statistical results in light of his or her
knowledge of the field of study. A finding that would
have little meaning in one field of study might be
important in another (Good, 1983; Kerlinger & Lee,
2000). Probability is expressed as a lowercase p, with
values expressed as percentages or as a decimal value
ranging from 0 to 1. For example, if the exact prob-
ability is known to be 0.23, it would be expressed as
p = 0.23. The p in statistics is defined as the probability
of rejecting the null hypothesis when the null is actu-
ally true. Nurse researchers typically consider a p =
0.05 value or less to indicate a real effect.
Classical Hypothesis Testing
Classical hypothesis testing refers to the process of
testing a hypothesis to infer the reality of an effect.
This process starts with the statement of a null hypoth-
esis, which assumes no effect (e.g., no difference
between groups, or no relationship between variables).
The researcher sets the values of two theoretical prob-
abilities: (1) the probability of rejecting the null
hypothesis when it is in fact true (alpha [α], Type I
error) and (2) the probability of retaining the null
hypothesis when it is in fact false (beta [β], Type II
error). In nursing research, alpha is usually set at
0.05, meaning that the researcher will allow a 5% or
lower chance of making a Type I error. The beta is
frequently set at 0.20, meaning that the researcher will
allow for a 20% or lower chance of making a Type II
error.
After conducting the study, the researcher culmi-
nates the hypothesis testing process by making a ratio-
nal decision either to reject or to retain the null
hypothesis, based on the statistical results. The follow-
ing steps outline each of the components of statistical
hypothesis testing.
1. State your primary null hypothesis. (Chapter 8 dis-
cusses the development of the null hypothesis.)
2. Set your study alpha (Type I error); this is usually
α = 0.05.
3. Set your study beta (Type II error); this is usually
β = 0.20.
4. Conduct power analyses (Aberson, 2010; Cohen,
1988).
5. Design and conduct your study.
6. Compute the appropriate statistic on your obtained
data.
7. Compare your obtained statistic with its corre-
sponding theoretical distribution in the tables pro-
vided in the Appendices at the back of this book.
For example, if you analyzed your data with a t-
test, you would compare the t value from your
study with the critical values of t in the table.
8. If your obtained statistic exceeds the critical value
in the distribution table, you can reject your null
hypothesis. If not, you must accept your null
hypothesis. These ideas are discussed in more
depth in Chapters 23 through 25 when the results
of statistics are presented.
Cox (1958, p. 159) stated, “Significance tests, from
this point of view, measure the adequacy of the data
to support the qualitative conclusion that there is a true
effect in the direction of the apparent difference.”
Thus, the decision is a judgment and can be in error.
The level of statistical significance attained indicates
the degree of uncertainty in taking the position that the
difference between the two groups is real. Classical
hypothesis testing has been largely criticized for such
errors in judgments (Cohen, 1994; Loftus 1993). Much
emphasis has been placed on researchers providing
indicators of effect, rather than just relying on p values,
specifically, providing the magnitude of the obtained
effect (e.g., a difference or relationship) as well as
confidence intervals associated with the statistical find-
ings. These additional statistics give consumers of
536 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
TABLE 21-1  Type I and Type II Errors
True Population Status Null Is True
Null Is False
research more information about the phenomenon
being studied (Cohen 1994).
Type I and Type II Errors
We choose the probability of making a Type I error
when we set alpha, and if we decrease the probability
of making a Type I error, we increase the probability
of making a Type II error. The relationships between
Type I and Type II errors are defined in Table 21-1.
Type II error occurs as a result of some degree of
overlap between the values of different populations,
so in some cases a value with a greater than 5% prob-
ability of being within one population may be within
the dimensions of another population.
It is impossible to decrease both types of error
simultaneously without a corresponding increase in
sample size. The researcher needs to decide which risk
poses the greatest threat within a specific study. In
nursing research, many studies are conducted with
small samples and instruments that lack precision and
accuracy in the measurement of study variables. Many
nursing situations include multiple variables that inter-
act to lead to differences within populations. However,
when one is examining only a few of the interacting
variables, small differences can be overlooked and
could lead to a false conclusion of no differences
between the samples. In this case, the risk of a Type
II error is a greater concern, and a more lenient level
of significance is in order. Nurse researchers usually
set the level of significance or α = 0.05 for their
studies versus a more stringent α = 0.01 or 0.001.
Setting α = 0.05 reduces the risk of a Type II error
of indicating study results are not significant when
they are.
Statistical Power
Power is the probability that a statistical test will
detect an effect when it actually exists. Power is the
inverse of Type II error and is calculated as 1 − β. Type
II error is the probability of retaining the null hypoth-
esis when it is in fact false. When the researcher sets
Type II error at 0.20 before conducting a study, this
Decision
Reject Null Accept Null
Type I Error Correct Decision
α 1−α
Correct Decision Type II Error
1−β β
means that the power of the planned statistic has been
set to 0.80. In other words, the statistic will have an
80% chance of detecting an effect if it actually exists.
Reported studies failing to reject the null hypoth-
esis (in which power is unlikely to have been exam-
ined) often have a low power level to detect an effect
if one exists. Until more recently, the researcher’s
primary interest was in preventing a Type I error.
Therefore, great emphasis was placed on the selection
of a level of significance, but little emphasis was
placed on power. This point of view is changing as we
recognize the seriousness of a Type II error in nursing
studies.
As stated in the steps of classical hypothesis testing
previously, step 4 is “conducting a power analysis.”
Power analysis involves determining the required
sample size needed to conduct your study after per-
forming steps 1, 2, and 3. Cohen (1988) identified four
parameters of power: (1) significance level, (2) sample
size, (3) effect size, and (4) power (standard of 0.80).
If three of the four are known, the fourth can be cal-
culated by using power analysis formulas. Signifi-
cance level and sample size are straightforward.
Chapter 15 provides a detailed discussion of determin-
ing sample size in quantitative studies that includes
power analysis. Effect size is “the degree to which the
phenomenon is present in the population or the degree
to which the null hypothesis is false” (Cohen, 1988,
pp. 9-10). For example, suppose you were measuring
changes in anxiety levels, measured first when the
patient is at home and then just before surgery. The
effect size would be large if you expected a great
change in anxiety. If you expected only a small change
in the level of anxiety, the effect size would be small.
Small effect sizes require larger samples to detect
these small differences (see Chapter 15 for a detailed
discussion of effect size). If the power is too low, it
may not be worthwhile conducting the study unless a
large sample can be obtained because statistical tests
are unlikely to detect differences or relationships that
exist. Deciding to conduct a study in these circum-
stances is costly in time and money, frequently does
 CHAPTER 21  Introduction to Statistical Analysis 537
TABLE 21-2  Software Applications for
Statistical Analysis
Software Application Website
NCSS (Number Cruncher www.ncss.com
Statistical System)
SPSS (Statistical Packages for www.spss.com
the Social Sciences)
SAS (Statistical Analysis System) www.sas.com
S+ spotfire.tibco.com
Stata www.stat.com
JMP www.jmp.com
not add to the body of nursing knowledge, and can
lead to false conclusions. Power analysis can be con-
ducted via hand calculations, computer software, or
online calculators and should be performed to deter-
mine the sample size necessary for a particular study
(Aberson, 2010). Power analysis can be calculated by
using the free power analysis software G*Power
(Faul, Erdfelder, Lang, & Buchner, 2007) or statisti-
cal software such as NCSS, SAS, and SPSS (Table
21-2). In addition, many free sample size calculators
are available online that are easy to use and under-
stand. If you have questions, you could consult a
statistician.
The power achieved should be reported with the
results of the studies, especially studies that fail to
reject the null hypothesis (have nonsignificant results).
If power is high, it strengthens the meaning of the
findings. If power is low, researchers need to address
this issue in the discussion of limitations and implica-
tions of the study findings. Modifications in the
research methodology that resulted from the use of
power analysis also need to be reported.
Statistical Significance versus
Clinical Importance
The findings of a study can be statistically significant
but may not be clinically important. For example, one
group of patients might have a body temperature 0.1°
F higher than that of another group. Data analysis
might indicate that the two groups are statistically
significantly different. However, the findings have
little or no clinical importance because of the small
difference in temperatures between groups. It is often
important to know the magnitude of the difference
between groups in studies. However, a statistical test
that indicates significant differences between groups
(e.g., a t-test) provides no information on the magni-
tude of the difference. The extent of the level of sig-
nificance (0.01 or 0.0001) tells you nothing about the
magnitude of the difference between the groups or the
relationship between two variables. The magnitude of
group differences can best be determined through cal-
culating effect sizes and confidence intervals (see
Chapters 22 through 25).
Inference
Statisticians use the terms inference and infer in a
similar way that a researcher uses the term generalize.
Inference requires the use of inductive reasoning. One
infers from a specific case to a general truth, from a
part to the whole, from the concrete to the abstract,
from the known to the unknown. When using inferen-
tial reasoning, you can never prove things; you can
never be certain. However, one of the reasons for the
rules that have been established with regard to statisti-
cal procedures is to increase the probability that infer-
ences are accurate. Inferences are made cautiously
and with great care. Researchers use inferences to
infer from the sample in their study to the larger
population.
Samples and Populations
Use of the terms statistic and parameter can be con-
fusing because of the various populations referred to
in statistical theory. A statistic, such as a mean (X), is
a numerical value obtained from a sample. A param-
eter is a true (but unknown) numerical characteristic
of a population. For example, µ is the population mean
or arithmetic average. The mean of the sampling dis-
tribution (mean of samples’ means) can also be shown
to be equal to µ. A numerical value that is the mean
(X) of the sample is a statistic; a numerical value that
is the mean of the population (µ) is a parameter
(Barnett, 1982).
Relating a statistic to a parameter requires an infer-
ence as one moves from the sample to the sampling
distribution and then from the sampling distribution to
the population. The population referred to is in one
sense real (concrete) and in another sense abstract.
These ideas are illustrated as follows:
Abstract
Infer Infer
Sample Sampling Population Infer
distribution
Concrete
For example, perhaps you are interested in the cho-
lesterol levels of women in the United States. Your
population is women in the United States. You cannot
measure the cholesterol level of every woman in the
United States; therefore, you select a sample of women
from this population. Because you wish your sample
538 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
to be as representative of the population as possible,
you obtain your sample by using random sampling
techniques (see Chapter 15). To determine whether the
cholesterol levels in your sample are similar to those
in the population, you must compare the sample with
the population. One strategy would be to compare the
mean of your sample with the mean of the entire
population. However, it is highly unlikely that you
know the mean of the entire population; you must
make an estimate of the mean of that population. You
need to know how good your sample statistics are as
estimators of the parameters of the population. First,
you make some assumptions. You assume that the
mean scores of cholesterol levels from multiple, ran-
domly selected samples of this population would be
normally distributed. This assumption implies another
assumption: that the cholesterol levels of the popula-
tion will be distributed according to the theoretical
normal curve—that difference scores and standard
deviations can be equated to those in the normal curve.
The normal curve is discussed later in this chapter.
If you assume that the population in your study is
normally distributed, you can also assume that this
population can be represented by a normal sampling
distribution. You infer from your sample to the sam-
pling distribution, the mathematically developed theo-
retical population made up of parameters such as the
mean of means and the standard error. The parameters
of this theoretical population are the measures of the
dimensions identified in the sampling distribution.
You can infer from the sampling distribution to the
population. You have both a concrete population and
an abstract population. The concrete population con-
sists of all the individuals who meet your study sample
criteria, whereas the abstract population consists of
individuals who will meet your sample criteria in the
future or the groups addressed theoretically by your
framework.
Types of Statistics
There are two major classes of statistics: descriptive
statistics and inferential statistics. Descriptive statis-
tics are computed to reveal characteristics of the
sample and to describe study variables. Inferential
statistics are computed to draw conclusions and make
inferences about the greater population, based on the
sample data set. The following sections define the
concepts and rationale associated with descriptive and
inferential statistics.
Descriptive Statistics
A basic yet important way to begin describing a
sample is to create a frequency distribution of the
variable or variables being studied. A frequency dis-
tribution is a plot of one variable, whereby the x-axis
consists of the possible values of that variable, and the
y-axis is the tally of each value. For example, if you
assessed a sample for a variable such as pain using a
visual analogue scale, and your subjects reported par-
ticular values for pain, you could create a frequency
distribution as illustrated in Figure 21-1.
Measures of Central Tendency
The measures of central tendency are descriptive sta-
tistics. The statistics that represent measures of central
tendency are the mean, median, and mode. All of these
statistics are representations or descriptions of the
center or middle of a frequency distribution. The
mean is the arithmetic average of all of the values of
a variable. The median is the exact middle value (or
the average of the middle two values if there is an
even number of observations). The mode is the most
commonly occurring value in a data set (Grove,
2007; Munro, 2005). It is possible to have more
than one mode in a sample, which is discussed
later in this chapter. In a normal curve, the mean,
median, and mode are equal or approximately equal
(see Figure 21-2).
Normal Curve
The theoretical normal curve is an expression of sta-
tistical theory. It is a theoretical frequency distribution
of all possible scores (see Figure 21-2). However, no
real distribution fits the normal curve exactly. The idea
of the normal curve was developed by an 18-year-old
mathematician, Gauss, in 1795, who found that data
measured repeatedly in many samples from the same
population by using scales based on an underlying
continuum can be combined into one large sample.
From this large sample, one can develop a more accu-
rate representation of the pattern of the curve in that
population than is possible with only one sample. In
most cases, the curve is similar, regardless of the spe-
cific data that have been examined or the population
being studied. This theoretical normal curve is sym-
metrical and unimodal and has continuous values. The
mean, median, and mode are equal. The distribution
is completely defined by the mean and standard devia-
tion, which are calculated and discussed further in
Chapter 22.
Sampling Distributions
The shape of the distribution provides important
information about the data. The outline of the distri­
bution shape is obtained by using a histogram. Within
this outline, the mean, median, mode, and standard
 CHAPTER 21  Introduction to Statistical Analysis 539

Frequency 5

0
1 2 3 4 5 6 7 8
Pain scores

Figure 21-1  Frequency distribution of visual analogue scale pain scores.

Mean
Median
Mode

2.28% 13.59% 34.13% 34.13% 13.59% 2.28%

Standard
deviation –2 s –1 s X +1 s +2 s

Figure 21-2  Normal curve.

deviation can be graphically illustrated (see Figure
21-2). This visual presentation of combined summary
statistics provides insight into the nature of the distri-
bution. As the sample size becomes larger, the shape
of the distribution more accurately reflects the shape
of the population from which the sample was taken.
Even when statistics, such as means, come from a
population with a skewed (asymmetrical) distribution,
the sampling distribution developed from multiple
means obtained from that skewed population tends to
fit the pattern of the normal curve. This phenomenon
is referred to as the central limit theorem.
Symmetry
Several terms are used to describe the shape of the
curve (and the nature of a particular distribution). The
shape of a curve is usually discussed in terms of
symmetry, skewness, modality, and kurtosis. A
540 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
symmetrical curve is one in which the left side is a
mirror image of the right side (see Figure 21-3). In
these curves, the mean, median, and mode are equal
and are the dividing point between the left and right
sides of the curve.
Skewness
Any curve that is not symmetrical is referred to as
skewed or asymmetrical. Skewness may be exhib-
ited in the curve in various ways. A curve may be
positively skewed, which means that the largest
portion of data is below the mean. For example, data
on length of enrollment in hospice are positively
skewed. Most people die within the first 3 weeks of
enrollment, whereas increasingly smaller numbers
survive as time increases. A curve can also be nega-
tively skewed, which means that the largest portion
of data is above the mean. For example, data on the
Mean
Median
Mode
Figure 21-3  Symmetrical curve.
Mode
Median
Mean
Positively skewed
Figure 21-4  Skewness.
occurrence of chronic illness by age in a population
are negatively skewed, with most chronic illnesses
occurring in older age groups. Figure 21-4 includes
both a positively skewed distribution and a negatively
skewed distribution.
In a skewed distribution, the mean, median, and
mode are not equal. Skewness interferes with the
validity of many statistical analyses; therefore, statisti-
cal procedures have been developed to measure the
skewness of the distribution of the sample being
studied. Few samples are perfectly symmetrical;
however, as the deviation from symmetry increases,
the seriousness of the impact on statistical analysis
increases. In a positively skewed distribution, the
mean is greater than the median, which is greater than
the mode. In a negatively skewed distribution, the
mean is less than the median, which is less than the
mode (see Figure 21-4).
Modality
Another characteristic of distributions is their modal-
ity. Most curves found in practice are unimodal,
which means that they have one mode, and frequen-
cies progressively decline as they move away from the
mode. Symmetrical distributions are usually uni-
modal. However, curves can also be bimodal (see
Figure 21-5) or multimodal. When you find a bimodal
Figure 21-5  Bimodal distribution.
Mode
Median
Mean
Negatively skewed
 CHAPTER 21  Introduction to Statistical Analysis 541
Leptokurtic Mesokurtic
Figure 21-6  Kurtosis.
sample, it usually means that you have not defined
your population adequately.
Kurtosis
Another term used to describe the shape of the distri-
bution curve is kurtosis. Kurtosis explains the degree
of peakedness of the curve, which is related to the
spread or variance of scores. An extremely peaked
curve is referred to as leptokurtic, an intermediate
degree of kurtosis is referred to as mesokurtic, and a
relatively flat curve is referred to as platykurtic (see
Figure 21-6). Extreme kurtosis can affect the validity
of statistical analysis because the scores have little
variation in a leptokurtic curve. Many computer pro-
grams analyze kurtosis before conducting statistical
analyses. A kurtosis of zero indicates that the curve is
mesokurtic. Kurtosis values above zero indicate that
the curve is leptokurtic, and values below zero that are
negative indicate a platykurtic curve (Box, Hunter, &
Hunter, 1978).
Tests of Normality
oasis-ebl|Rsalles|1476461060
Statistics are computed to obtain an indication of the
skewness and kurtosis of a given frequency distribu-
tion. The Shapiro-Wilk’s W test is a formal test of
normality that assesses whether the distribution of a
variable is skewed or kurtotic or both. This test has
the ability to calculate both skewness and kurtosis for
a study variable such as pain measured with a visual
analog scale. For large samples (n > 2000), the
Kolmogorov-Smirnov D test is an alternative test of
normality for large samples.
Variation
The range, standard deviation, and variance are statis-
tics that describe the extent to which the values in the
sample vary from one another. The most common of
Platykurtic
these statistics to be reported in the literature is the
standard deviation because of its direct association
with the normal curve. If the frequency distribution of
any given variable is approximately normal, knowing
the standard deviation of that variable allows us to
know what percentages of subjects’ values on that
variable fall between +1 and −1 standard deviation.
Referring back to the hypothetical frequency distribu-
tion of pain in Figure 21-1, when we calculate a stan-
dard deviation, we know that 34.13% of the subjects’
pain scores were between the mean pain score and 1
standard deviation above the mean pain score. We also
know that 34.13% of the subjects’ pain scores were
between the mean pain score and 1 standard deviation
below the mean. The middle 95.44% of the subjects’
scores were between −2 standard deviations and +2
standard deviations.
Confidence Intervals
When the probability of including the value of the
parameter within the interval estimate is known, this
is referred to as a confidence interval. Calculating a
confidence interval involves the use of two formulas
to identify the upper and lower ends of the interval
(see Chapter 22 for calculations). Confidence intervals
are usually expressed as “(38.6, 41.4),” with 38.6
being the lower end and 41.4 being the upper end of
the interval. Theoretically, we can produce a confi-
dence interval for any parameter of a distribution. It
is a generic statistical procedure. Confidence intervals
can also be developed around correlation coefficients
(Glass & Stanley, 1970). Estimation can be used for a
single population or for multiple populations. In esti-
mation, we are inferring the value of a parameter from
sample data and have no preconceived notion of the
value of the parameter. In contrast, in hypothesis
testing, we have an a priori theory about the value of
542 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
the parameter or parameters or some combination of
parameters. A formula is provided for calculating con-
ference intervals and example confidence intervals are
provided for different analysis results in Chapters 22
through 25.
Inferential Statistics
Inferential statistics are computed to draw conclusions
and make inferences about the greater population,
based on the sample data set. There are two classes of
inferential statistics: parametric and nonparametric
statistics.
Parametric Statistics
The most commonly used type of statistical analysis
is parametric statistics. The analysis is referred to as
parametric statistical analysis because the findings
are inferred to the parameters of a normally distributed
population. These approaches to analysis emerged
from the work of Fisher (1935) and require meeting
the following three assumptions before they can justi-
fiably be used.
1. The sample was drawn from a population for which
the variance can be calculated. The distribution is
usually expected to be normal or approximately
normal (Conover, 1971, Munro, 2005).
2. Because most parametric techniques deal with con-
tinuous variables rather than discrete variables, the
level of measurement should be at least interval
level data or ordinal data with an approximately
normal distribution.
3. The data can be treated as random samples (Box
et al., 1978).
Nonparametric Statistics
Nonparametric statistical analysis, or distribution-
free techniques, can be used in studies that do not meet
the first two assumptions of normal distribution and at
least interval level data. Most nonparametric tech-
niques are not as powerful as their parametric coun-
terparts (Tanizaki, 1997). In other words, nonparametric
techniques are less able to detect differences and have
a greater risk of a Type II error if the data meet the
assumptions of parametric procedures; this is gener-
ally because nonparametric statistics are actually per-
formed on ranks of the original data. When data have
been converted into ranks, they inevitably lose accu-
racy. Because nonparametric statistics have lower sta-
tistical power, many researchers choose to submit
ordinal data to parametric statistical procedures. If the
instrument or measurement procedure yielding ordinal
data has been rigorously evaluated, parametric statis-
tics are justified (Fife-Schaw, 1995). For example,
researchers often analyze data from a Likert scale with
strong reliability and validity as though they are inter-
val level data (see Chapter 17 for a description of
Likert scales).
Practical Aspects of
Data Analysis
Statistics can be used for a variety of purposes, such
as to (1) summarize, (2) explore the meaning of devia-
tions in the data, (3) compare or contrast descriptively,
(4) test the proposed relationships in a theoretical
model, (5) infer that the findings from the sample are
indicative of the entire population, (6) examine causal-
ity, (7) predict, or (8) infer from the sample to a
theoretical model. These different purposes for data
analysis are addressed in Chapters 22 through 25.
The process of quantitative data analysis consists
of several stages: (1) preparation of the data for analy-
sis; (2) description of the sample; (3) testing the reli-
ability of measurement; (4) exploratory analysis of the
data; (5) confirmatory analysis guided by the hypoth-
eses, questions, or objectives; and (6) post hoc analy-
sis. Statisticians such as Tukey (1977) divided the role
of statistics into two parts: exploratory data analysis
and confirmatory data analysis. You can perform
exploratory data analysis to obtain a preliminary
indication of the nature of the data and to search the
data for hidden structure or models. Confirmatory
data analysis involves traditional inferential statistics,
which you can use to make an inference about a popu-
lation or a process based on evidence from the study
sample.
Although not all of these six stages are reflected in
the final published report of the study, they all contrib-
ute to the insight you can gain from analyzing the data.
Many novice researchers do not plan the details of
data analysis until the data are collected and they are
confronted with the analysis task. This research tech-
nique is poor and often leads to the collection of unus-
able data or the failure to collect the data needed to
answer the research questions. Plans for data analysis
need to be made during development of the study
methodology. The following section covers the six
stages of quantitative data analysis.
Preparing the Data for Analysis
Except in very small studies, computers are almost
universally used for data analysis. Use of computers
has increased over the last decade with easy-to-use
data analysis packages becoming available for per-
sonal computers (PCs). When computers are used
 CHAPTER 21  Introduction to Statistical Analysis 543
for analysis, the first step of the process is entering
the data into the computer. Table 21-2 lists examples
of common statistical packages used for nursing
research.
Before entering data into the computer, the com-
puter file that will contain the data needs to be care-
fully prepared with information from the codebook as
described in Chapter 20. The location of each variable
in the computer file needs to be identified. Each vari-
able must be labeled in the computer so that the vari-
ables involved in a particular analysis are clearly
designated on the computer printouts. Develop a sys-
tematic plan for data entry that is designed to reduce
errors during the entry phase, and enter data during
periods when you have few interruptions.
In some cases, data must be reverse-scored before
initiating data analysis. Items in scales are often ar­
ranged so that sometimes a higher numbered response
indicates more of the construct being studied, and
sometimes a higher numbered response indicates less
of the construct being studied. This arrangement pre-
vents the subject from giving a global response to all
items in the scale. To reduce errors, the values on these
items need to be entered into the computer exactly as
they appear on the data collection form. Values on the
items are reversed by computer commands.
Cleaning the Data
Print the data file so that the data can be examined for
errors. When the size of the data file allows, you need
to cross-check every datum on the printout with the
original datum for accuracy. Otherwise, randomly
check the accuracy of data points. Correct all errors
found in the computer file. Perform a computer analy-
sis of the frequencies of each value of every variable
as a second check of the accuracy of the data. Search
for values outside the appropriate range of values for
that variable. Data that have been scanned into a com-
puter are less likely to have errors but should still be
checked.
Identifying Missing Data
Identify all missing data points. Determine whether
the information can be obtained and entered into the
data file. If a large number of subjects have missing
data on specific variables, you need to make a judg-
ment regarding the availability of sufficient data to
perform analysis with those variables. In some cases,
subjects must be excluded from the analysis because
of missing essential data. Missing data can also be
imputed via missing data statistical procedures. The
rules involving the appropriateness of missing data
imputations are complex, and there are many choices
of statistical applications. The seminal publication on
the subject of missing data imputation was written by
Rubin (1976).
Data Transformations
Skewed or non–normally distributed data that do not
meet the assumptions of parametric analysis can
sometimes be transformed in such a way that the
values are distributed closer to the normal curve.
Various mathematical operations are used for this
purpose. Examples of these operations include squar-
ing each value, calculating the square root of each
value, or calculating the logarithm of each value.
These operations can allow the researcher to yield a
frequency distribution that more closely approximates
normality, freeing the researcher to compute paramet-
ric statistics.
Data Calculations and Scoring
Sometimes a variable used in the analysis is not col-
lected but calculated from other variables and is
referred to as a calculated variable. For example, if
data are collected on the number of patients on a
nursing unit and on the number of nurses on a shift,
one might calculate a ratio of nurse to patient for a
particular shift. The data are more accurate if this
calculation is performed by computer rather than man-
ually. The results can be stored in the data file as a
variable rather than being recalculated each time
the variable is used in an analysis (Shortliffe &
Cimino, 2006).
Data Storage and Documentation
When the data-cleaning process is complete, backups
need to be made again; labeled as the complete,
cleaned data set; and carefully stored. Data cleaning
is a time-consuming process that you will not wish to
repeat unnecessarily. If your data are being stored on
a PC hard disk drive, be sure to back up the informa-
tion each time you enter more data. It is wise to keep
a second copy of the data filed at a separate, carefully
protected site. If your data are being stored on a
network, ensure that the network drive is being backed
up at least once a day. After data entry, you need to
store the original data in secure files for safekeeping.
The data files need to be secured as designated by
institutional review board policies. This usually
includes password protecting computer data files or
storing data on encrypted flash drives to which only
the research team has access.
Results of data analysis can easily become lost in
the mountain of printer paper. Rather than keep paper
printouts of statistical output, it is recommended that
544 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
you make pdf (portable document format) files of each
output file and store these files in the same folder as
your data sets and reports. There are many free pdf
converters available on the Internet for download. A
pdf converter allows you to convert any file into a pdf
file, which can be read by most computer operating
systems. Converting output files into pdf files allows
the researcher to transport those files and read
them on any computer, even a computer that does not
house the statistical software that created the original
output file.
All files, including data sets and output files, need
to be systematically named to allow easy access later
when theses or dissertations are being written or
research papers are being prepared for publication. We
recommend naming files by time sequence. Name the
file by its contents, and at the end of the file name,
identify the date (month, day, and year) that the file
was created or the analysis was performed. For
example, the files named “Rehab Outcomes Data
03-29-12” and “Means and Standard Deviations of
Pain Subscales 06-23-12” represent a data file saved
on March 29, 2012, and a statistical output file con-
taining means and standard deviations of subscale
scores saved on June 23, 2012.
When you are preparing papers that describe your
study, the results of each analysis reported in the paper
need to be cross-indexed with the output file for refer-
ence as needed. As interpretation of the results pro-
ceeds and you attempt to link various findings, you
may question some of the results. Selected results may
not seem to fit with the rest of the results, or they may
not seem logical. You may find that you have failed to
include necessary statistical information. When rewrit-
ing a paper for publication, you may need to report
additional results requested by reviewers. The search
for a particular output can be time-consuming and
frustrating if the printouts have not been carefully
organized. It is easy to lose needed results and have
to repeat the analysis.
Description of the Sample
After the data have been successfully entered in the
computer and stored, researchers start conducting the
essential analysis techniques for their studies. The first
step is to obtain as complete a picture as possible of
the sample. The demographic variables, such as age,
gender, and economic status, are analyzed with the
appropriate analysis techniques and used to develop
the characteristics of the sample. The analysis tech-
niques used in describing the sample are covered in
Chapter 22.
Testing the Reliability
of Measurement Methods
Examine the reliability of the methods of measure-
ment used in the study. The reliability of observational
measures or physiological measures may have been
obtained during the data collection phase, but it needs
to be noted at this point. Additional examination of the
reliability of measurement methods, such as a Likert
scale, is possible at this point. If you used a multiple-
item Likert scale in data collection, alpha coefficients
need to be calculated (Waltz, Strickland, & Lenz,
2010). The value of the coefficient needs to be com-
pared with values obtained for the instrument in previ-
ous studies. If the alpha coefficient is unacceptably
low (<0.6), you need to determine if you are justified
in performing analysis on data from the instrument
(see Chapter 16).
Exploratory Analysis of the Data
Examine all the data descriptively, with the intent of
becoming as familiar as possible with the nature of the
data. You might explore the data by conducting mea-
sures of central tendency and dispersion and examin-
ing outliers of the data. Neophyte researchers often
omit this step and jump immediately into the analyses
that were designed to test their hypotheses, questions,
or objectives. However, they omit this step at the risk
of missing important information in the data and per-
forming analyses that are inappropriate for the data.
The researcher needs to examine data on each variable
by using measures of central tendency and dispersion.
Are the data skewed or normally distributed? What is
the nature of the variation in the data? Are there outli-
ers with extreme values that appear different from the
rest of the sample that cause the distribution to be
skewed? The most valuable insights from a study
sometimes come from careful examination of outliers
(Tukey, 1977).
In many cases, as a part of exploratory analysis,
inferential statistical procedures are used to examine
differences and associations within the sample. From
an exploratory perspective, these analyses are relevant
only to the sample under study. There should be no
intent to infer to a population. If group comparisons
are made, effect sizes need to be determined for the
variables involved in the analyses.
In some nursing studies, the purpose of the study
is exploratory analysis. In such studies, it is often
found that sample sizes are small, power is low, mea-
surement methods have limited reliability and validity,
and the field of study is relatively new. If treatments
are tested, the procedure might be approached as a
 CHAPTER 21  Introduction to Statistical Analysis 545
pilot study. The most immediate need is tentative
exploration of the phenomena under study. Confirming
the findings of these studies requires more rigorously
designed studies with much larger samples. Many of
these exploratory studies are reported in the literature
as confirmatory studies, and attempts are made to
infer to larger populations. Because of the unaccept-
ably high risk of a Type II error in these studies,
negative findings should be viewed with caution.
Using Tables and Graphs for Exploratory Analysis
Although tables and graphs are commonly thought of
as a way of presenting the findings of a study, these
tools may be even more useful in helping the researcher
to become familiar with the data (see Figure 21-1 of
the frequency distribution of visual analogue scale
pain scores). Tables and graphs need to illustrate the
descriptive analyses being performed, even though
they will probably not be included in a research report.
These tables and figures are prepared for the sole
purpose of helping researchers to identify patterns in
their data and interpret exploratory findings, but they
are sometimes useful in reporting study results to
selected groups (Munro, 2005). Visualizing the data in
various ways can greatly increase insight regarding
the nature of the data (see Chapter 22).
Confirmatory Analysis
As the name implies, confirmatory analysis is per-
formed to confirm expectations regarding the data that
are expressed as hypotheses, questions, or objectives.
The findings are inferred from the sample to the popu-
lation. Thus, inferential statistical procedures are used.
The design of the study, the methods of measurement,
and the sample size must be sufficient for this confir-
matory process to be justified. A written analysis plan
needs to describe clearly the confirmatory analyses
that will be performed to examine each hypothesis,
question, or objective.
1. Identify the level of measurement of the data avail-
able for analysis with regard to the research objec-
tive, question, or hypothesis (see Chapter 16).
2. Select a statistical procedure or procedures appro-
priate for the level of measurement that will
respond to the objective, answer the question, or
test the hypothesis.
3. Select the level of significance that you will use
to interpret the results, which is usually α = 0.05.
4. Choose a one-tailed or two-tailed test if appropri-
ate to your analysis. The extremes of the normal
curve are referred to as tails. In a one-tailed test
of significance, the hypothesis is directional, and
the extreme statistical values that occur in a single
tail of the curve are of interest. In a two-tailed
test of significance, the hypothesis is nondirec-
tional or null, and the extreme statistical values in
both ends of the curve are of interest. Tailedness
is discussed in more detail in Chapter 25.
5. Determine the sample size available for the analy-
sis. If several groups will be used in the analysis,
identify the size of each group.
6. Evaluate the representativeness of the sample (see
Chapter 15).
7. Determine the risk of a Type II error in the analy-
sis by performing a power analysis.
8. Develop dummy tables and graphics to illustrate
the methods that you will use to display your
results in relation to your hypotheses, questions,
or objectives.
9. Determine the degrees of freedom for your analy-
sis. Degrees of freedom (df) involve the freedom
of a score’s value to vary given the other existing
scores’ values. The calculation of df varies based
on the analysis techniques conducted; Chapters
22 through 25 provide additional information and
examples of df.
10. Perform the analyses with a computer and rarely
manually.
11. Compare the statistical value obtained with the
table value by using the level of significance,
tailedness of the test, and df previously identified.
12. Most analyses are conducted by computer, and the
computer printout includes the statistical value
obtained by analyzing the data, p value, and df for
each inferential analysis technique.
13. Reexamine the analysis to ensure that the proce-
dure was performed with the appropriate vari-
ables and that the statistical procedure was
correctly specified in the computer program.
14. Interpret the results of the analysis in terms of the
hypothesis, question, or objective.
15. Interpret the results in terms of the framework.
Post Hoc Analysis
Post hoc analyses are commonly performed in studies
with more than two groups when the analysis indicates
that the groups are significantly different but does not
indicate which groups are different. For example, an
analysis of variance is conducted to examine the dif-
ferences among three groups—experimental group,
control group, and placebo group—and the groups are
found to be significantly different. A post hoc analysis
must be performed to determine which of the three
groups are significantly different. Post hoc analysis is
546 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
discussed in more detail in Chapter 25. In other
studies, the insights obtained through the planned
analyses generate further questions that can be exam-
ined with the available data.
Choosing Appropriate Statistical
Procedures for a Study
Multiple factors are involved in determining the suit-
ability of a statistical procedure for a particular study.
These factors can be related to the nature of the study,
the nature of the researcher, and the nature of statisti-
cal theory. Specific factors include (1) the purpose of
the study; (2) hypotheses, questions, or objectives; (3)
design; (4) level of measurement; (5) previous experi-
ence in statistical analysis; (6) statistical knowledge
level; (7) availability of statistical consultation; (8)
financial resources; and (9) access to computers. Use
items 1 to 4 to identify statistical procedures that meet
the requirements of the study, and narrow your options
further through the process of elimination based on
items 5 through 9.
The most important factor to examine when choos-
ing a statistical procedure is the study hypothesis. The
hypothesis that is clearly stated indicates the statistics
needed to test it. An example of a clearly developed
hypothesis is, “There is a difference in employment
rates between veterans who receive vocational reha-
bilitation and veterans who are on a wait-list control.”
This statement tells the researcher that a statistic to
determine differences between two groups is appropri-
ate to address this hypothesis.
One approach to selecting an appropriate statistical
procedure or judging the appropriateness of an analy-
sis technique is to use a decision tree. A decision tree
directs your choices by gradually narrowing your
options through the decisions you make. A decision
tree that can been helpful in selecting statistical pro-
cedures is presented in Figure 21-7.
One disadvantage of decision trees is that if you
make an incorrect or uninformed decision (guess), you
can be led down a path where you might select an
inappropriate statistical procedure for your study.
Decision trees are often constrained by space and do
not include all the information needed to make an
appropriate selection. A more extensive decision tree
can be found in A Guide for Selecting Statistical Tech-
niques for Analyzing Social Science Data by Andrews,
Klem, Davidson, O’Malley, and Rodgers (1981). The
following examples of questions designed to guide the
selection or evaluation of statistical procedures were
extracted from this book:
1. How many variables does the problem involve?
2. How do you want to treat the variables with
respect to the scale of measurement?
3. What do you want to know about the distribution
of the variable?
4. Do you want to treat outlying cases differently
from others?
5. What is the form of the distribution?
6. Is a distinction made between a dependent and an
independent variable?
7. Do you want to test whether the means of the two
variables are equal?
8. Do you want to treat the relationship between
variables as linear?
9. How many of the variables are dichotomous?
10. Is the dichotomous variable a collapsing of a con-
tinuous variable?
11. Do you want to treat the ranks of ordered catego-
ries as interval scales?
12. Do the variables have the same distribution?
13. Do you want to treat the ordinal variable as
though it were based on an underlying normally
distributed interval variable?
14. Is the interval variable dependent?
15. Do you want a measure of the strength of the
relationship between the variables or a test of the
statistical significance of differences between
groups (Kenny, 1979)?
16. Are you willing to assume that an interval-scaled
variable is normally distributed in the popula-
tion?
17. Is there more than one dependent variable?
18. Do you want to statistically remove the linear
effects of one or more covariates from the depen-
dent variable?
19. Do you want to treat the relationships among the
variables as additive?
20. Do you want to analyze patterns existing among
variables or among individual cases?
21. Do you want to find clusters of variables that are
more strongly related to one another than to the
remaining variables?
Each question confronts you with a decision. The
decision you make narrows the field of available sta-
tistical procedures (see Figure 21-7). Decisions must
be made regarding the following:
1. Number of variables (one, two, or more than two)
2. Level of measurement (nominal, ordinal, or
interval)
3. Type of variable (independent, dependent, or
research)
4. Distribution of variable (normal or non-normal)
5. Type of relationship (linear or nonlinear)
 Nature of Research Measurement Scale of Number of Groups Research Recommended
Question Dependent Variable (or Levels of Design Statistic
or Research Variable Independent Variable)

One-sample
One Group
chi square ( c 2)

Independent Fisher’s exact

samples test or c 2
Two Groups
Nominal
Paired McNemar
samples test

Independent
samples c2
More than
two groups
Paired Cochran’s Q
samples statistic

Independent Mann-
samples Whitney U
Two Groups
Paired Wilcoxon signed-
samples rank test
Differences
Ordinal
Independent Kruskal-
samples Wallis test
More than
two groups
Paired
Friedman test
samples

One-sample
One Group
t-test

Independent Independent
samples samples t-test
Two Groups

Interval/Ratio Paired Paired samples

samples t-test

One-way
Independent analysis
samples of variance
More than (ANOVA)
oasis-ebl|Rsalles|1476461066

two groups
Paired Repeated
measures
samples ANOVA

Phi, Contingency
Nominal coefficient,
Cramer’s V

Spearman rank-order
Associations Ordinal correlation, Kendall’s, tau,
Somer’s D

Pearson product-moment
Interval/Ratio correlation coefficient,
Simple linear regression

Figure 21-7  Statistical decision tree for selecting an appropriate analysis technique.
548 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
6. What you want to measure (strength of relationship
or difference between groups)
7. Nature of the groups (equal or unequal in size,
matched or unmatched, dependent [paired] or
independent)
8. Type of analysis (descriptive, classification, meth-
odological, relational, comparison, predicting out-
comes, intervention testing, causal modeling,
examining changes across time)
Selecting and evaluating statistical procedures
requires that you make many judgments regarding the
nature of the data and what you want to know. Knowl-
edge of the statistical procedures and their assump-
tions is necessary for selecting appropriate procedures.
You must weigh the advantages and disadvantages of
various statistical options. Access to a statistician can
be invaluable in selecting the appropriate procedures.
KEY POINTS
• This chapter introduces you to the concepts of
statistical theory and discusses some of the more
pragmatic aspects of quantitative data analysis,
including the purposes of statistical analysis, the
process of performing data analysis, the choice of
the appropriate statistical procedures for a study,
and resources for statistical analysis.
• Two types of errors can occur when making deci-
sions about the meaning of a value obtained from
a statistical test: Type I errors and Type II errors.
• A Type I error occurs when the researcher con-
cludes a significant effect when no significant
effect actually exists.
• A Type II error occurs when the researcher con-
cludes no significant effect when an effect actually
exists.
• The formal definition of the level of significance,
or alpha (α), is the probability of making a Type I
error when the null hypothesis is true.
• The p value is the exact value that can be calculated
during a statistical computation to indicate the
probability of making a Type I error.
• Power is the probability that a statistical test will
detect a significant effect when it actually exists.
• Statistics can be used for various purposes, such as
to (1) summarize, (2) explore the meaning of devia-
tions in the data, (3) compare or contrast descrip-
tively, (4) test the proposed relationships in a
theoretical model, (5) infer that the findings from
the sample are indicative of the entire population,
(6) examine causality, (7) predict, or (8) infer from
the sample to a theoretical model.
• The quantitative data analysis process consists of
several stages: (1) preparation of the data for analy-
sis; (2) description of the sample; (3) testing the
reliability of measurement; (4) exploratory analysis
of the data; (5) confirmatory analysis guided by
hypotheses, questions, or objectives; and (6) post
hoc analysis.
• A decision tree is provided to assist you in selecting
appropriate analysis techniques to use in analyzing
study or clinical data.
REFERENCES
Aberson, C. L. (2010). Applied power analysis for the behavioral
sciences. New York, NY: Routledge Taylor & Francis Group.
Andrews, F. M., Klem, L., Davidson, T. N., O’Malley, P. M., &
Rodgers, W. L. (1981). A guide for selecting statistical techniques
for analyzing social science data (2nd ed.). Ann Arbor, MI:
Survey Research Center, Institute for Social Research, University
of Michigan.
Barnett, V. (1982). Comparative statistical inference. New York,
NY: Wiley.
Box, G. E. P., Hunter, W. G., & Hunter, J. S. (1978). Statistics for
experimenters. New York, NY: Wiley.
Cohen, J. (1988). Statistical power analysis for the behavioral sci-
ences (2nd ed.). New York, NY: Academic Press.
Cohen, J. (1994). The earth is round (p<.05). American Psycholo-
gist, 49(12), 997–1003.
Conover, W. J. (1971). Practical nonparametric statistics. New
York, NY: Wiley.
Cox, D. R. (1958). Planning of experiments. New York, NY: Wiley.
Faul, F., Erdfelder, E., Lang, A.-G., & Buchner, A. (2007). G*Power
3: A flexible statistical power analysis program for the social,
behavioral, and biomedical sciences. Behavior Research Methods,
39(2), 175–191.
Fife-Schaw, C. (1995). Levels of measurement. In G. M. Breakwell,
S. Hammond, & C. Fife-Schaw (Eds.), Research methods in psy-
chology. Thousand Oaks, CA: Sage.
Fisher, R. A., Sir. (1935). The designs of experiments. New York,
NY: Hafner.
Glass, G. V. & Stanley, J. C. (1970). Statistical methods in education
and psychology. Englewood Cliffs, NJ: Prentice-Hall.
Good, I. J. (1983). Good thinking: The foundations of probability
and its applications. Minneapolis, MN: University of Minnesota
Press.
Grove, S. K. (2007). Statistics for health care research: A practical
workbook. St. Louis, MO: Saunders.
Kenny, D. A. (1979). Correlation and causality. New York, NY:
Wiley.
Kerlinger, F. N. & Lee, H. B. (2000). Foundations of behavioral
research (4th ed.). New York, NY: Harcourt Brace.
Loftus, G. R. (1993). A picture is worth a thousand p values: On
the irrelevance of hypothesis testing in the microcomputer age.
Behavior Research Methods, Instrumentation, & Computers,
25(2), 250–256.
Munro, B. H. (2005). Statistical methods for health care research
(5th ed.). Philadelphia, PA: Lippincott Williams & Wilkins.
 CHAPTER 21  Introduction to Statistical Analysis 549

Rubin, D. B. (1976). Inference and missing data. Biometrika, 63(3),
581–592.
Shortliffe, E. H. & Cimino, J. J. (2006). Biomedical informatics:
Computer applications in health care and biomedicine. New
York, NY: Springer Science.
Tanizaki, H. (1997). Power comparison of non-parametric tests:
Small-sample properties from Monte Carlo experiments. Journal
of Applied Statistics, 24(5), 603–632.
Tukey, J. W. (1977). Exploratory data analysis. Reading, MA:
Addison-Wesley.
Waltz, C. F., Strickland, O. L., & Lenz, E. R. (2010). Measurement
in nursing and health research (4th ed.). New York, NY: Springer
Publishing Company.
   http://evolve.elsevier.com/Grove/practice/

22
CHAPTER
 

Using Statistics to Describe Variables

T
here are two major classes of statistics: organizing the data by listing every possible value in
descriptive statistics and inferential statistics. the first column of numbers and the frequency (tally)
Descriptive statistics are computed to reveal of each value in the second column of numbers. For
characteristics of the sample data set. Inferential sta- example, consider the following hypothetical age data
tistics are computed to gain information about effects for patients from a primary care clinic. The ages of 20
in the population being studied. For some types of patients were:
studies, descriptive statistics are the only approach to
analysis of the data. For other studies, descriptive sta- 45, 26, 59, 51, 42, 28, 26, 32, 31, 55,
tistics are the first step in the data analysis process, 43, 47, 67, 39, 52, 48, 36, 42, 61, 57
to be followed by inferential statistics. For all studies
that involve numerical data, descriptive statistics are First, we must sort the patients’ ages from lowest to
crucial in understanding the fundamental properties of highest values:
the variables being studied. This chapter focuses on
26
descriptive statistics and includes the most common
26
descriptive statistics conducted in nursing research
28
with examples from clinical studies.
31
32
36
Using Statistics to 39
Summarize Data 42
42
Frequency Distributions 43
A basic yet important way to begin describing a 45
sample is to create a frequency distribution of the 47
variable or variables being studied. A frequency dis- 48
tribution can be displayed in a table or figure. A line 51
graph figure can be used to plot one variable, whereby 52
the x-axis consists of the possible values of that vari- 55
able, and the y-axis is the tally of each value. The 57
frequency distributions presented in this chapter 59
include values of continuous variables. With a con- 61
tinuous variable, the higher numbers represent more 67
of that variable, and the lower numbers represent less
of that variable. Common examples of continuous Next, each age value is tallied to create the frequency.
variables are age, income, blood pressure, weight, This is an example of an ungrouped frequency distri-
height, pain levels, and perception of quality of life. bution. In an ungrouped frequency distribution,
The frequency distribution of a variable can be researchers list all categories of the variable on which
presented in a frequency table, which is a way of they have data and tally each datum on the listing

550
 CHAPTER 22  Using Statistics to Describe Variables 551

TABLE 22-1  Grouped Frequency Distribution of Patient Ages with Percentages

Adult Age Range Frequency (f) Percentage Cumulative Percentage
20-29 3 15% 15%
30-39 4 20% 35%
40-49 6 30% 65%
50-59 5 25% 90%
60-69 2 10% 100%
Total 20 100%

(Corty, 2007). In this example, all the different 7

ages of the 20 patients are listed and then tallied for
6
each age.
Age Frequency 5

Frequency
26 2
4
28 1
31 1 3
32 1
36 1 2
39 1
1
42 2
43 1 0
45 1 20-29 30-39 40-49 50-59 60-69
47 1 Age Range
48 1
51 1 Figure 22-1  Frequency distribution of patient age ranges.
52 1
55 1
57 1
Table 22-1 presents a grouped frequency distribution
59 1
of patient ages classified by ranges of 10 years. The
61 1
67 1
range starts at “20” because there are no patient
ages lower than 20; also, there are no ages higher
Because most of the ages in this data set have than 69.
frequencies of “1,” it is better to group the ages into Table 22-1 also includes percentages of patients
ranges of values. These ranges must be mutually with an age in each range and the cumulative percent-
exclusive. A patient’s age can be classified into only ages for the sample, which should add to 100%. This
one of the ranges. In addition, the ranges must be table provides an example of a percentage distribu-
exhaustive, meaning that each patient’s age fits into tion that indicates the percentage of the sample with
at least one of the categories. For example, we may scores falling in a specific group or range (Corty,
choose to have ranges of 10, so that the age ranges 2007; Munro, 2005). Percentage distributions are par-
are 20 to 29, 30 to 39, 40 to 49, 50 to 59, and 60 to ticularly useful in comparing the data of the present
69. We may choose to have ranges of 5, so that the study with results from other studies.
age ranges are 20 to 24, 25 to 29, 30 to 34, and so As discussed earlier, frequency distributions can be
on. The grouping should be devised to provide the presented in figures. Frequencies are commonly pre-
greatest possible meaning to the purpose of the sented in graphs, charts, histograms, and frequency
study. If the data are to be compared with data in polygons. Figure 22-1 is a line graph of the frequency
other studies, groupings should be similar to group- distribution for age ranges, where the x-axis (horizon-
ings of other studies in this field of research. Classi- tal line) represents the different age ranges, and the
fying data into groups results in the development of y-axis (vertical line) represents the frequencies of
a grouped frequency distribution (Munro, 2005). patients with ages in each of the ranges.
552 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research

TABLE 22-2  Frequency Table of Smoking Status TABLE 22-3  Data of Medication Use in
Smoking Status Frequency Percentage (%) Veterans with Rheumatoid
Current smoker 1 10% Arthritis
Former smoker 6 60% Duration of Medication Use (years)
Never smoked 3 30%
0.1
Total 10 100%
0.3
1.3
1.5
1.5
2.0
2.2
70%
3.0
60% 3.0
50% 4.0
40%
30%
20% (MD), and mean (X) (Corty, 2007; Munro, 2005). The
10% mode, median, and mean are defined and calculated in
0% this section using data collected from veterans with
Current Never Former rheumatoid arthritis (Tran et al., 2009). The data were
Smoker Smoked Smoker extracted from a larger sample of veterans who had
a history of biologic medication use. Examples of
Figure 22-2  Histogram of smoking status. common biologic medications used to treat rheuma-
toid arthritis are adalimumab, etanercept, and inflix-
imab (Deighton, O’Mahony, Tosh, Turner, & Rudolf,
2009). Table 22-3 contains the data collected from 10
A frequency table is also an important method to veterans who had stopped taking biologic medica-
represent nominal data (Corty, 2007; Munro, 2005; tions, and the variable represents the number of years
Tukey, 1977). For example, a common nominal vari- that each veteran had taken the medication before the
able is smoking history. Many researchers assess sub- veteran stopped. Because the number of study subjects
jects’ history of smoking using nominal categories represented is 10, the correct statistical notation to
such as “never smoked,” “former smoker,” and reflect that number is:
“current smoker.” Table 22-2 presents frequency and
n = 10
percentage distributions for data extracted from a
sample of veterans with rheumatoid arthritis (Tran, The letter “n” is lowercase because we are referring
Hooker, Cipher, & Reimold, 2009). to a sample of veterans. If the data being presented
As shown in Table 22-2, the frequencies indicate represented the entire population of veterans, the
that 6 of 10 (60%) veterans were former smokers, and correct notation would be uppercase “N” (Zar, 1999).
3 (30%) never smoked. For nominal variables such as Because most nursing research is conducted using
smoking status, tables are a helpful method to inform samples, not populations, all formulas in Chapters 22
researchers and others about the variable being studied. to 25 incorporate the sample notation, n.
Graphically representing the values in a frequency
table can yield visually important trends. Figure 22-2 Mode
is a histogram that was developed to represent the The mode is the numerical value or score that occurs
smoking status data visually. with the greatest frequency in a data set. It does not
indicate the center of the data set. The data in Table
Measures of Central Tendency 22-3 contain two modes: 1.5 years and 3.0 years of
A measure of central tendency is a statistic that rep- medication use. Each of these numbers occurred twice
resents the center or middle of a frequency distribution in the data set. When two modes exist, the data set is
(Corty, 2007; Glass & Stanley, 1970; Grove, 2007). referred to as bimodal (see Chapter 21). A data set
The three measures of central tendency commonly that contains more than two modes is referred to as
reported in nursing studies include mode, median multimodal (Zar, 1999).
 CHAPTER 22  Using Statistics to Describe Variables 553

Median 3.0
The median (MD) is the score at the exact center of
2.5
the ungrouped frequency distribution. It is the 50th
percentile. To obtain the MD, sort the values from 2.0

Frequency
lowest to highest. If the number of values is an uneven
number, exactly 50% of the values are above the MD 1.5
and 50% are below it. If the number of values is an
1.0
even number, the MD is the average of the two middle
values; thus, the MD may not be an actual value in the 0.5
data set (Zar, 1999). For example, the data in Table
22-3 consist of 10 observations, and the MD is calcu- 0.0
lated as the average of the two middle values. 0-.9 1-1.9 2-2.9 3-3.9 4-4.9
A Years of Biologic Medication Use
(1.5 + 2.0)
MD = = 1.75 3.0
2
2.5
Mean
2.0

Frequency
The mean is the arithmetic average of all the values
of a variable in a study and is the most commonly 1.5
reported measure of central tendency. The mean is the 1.0
sum of the scores divided by the number of scores
being summed. Similar to the MD, the mean may not 0.5
be a member of the data set. The formula for calculat- 0.0
ing the mean is as follows:

1- 9
9
9

4- 9
9

6- 9
9

8- 9
9- 9
10 9.9
11 0.9

9
.
1.
2.
3.
4.
5.
6.
7.
8.

1.
0-

2-
3-

5-

7-

ΣX
X =
n
where
X = mean
Σ = sigma, the statistical symbol for summation
X = a single value in the sample
n = total number of values in the sample
The mean number of years that the veterans used a
biologic medication is calculated as follows:
(0.1 + 0.3 + 1.3 + 1.5 + 1.5
+ 2.0 + 2.2 + 3.0 + 3.0 + 4.0)
X =
10
18.9
X = = 1.89 years
10
The mean is an appropriate measure of central ten-
dency to calculate for approximately normally distrib-
uted populations with variables measured at the
interval or ratio levels. It is also appropriate for ordinal
level data such as Likert scale or rating scale values
(as described in Chapter 17), where higher numbers
represent more of the construct being measured, and
lower numbers represent less of the construct, such as
a 10-point pain rating scale, where 1 represents no
noticeable pain and 10 represents dramatically intense
pain (Clifford & Cipher, 2005).
-1
-1
B Years of Biologic Medication Use
Figure 22-3  A and B, Frequency distribution of years of biologic
medication use, without outlier and with outlier.
The mean is sensitive to extreme scores such as
outliers. An outlier is a value in a sample data set that
is unusually low or unusually high in the context of
the rest of the sample data (Zar, 1999). An example of
an outlier in the data presented in Table 22-3 might be
a value such as medication use for 11 years. The exist-
ing values range from 0.1 to 4.0, indicating that no
veteran used a biologic medication beyond 4 years. If
an additional veteran was added to the sample, and
that person used a biologic medication for 11 years,
the mean would be much larger: 2.72 years (mean =
18.9 + 11 = 29.9 ÷ 11 = 2.718 rounded to 2.72). Simply
adding this outlier to the sample nearly doubled the
mean value. The outlier would also change the fre-
quency distribution. Without the outlier, the frequency
distribution is approximately normal, as shown in
Figure 22-3. The inclusion of the outlier changes the
shape from an approximately normal distribution to a
positively skewed distribution (see Figure 22-3) (Zar,
1999). The median is a better measure of central ten-
dency than the mean for these data that are positively
 554 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
skewed by the outlier (see Chapter 21 for discussion
of skewness).
Using Statistics to Explore
Deviations in the Data
Although the use of summary statistics has been the
traditional approach to describing data or describing
the characteristics of the sample before inferential sta-
tistical analysis, the ability of summary statistics to
clarify the nature of data is limited. For example, using
measures of central tendency, particularly the mean,
to describe the nature of the data obscures the impact
of extreme values or deviations in the data. Significant
features in the data may be concealed or misrepre-
sented. Often, anomalous, unexpected, problematic
data and discrepant patterns are evident, but these are
not regarded as meaningful. Measures of dispersion,
such as the range, difference scores, variance, and
standard deviation, provide important insight into the
nature of the data.
Measures of Dispersion
Measures of dispersion or variability are measures
of individual differences of the members of the pop-
ulation and sample (Munro, 2005). They indicate
how values in a sample are dispersed around the
mean. These measures provide information about the
data that is not available from measures of central
tendency. They indicate how different the scores
are—the extent to which individual values deviate
from one another. If the individual values are
similar, measures of variability are small, and the
sample is relatively homogeneous in terms of those
values. When there are wide variations or differ-
ences in the scores, the sample is considered hetero-
geneous. The heterogeneity of sample scores or
oasis-ebl|Rsalles|1476461070
values is determined by measures of dispersion or
variability (Zar, 1999). The measures of dispersion
most commonly reported in nursing research are
range, difference scores, variance, and standard
deviation.
Range
The simplest measure of dispersion is the range
(Corty, 2007; Munro, 2005). In published studies,
range is presented in two ways: (1) the range is the
lowest and highest scores, or (2) the range is calcu-
lated by subtracting the lowest score from the highest
score. The range for the scores in Table 22-3 is 0.3 to
4.0 or can be calculated as follows: 4.0 − 0.3 = 3.7. In
this form, the range is a difference score that uses only
the two extreme scores for the comparison. The range
is generally reported in published studies but is not
used in further analyses.
Difference Scores
Difference scores are obtained by subtracting the
mean from each score. Sometimes a difference score
is referred to as a deviation score because it indicates
the extent to which a score deviates from the mean.
Most variables in nursing research are not “scores”;
however, the term difference score is used to represent
the deviation of a value from the mean. The difference
score is positive when the score is above the mean,
and it is negative when the score is below the mean.
The difference scores (both positive and negative) add
to zero or approximately zero based on rounding. Dif-
ference scores are the basis for many statistical analy-
ses and can be found within many statistical equations.
The formula for difference scores is:
X−X
The mean deviation is the average difference
score, using the absolute values. The formula for the
mean deviation is:
ΣX−X
X deviation =
n
1.8 + 1.6 + 0.6 + 0.4 + 0.4
+ 0.1 + 0.3 + 1.1 + 1.1 + 2.1
X deviation =
10
9.5
X deviation =
10
X deviation = 0.95
In this example using the data from Table 22-4, the
mean deviation is 0.95. The result indicates that, on
average, subjects’ duration of biologic medication use
deviated from the mean by 0.95 year.
Variance
Variance is another measure of dispersion commonly
used in statistical analysis. The equation for a sample
variance (s2) is provided. The lowercase letter “s2” is
used to represent a sample variance. The lowercase
Greek sigma “σ2” is used to represent a population
variance, in which the denominator is “N” instead of
“n − 1.” Because most nursing research is conducted
using samples, not populations, all formulas in the
next several chapters that contain a variance or stan-
dard deviation incorporate the sample notation and use
 CHAPTER 22  Using Statistics to Describe Variables 555

TABLE 22-4  Difference Scores of Duration of TABLE 22-5  Calculation of Variance for
Biologic Medication Use Duration of Medication Use
X −X X−X X X X−X (X − X)2
0.1 −1.9 −1.8 0.1 −1.9 −1.8 3.24
0.3 −1.9 −1.6 0.3 −1.9 −1.6 2.56
1.3 −1.9 −0.6 1.3 −1.9 −0.6 0.36
1.5 −1.9 −0.4 1.5 −1.9 −0.4 0.16
1.5 −1.9 −0.4 1.5 −1.9 −0.4 0.16
2.0 −1.9 0.1 2.0 −1.9 0.1 0.01
2.2 −1.9 0.3 2.2 −1.9 0.3 0.09
3.0 −1.9 1.1 3.0 −1.9 1.1 1.21
3.0 −1.9 1.1 3.0 −1.9 1.1 1.21
4.0 −1.9 2.1 4.0 −1.9 2.1 4.41
Σ of absolute values = 9.5 Σ 13.41

“n − 1” as the denominator. Statistical software pack- Table 22-5 displays the computations for the vari-
ages compute the variance and standard deviation ance. To compute the standard deviation, simply take
using the sample formulas, not the population the square root of the variance. You know that the
formulas. variance of biologic medication duration use is s2 =
1.49. Therefore, the standard deviation of biologic
duration is s = 1.22. In published studies, sometimes
Σ (X − X )
2

s2 = the statistic reported by researchers for standard devia-

n −1
The variance is always a positive value and has no
upper limit. In general, the larger the calculated vari-
ance for a study variable is, the larger the dispersion
or spread of scores is for the variable. The variance is
most often computed to derive the standard deviation
because in contrast to the variance, the standard devia-
tion reflects important properties about the frequency
distribution of the variable it represents. Table 22-4
displays how you might compute a variance by hand,
using the medication use duration data. Table 22-5
shows calculation of variance for duration of medica-
tion use.
13.41
s =
2
9 distribution of a variable that is perfectly normally
s2 = 1.49
Standard Deviation
Standard deviation (s) is a measure of dispersion that
is the square root of the variance. The equation for
obtaining a standard deviation is:
Σ (X − X )
2
tion is SD. Either SD or s might be used in a research
report to indicate the standard deviation for a study
variable.
The standard deviation is an important statistic,
both for understanding dispersion within a distribution
and for interpreting the relationship of a particular
value to the distribution. Grove’s (2007) statistical
workbook provides you with a resource for calculating
and interpreting the measures of central tendency and
measures of dispersion in published studies. The fol-
lowing section summarizes the properties of the stan-
dard deviation as it relates to a normal distribution.
Normal Curve
The standard deviation of a variable tells researchers
much about the entire sample of values. A frequency
distributed is shown in Figure 22-4, otherwise known
as the normal curve.
The normal curve is a perfectly symmetrical fre-
quency distribution. The value at the exact center of a
normal curve is the mean of the values. Note the lines
to the left and to the right of the mean. Those lines are
drawn at +1 standard deviation (which indicates 1 s
above the mean) and −1 standard deviation (which
indicates 1 s below the mean), +2 standard deviations
above the mean, −2 standard deviations below the

s= mean, and so forth. When a frequency distribution is

n −1 shaped like the normal curve, we know that 34.13%
556 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research

2.28% 13.59% 34.13% 34.13% 13.59% 2.28%

–2 s –1 s X +1 s +2 s

Figure 22-4  Normal curve. s = Standard deviation

2.28% 13.59% 34.13% 34.13% 13.59% 2.28%

30 40 50 60 70
X

Figure 22-5  Frequency distribution of SF-36 Physical Functioning Scale values.

of the subjects scored between the mean and 1 stan-
dard deviation above the mean, and 34.13% of the
subjects scored between the mean and 1 standard
deviation below the mean. Because the normal curve
is perfectly symmetrical, we also know that 50% of
the subjects scored above the mean, and 50% of the
subjects scored below the mean.
You can also say that 68.26% of the subjects scored
between −1 and +1 standard deviation. This number
is obtained by adding 34.13% and 34.13%. Further-
more, we can say that 95.44% of the subjects scored
between −2 and +2 standard deviations. If we are
given a mean and a standard deviation value for any
variable that is normally distributed, we know certain
facts about those data. For example, consider a score
obtained on a subscale of the Short Form (36) Health
survey (SF-36). The SF-36 is a widely used health
survey that yields eight subscales that each represent
a domain of subjective health status (Ware &
Sherbourne, 1992). The subscales have been normed
on populations of respondents as having a mean of 50
and standard deviation of 10. The frequency distribu-
tion of responses for the subscale “Physical Function-
ing” can be drawn as seen in Figure 22-5.
The mean is marked as “50” in the middle, and the
standard deviations are marked at the lines. Therefore,
you know that 34.13% of the population scores
between a 50 and a 60 on the Physical Functioning
subscale. You also know that 95.44% of the population
scores between 30 and 70 on the Physical Functioning
subscale. Figure 22-5 shows that only 2.28% of the
population scores above a value of 70 (this is com-
puted by subtracting 34.13% and 13.59% from 50%).
Likewise, only 2.28% of the population scores below
a value of 30.
When using examples such as these, researchers
often use the statistic “z” instead of the term “standard
deviation.” A z value is synonymous with a standard
 CHAPTER 22  Using Statistics to Describe Variables 557
deviation unit. A z value of 1.0 represents 1 standard
deviation unit above the mean. A z value of −1.0 rep-
resents 1 standard deviation unit below the mean.
Although a standard deviation value cannot have a
negative value, a z value can be negative or positive.
A z of 0 represents exactly the mean value. Any value
in a data set can be converted to a z by using the fol-
lowing formula:
z=
(X − X )
s 2.58 (see Figure 22-6). A table of z values can be
For example, a person scoring a 61 on the SF-36 Phy­
sical Functioning scale would have a z value of 1.1:
(61 − 50)
z=
10
z = 1.1
It is important to note how z values represent stan-
dard deviations on the normal curve because this
knowledge becomes necessary when performing sig-
nificance testing in inferential statistics. For example,
observe how a z value of 1.0 or −1.0 is much more
common than a z value of 3.0 or −3.0. The farther the
z value is from the mean, the more uncommon,
unusual, and unlikely that value is to occur. This prin-
ciple is revisited in Chapters 23 through 25.
–2.58 z –1.96 z –1 z
Unusual z values
Figure 22-6  Distribution of z values.
The distribution of the normal curve is drawn once
more in Figure 22-6 but this time with the z statistic,
where z represents 1 standard deviation unit. Common
values of z are smaller values and closer to the mean.
Uncommon and unusual z values are farther away
from the mean (either lower than the mean or higher
than the mean). When a variable is normally distrib-
uted, 95% of z values for that variable fall somewhere
between a z of −1.96 and 1.96; 99% of z values for
that variable fall somewhere between a z of −2.58 and
found in Appendix A.
Sampling Error
A standard error describes the extent of sampling
error. A standard error of the mean is calculated to
determine the magnitude of the variability associated
with the mean. A small standard error is an indication
that the sample mean is close to the population mean.
A large standard error yields less certainty that the
sample mean approximates the population mean. The
formula for the standard error of the mean (sX ) is:
s
sX =
n
where
sX = standard error of the mean
s = standard deviation
n = sample size
0 +1 z +1.96 z +2.58 z
Unusual z values
558 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
Using the biologic medication duration data, we
know that the standard deviation of biologic duration
is s = 1.22. Therefore, the standard error of the mean
for biologic duration is computed as follows:
1.22
sX =
10
sX = 0.39
The standard error of the mean for biologic duration
is 0.39.
Confidence Intervals
To determine how closely the sample mean approxi-
mates the population mean, the standard error of the
mean is used to build a confidence interval. A confi-
dence interval can be created for many statistics, such
as a mean, proportion, and odds ratio. To build a
confidence interval around a statistic, you must have
the standard error value and the t value to adjust
the standard error. The t is a statistic for the t-test
that is calculated to determine group differences
and is discussed in more detail in Chapter 25. The
degrees of freedom (df) to use to compute a confidence
interval is:
df = n − 1
To compute the confidence interval for a mean, the
lower and upper limits of that interval are created by
multiplying the sX by the t statistic, where df = n − 1.
For a 95% confidence interval, the t value should be
selected at alpha (α) = 0.05. For a 99% confidence
interval, the t value should be selected at α = 0.01.
Using the biologic medication duration data, we
know that the standard error of the mean duration of
biologic medication use is sX = 0.39. The mean dura-
tion of biologic medication use is 1.89. The 95% con-
fidence interval for the mean duration of biologic
medication use is computed as follows:
X ± sX t
1.89 ± (0.39)(2.26)
1.89 ± .88
As referenced in Appendix B, the t value required
for the 95% confidence interval with df = 9 is 2.26.
The previous computation results in a lower limit of
1.01 and an upper limit of 2.77. Thus, it is 95% prob-
able that the population mean is between 1.01 and
2.77. If we were to compute a 99% confidence inter-
val, we would require the t value that is referenced at
α = 0.01. The 99% confidence interval for the mean
duration of biologic medication use is computed as
follows:
1.89 ± (0.39)(3.25)
1.89 ± 1.27
As referenced in Appendix B, the t value required
for the 99% confidence interval with df = 9 is 3.25.
The previous computation results in a lower limit
of 0.62 and an upper limit of 3.16. Thus, it is 99%
probable that the population mean is between 0.62
and 3.16.
Degrees of Freedom
The concept of degrees of freedom was used in refer-
ence to computing a confidence interval. For any sta-
tistical computation, degrees of freedom is the
number of independent pieces of information that are
free to vary to estimate another piece of information
(Zar, 1999). In the case of the confidence interval, the
degrees of freedom (df) is n − 1. This means that there
are n − 1 independent observations in the sample that
are free to vary (to be any value) to estimate the lower
and upper limits of the confidence interval.
KEY POINTS
• Data analysis begins with descriptive statistics in
any study in which the data are numerical, includ-
ing demographic variables for samples in quantita-
tive and qualitative studies.
• Descriptive statistics allow the researcher to orga-
nize the data in ways that facilitate meaning and
insight.
• Three measures of central tendency are the mode,
median, and mean.
• The measures of dispersion most commonly
reported in nursing studies are range, difference
scores, variance, and standard deviation.
• The standard deviation and z represent certain
properties of the normal curve that are used in
significance testing.
• Standard error indicates the extent of sampling
error, and a formula is provided for calculation of
sampling error.
• To determine how closely the sample mean approx-
imates the population mean, the standard error of
the mean is used to build a confidence interval. The
 CHAPTER 22  Using Statistics to Describe Variables 559

formulas and calculations for the 95% and 99% Glass, G. V. & Stanley, J. C. (1970). Statistical methods in education
confidence intervals are presented. and psychology. Englewood Cliffs, NJ: Prentice-Hall.
• For any statistical computation, degrees of free­ Grove, S. K. (2007). Statistics for health care research: A practical
workbook. St. Louis, MO: Saunders Elsevier.
dom are the number of independent pieces of
Munro, B. H. (2005). Statistical methods for health care research
information that are free to vary to estimate another (5th ed.). Philadelphia, PA: Lippincott Williams & Wilkins.
piece of information. Tran, S., Hooker, R. S., Cipher, D. J., & Reimold, A. (2009). Pat-
terns of biologic use in inflammatory diseases: Older males.
Drugs & Aging, 26(7), 607–615.
REFERENCES Tukey, J. W. (1977). Exploratory data analysis. Reading, MA:
Addison-Wesley.
Clifford, P. A. & Cipher, D. J. (2005). The Geriatric Multidimen-
Ware, J. E. & Sherbourne, C. D. (1992). The MOS 36-Item Short-
sional Pain and Illness Inventory: A new instrument measuring
Form Health Survey (SF-36(r)): Conceptual framework and item
pain and illness in long-term care. Clinical Gerontologist, 28(3),
selection. Medical Care, 30(6), 473–483.
45–61.
Zar, J. H. (1999). Biostatistical analysis (4th ed.). Upper Saddle
Corty, E. W. (2007). Using and interpreting statistics: A practical
River, NJ: Prentice-Hall.
text for the health, behavioral, and social sciences. St. Louis, MO:
Mosby Elsevier.
Deighton, C., O’Mahony, R., Tosh, J., Turner, C., & Rudolf, M.
(2009). Guideline Development Group. Management of rheuma-
toid arthritis: Summary of NICE guidelines. BMJ, 338, b702.

23
CHAPTER
Using Statistics to Examine
Relationships
C
orrelational analyses identify relationships or
associations among variables. There are many
different kinds of statistics that yield a measure
of correlation. All of these statistics address a research
question or hypothesis that involves an association or
a relationship. Examples of research questions that are
answered with correlation statistics are as follows: “Is
there an association between weight loss and depres-
sion?” “Is there a relationship between patient satis-
faction and health status?” A hypothesis is developed
to identify the nature (positive or negative) of the
relationship between the variables being studied. For
example, a researcher may hypothesize that better
self-reported health is associated with higher levels of
patient satisfaction (Cipher & Hooker, 2006).
This chapter presents the common analysis tech-
niques used to examine relationships in studies.
The analysis techniques discussed include the use of
scatter diagrams before correlational analysis, bivari-
ate correlational analysis, testing the significance of a
correlational coefficient, spurious correlations, corre-
lations between two raters or measurements, the role
of correlation in understanding causality, and multi-
variate correlational procedure–factor analysis.
Scatter Diagrams
Scatter plots or scatter diagrams provide useful pre-
liminary information about the nature of the relation-
ship between variables (Corty, 2007; Munro, 2005).
The researcher should develop and examine scatter
diagrams before performing a correlational analysis.
Scatter plots may be useful for selecting appropriate
correlational procedures, but most correlational proce-
dures are useful for examining linear relationships
only. A scatter plot can easily identify nonlinear rela-
tionships; if the data are nonlinear, the researcher
should select statistical alternatives such as nonlinear
regression analysis (Zar, 1999). A scatter plot is created
560
  http://evolve.elsevier.com/Grove/practice/
by plotting the values of two variables on an x-axis
and y-axis. As shown in Figure 23-1, pain levels from
14 long-term care residents were plotted against their
number of behavioral disturbances (Cipher, Clifford,
& Roper, 2006). Specifically, each resident’s pair of
values (pain score, behavior value) was plotted on the
diagram. Pain was measured with the Geriatric Multi­
dimensional Pain and Illness Inventory (GMPI), and
behavioral disturbances were measured by the Geriat-
ric Level of Dysfunction Scale (Clifford, Cipher, &
Roper, 2005). The resulting scatter plot reveals a linear
trend whereby higher levels of pain tend to correspond
with higher behavioral disturbance values. The line
drawn in Figure 23-1 is a regression line that repre-
sents the concept of least-squares. A least-squares
regression line is a line drawn through a scatter plot
that represents the smallest deviation of each value
from the line (Cohen & Cohen, 1983). Regression
analysis is discussed in detail in Chapter 24.
Bivariate Correlational Analysis
Bivariate correlational analysis measures the mag-
nitude of linear relationship between two variables
and is performed on data collected from a single
sample (Munro, 2005). The particular correlation sta-
tistic that is computed depends on the scale of mea-
surement of each variable. Correlational techniques
are available for all levels of data: nominal (phi, con-
tingency coefficient, Cramer’s V, and lambda), ordinal
(Spearman rank order correlation coefficient, gamma,
Kendall’s tau, and Somers’ D), or interval and ratio
(Pearson’s product-moment correlation coefficient).
Figure 21-7 in Chapter 21 illustrates the level of mea-
surement for which each of these statistics is appropri-
ate. Many of the correlational techniques (Kendall’s
tau, contingency coefficient, phi, and Cramer’s V) are
used in conjunction with contingency tables, which
illustrate how values of one variable vary with values
 CHAPTER 23  Using Statistics to Examine Relationships 561

Number of Behavioral Disturbances 9 TABLE 23-1  Strength of Relationships

8 Strength of Positive Negative
7 Relationship Relationship Relationship
6 Weak relationship 0.00 to <0.30 0.00 to <−0.30
(y axis)

5 Moderate relationship 0.30 to 0.50 −0.30 to −0.50

4 Strong relationship >0.50 >−0.50
3
2
r = 0.0
1
0
0 2 4 6 8 10
Pain Subscale Score (x axis)

Figure 23-1  Scatter plot of pain and behavioral disturbances.

for a second variable. Contingency tables are explained
further in Chapter 25.
Correlational analysis provides two pieces of infor-
mation about the data: the nature or direction of the
linear relationship (positive or negative) between
the two variables and the magnitude (or strength) of
the linear relationship. Correlation statistics are not
an indication of causality, no matter how strong the
statistical result.
In a positive linear relationship, the values being
correlated vary together (in the same direction). When
one value is high, the other value tends to be high;
when one value is low, the other value tends to be low.
The relationship between weight and blood pressure
is considered positive because the more a patient
weighs, usually the higher his or her blood pressure.
In a negative linear relationship, when one value is
high, the other value tends to be low. There is a nega-
tive linear relationship between calories consumed
and weight loss because the more calories a person
consumes, the less his or her weight loss. A negative
linear relationship is sometimes referred to as an
inverse linear relationship—the terms negative and
inverse are synonymous in correlation statistics.
Sometimes the relationship between two variables
is curvilinear, which reflects a relationship between
the variables that changes over the range of both vari-
ables. For example, one of the most famous curvilin-
ear relationships is that of stress and test performance.
Test performance tends to be better as test-takers have
more stress but only up to a point. When students
experience very high stress levels, test performance
deteriorates (Lupien, Maheu, Tu, Fiocco, & Schramek,
2007; Yerkes & Dodson, 1908). Analyses designed to
test for linear relationships or associations between
two variables, such as Pearson’s correlation, cannot
detect a curvilinear relationship.
Variable Y on y axis
Variable X on the x axis
Figure 23-2  Scatter plot of r equal to approximately 0.00 representing
no relationship between two variables.
Pearson’s Product-Moment
Correlation Coefficient
Pearson’s product-moment correlation was the first
of the correlation measures developed and is the most
commonly used (Corty, 2007; Munro, 2005). This
coefficient (statistic) is represented by the letter r, and
the value of r is always between −1.00 and +1.00. A
value of zero indicates no relationship between the
two variables. A positive correlation indicates that
higher values of X are associated with higher values
of Y, and lower values of X are associated with lower
values of Y. A negative or inverse correlation indicates
that higher values of X are associated with lower
values of Y. The r value is indicative of the slope of
the line (called a regression line) that can be drawn
through a standard scatter plot of the values of two
paired variables. The strengths of different relation-
ships are identified in Table 23-1 (Aberson, 2010;
Cohen, 1988). Figure 23-2 represents an r value
approximately equal to zero, indicating no relation-
ship or association between the two variables. An r
value is rarely, if ever, equal to exactly zero. Figure
23-3 shows an r value equal to 0.50, which is a
 562 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research

r = 0.50
TABLE 23-2  Computation of Pearson’s
Correlation
Variable Y on the y axis

Patient X (Behaviors) Y (Pain) X2 Y2 XY

1 2 1.0 4 1.0 2.0
2 4 2.3 16 5.3 9.2
3 3 5.0 9 25.0 15.0
4 8 7.7 64 59.3 61.6
5 5 8.7 25 75.7 43.5
6 2 4.3 4 18.5 8.6
7 2 1.0 4 1.0 2.0
8 2 2.0 4 4.0 4.0
9 6 3.3 36 10.9 19.8
10 2 4.7 4 22.1 9.4
Variable X on the x axis
11 1 4.7 1 22.1 4.7
12 3 2.0 9 4.0 6.0
Figure 23-3  Scatter plot of variables where r is 0.50, representing a 13 3 3.0 9 9.0 9.0
moderate positive correlation.
14 6 5.3 36 28.1 31.8
Total 49 55 225 286 226.6
r = –0.50

words, for every value of X, the distribution of Y

Variable Y on the y axis

values should have equal variability. If the data for the

two variables being correlated are not homoscedastic,
inferences made during significance testing could be
invalid (Cohen & Cohen, 1983).

Calculation
Pearson’s product-moment correlation coefficient is
computed using one of several formulas; the following
formula is considered the “computational formula”
Variable X on the x axis
because it makes computation by hand easier (Zar,
Figure 23-4  Scatter plot of variables where r is −0.50, representing a 1999).
moderate inverse correlation. nΣXY − ΣXΣY
r=
[ nΣX − (ΣX )2 ][ nΣY 2 − (ΣY )2 ]
2

moderate positive relationship. Figure 23-4 shows an where

r value equal to −0.50, which is a moderate negative
or inverse relationship.
As discussed earlier, Pearson’s product-moment
correlation coefficient is used to determine the rela-
tionship between two variables measured at least at
the interval level of measurement. The formula for
Pearson’s correlation coefficient is based on the fol-
lowing assumptions:
1. Interval or ratio measurement of both variables
(e.g., age, income, blood pressure, cholesterol
levels)
2. Normal distribution of at least one variable
3. Independence of observational pairs
4. Homoscedasticity
Data that are homoscedastic are evenly dispersed
above and below the regression line, which indicates
a linear relationship on a scatter plot. Homoscedastic-
ity reflects equal variance of both variables. In other
r = Pearson’s correlation coefficient
n = total number of subjects
X = value of the first variable
Y = value of the second variable
XY = X multiplied by Y
Table 23-2 displays how one would set up data to
compute a Pearson’s correlation coefficient. This
example includes a subset of data from a study intro-
duced earlier in this chapter of long-term care resi-
dents (n = 14) with the most severe levels of dementia
(Cipher et al., 2006). A subset of data was selected for
this illustration so that the computation example
would be small and manageable. In actuality, studies
involving correlational analysis need to be adequately
powered and involve a larger sample than is used
here (Aberson, 2010; Cohen, 1988). However, all
data presented in this chapter are actual, unmodified
clinical data.
 CHAPTER 23  Using Statistics to Examine Relationships 563
The first variable in Table 23-2 is the number
of behavioral disturbances being exhibited by the
residents. The second variable is the residents’ scores
on the Pain and Suffering subscale of the GMPI
(Clifford & Cipher, 2005). The GMPI is a compre-
hensive pain assessment instrument that was designed
to be used by nurses, social workers, and psycholo-
gists to assess pain experienced by a person residing
in a long-term care facility. The GMPI Pain and Suf-
fering subscale is scored on a scale of 1 to 10, with
higher numbers indicating more severe pain. Higher
numbers are also indicative of more behavioral dis-
turbances. The null hypothesis being tested is: There
is no significant association between pain and behav-
ioral disturbances among long-term care residents
with severe dementia. The data in Table 23-2 are
arranged in columns, which correspond to the ele-
ments of the formula. The summed values in the last
row of Table 23-2 are inserted into the appropriate
place in the formula:
14(226.6) − (49)(55)
r=
[14(225) − 492 ][14(286) − 552 ]
14(226.6) − (49)(55)
r=
(749)(979)
477.4
r= = 0.56
856.3
Interpretation of Results
The r value is 0.56, indicating a strong positive rela-
tionship between pain and behavioral disturbances
among long-term care residents with severe demen-
tia. To determine whether this relationship is improb-
able to have been caused by chance alone, we consult
the r probability distribution table. The formula for
degrees of freedom (df) for a Pearson’s r is n − 2.
Recall from Chapter 22 that every inferential statistic
has its own formula for degrees of freedom (numbers
of values that are free to vary). In our analysis, the df
Middle 95% of
2½% of r distribution;
r distribution
Improbable values of r
–0.53
Figure 23-5  Probability distribution of r at df = 12.
is 14 − 2 = 12. With r of 0.56 and df = 12, you need
to consult the table in Appendix C to identify the
critical value of r. The critical r value at α = 0.05,
df = 12 is 0.532. Our obtained r was 0.56, which
exceeds the critical value in the table. Therefore, we
can conclude that: There was statistically significant
positive correlation between pain and behavioral dis-
turbances exhibited in long-term care residents with
severe dementia, r (12) = 0.56, p < 0.05. The null
hypothesis is rejected.
Every inferential statistic can be reflected by a
probability distribution of that statistic. The table we
referred to in Appendix C to determine the signifi-
cance of our obtained r was actually drawn from the
probability distribution of r. Chapter 22 illustrated the
probability distribution of z, which appears identical
to the normal curve. The Pearson’s r can be reflected
by a theoretical distribution of r values. The shape of
this distribution changes, depending on the size of the
sample. When a Pearson’s correlation is computed
using a large number of values (n > 120), the corre-
sponding distribution of r values appears similar to the
normal curve. The smaller the sample size, the flatter
the r distribution, and the larger the sample size, the
more the r distribution approximates the normal curve
reflecting the range of paired values obtained. Sample
size matters because the shape of the probability dis-
tribution determines whether our obtained statistic is
statistically significant (Zar, 1999).
For example, consider our obtained r of 0.56, pre-
viously calculated. At 12 df, the r probability distribu-
tion looks like that of Figure 23-5. With a sample size
of 14 (and 12 df), the middle 95% of the r probability
distribution is marked by −0.53 and 0.53. The mean r,
theoretically, is r = 0. That is, most correlation coef-
ficients computed between two variables equal zero,
reflecting no relationships between the two variables.
Therefore, an r value of 0 is the most common and
probable r value. It is much more improbable to obtain
a high r value. At 12 df, r values within the limits of
−0.53 and 0.53 are considered common and likely, and
2½% of r distribution;
Improbable values of r
0.53
 564 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
Middle 95% of
2½% of r distribution;
r distribution
Improbable values of r
–0.19
Figure 23-6  Probability distribution of r at df = 100.
values outside of these limits are uncommon, unlikely,
and improbable to have occurred by chance. The
values outside of these limits constitute 5% of the r
distribution, which is where the concept of alpha
(Type I error) originates. We obtained an r of 0.56 and
rejected the null hypothesis. In rejecting the null
hypothesis, there is less than a 5% chance that we are
making a Type I error.
Compare Figure 23-5 with Figure 23-6, in which
the probability distribution of r at df = 100 is dis-
played. Appendix C tells us that the critical r value at
alpha (α) = 0.05, df = 100 (and a sample size of 102)
is r = 0.19. This means that the middle 95% of the r
probability distribution at df = 100 is marked by −0.19
and 0.19. Furthermore, r values within the limits of
−0.19 and 0.19 are considered common and likely, and
values outside of these limits are uncommon, unlikely,
and improbable to have occurred by chance. Observe
the difference that the larger sample size makes in the
critical r value to achieve significance. The larger the
sample size, the smaller the obtained r value can be to
be considered still statistically significant.
oasis-ebl|Rsalles|1476461078
Effect Size
After establishing the statistical significance of r, the
relationship subsequently must be examined for clini-
cal importance. There are ranges for strength of asso-
ciation suggested by Cohen (1988), as displayed in
Table 23-1. One can also assess the magnitude of
association by obtaining the coefficient of determina-
tion for the Pearson’s correlation. Computing the
coefficient of determination simply involves squaring
the r value. The r2 (multiplied by 100%) represents the
percentage of variance shared between the two vari-
ables (Cohen & Cohen, 1983). In our example, r was
0.56, and therefore r2 was 0.31; this indicates that pain
levels and behavioral disturbance shared 31% of the
2½% of r distribution;
Improbable values of r
0.19
same variance. More specifically, 31% of the variance
in pain levels can be explained by knowing the resi-
dent’s level of behavioral disturbances, and, vice
versa, 31% of the variance in behavioral disturbances
can be explained by knowing the resident’s pain level
(Cipher et al., 2006).
Nonparametric Alternatives
If one or both of your variables do not meet the
assumptions for a Pearson’s correlation, or if your
variables are scaled on an ordinal scale of measure-
ment (rather than interval or ratio), both Spearman’s
rank-order correlation coefficient and Kendall’s tau
are more appropriate statistics. The Spearman’s rank-
order correlation coefficient and Kendall’s tau calcula-
tions involve converting the data to ranks, discarding
any variance or normality issues associated with the
original values.
If your data meet the assumptions for the Pearson’s
correlation coefficient, it is the preferred analysis pro-
cedure. You would calculate a nonparametric alterna-
tive only if your data violate those assumptions.
Because Spearman’s correlation and Kendall’s tau are
based on ranks of the data, the properties of the origi-
nal data are lost when they are converted to ranks.
Because of this fact, most nonparametric statistics of
association yield lower statistical power (Daniel,
2000). Grove’s (2007) statistical workbook provides
examples of Spearman’s rank-order correlation coef-
ficient from published studies and provides guidance
in the interpretation of these results.
If both of your variables are dichotomous, the phi
coefficient is the appropriate statistic to determine an
association. If both of your variables are nominal and
one or both has more than two categories, Cramer’s V
statistic is the appropriate statistic. Spearman’s rank-
order correlation coefficient, Kendall’s tau, phi, and
 CHAPTER 23  Using Statistics to Examine Relationships 565
Cramer’s V are addressed in detail by Daniel (2000)
and Munro (2005).
Role of Correlation
in Understanding Causality
In any situation involving causality, a relationship
exists between the factors involved in the causal
process. Therefore, the first clue to the possibility
of a causal link is the existence of a relationship.
However, a relationship does not mean causality. For
example, blood glucose level may be related to or
correlated with body temperature; however, this does
not mean that one causes the other. Two variables can
be highly correlated but have no causal relationship.
However, as the strength of a relationship increases,
the possibility of a causal link increases. The absence
of a relationship precludes the possibility of a causal
connection between the two variables being exam-
ined, given adequate measurement of the variables
and absence of other variables that might mask the
relationship (Cohen & Cohen, 1983). A correlational
study can be the first step in determining the dynamics
important to nursing practice within a particular popu-
lation. Determining these dynamics can allow us to
increase our ability to predict and control the situation
studied. However, correlation cannot be used to show
causality.
Spurious Correlations
Spurious correlations are relationships between
variables that are not logical. In some cases, these
significant relationships are a consequence of chance
and have no meaning. When you choose a level of
significance of α = 0.05, 1 in 20 correlations that you
compute will be statistically significant by chance
alone. There is really no true relationship between
the two variables under study in the population; you
just happened to draw a sample that showed a rela-
tionship where there typically is not one. Other pairs
of variables may be correlated because of the influ-
ence of other unrelated or confounding variables.
For example, you might find a positive correlation
between the number of deaths on a nursing unit and
the number of nurses working on the unit. The
number of deaths cannot be explained as occurring
because of increases in the number of nurses. It is
more likely that a third variable (units having patients
with more critical conditions) explains both the
increased number of nurses and the increased number
of deaths. In many cases, the “other” variable remains
unknown, although the researcher can use reasoning
to identify and exclude most of these spurious
correlations.
Bland and Altman Plots
Bland and Altman plots are used to examine the
extent of agreement between two measurement tech-
niques. In nursing research, Bland and Altman plots
are used to display visually the extent of interrater
agreement and test-retest agreement (see Chapter 16
for discussion of reliability). For both instances, pairs
of data are collected from each subject (from rater 1
and rater 2, or administration 1 and administration 2),
and each subject’s two values are subtracted from one
another. The differences are plotted on a graph, dis-
playing a scatter diagram of the differences plotted
against the averages. Limits of agreement are defined
as twice the standard deviation above and below the
mean. Bland and Altman plots are primarily used to
see how many of the values are outside these limits.
Acceptable interrater or test-retest agreement is con-
sidered to be reflected when at least 95% of the values
are within the limits of agreement on the plot (Altman,
1991).
Example
Table 23-3 displays test-retest data obtained from the
Clifford and Cipher (2005) study introduced earlier.
The GMPI was administered to 22 long-term care
residents twice within 48 hours. The score from the
GMPI subscale Activity Interference was analyzed for
test-retest agreement. This subscale is indicative of the
level of functional impairment secondary to pain,
rated on a scale of 1 to 10, with 1 representing no
impairment and 10 representing extreme impairment
(Clifford & Cipher, 2005). Each subject’s subscale
score at Assessment 1 and Assessment 2 is displayed
in Table 23-3, along with the difference between each
pair of scores.
A Bland and Altman plot of these data is illustrated
in Figure 23-7. The line of perfect agreement is drawn
as a dotted line in the exact horizontal middle of the
graph, and the limits of agreement are the two outside
dotted lines. Only one value is outside of the limits of
agreement. Therefore, 21 of the 22 pairs (95.5%) were
within the limits of agreement. Incidentally, the r
between the first and second administrations of the
subscale was 0.96. However, the Bland and Altman
plot does not always corroborate a Pearson’s correla-
tion coefficient, and vice versa, because they are dis-
tinctly different methods (Bland & Altman, 1986).
The coefficient of repeatability was created by
Bland and Altman (1986) as an indication of the
repeatability of a single method of measurement.
Because the same method is being measured repeat-
edly, the mean difference should be zero. Use the
566 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research

following formula to calculate a coefficient of repeat- Table 23-3 displays each difference score, of which
ability (CR), where sx1 − x2 is the standard deviation of sx1 − x2 = 0.42. Therefore, the CR is calculated as:
the difference scores. CR = 1.96 (0.42)
CR = 1.96 (sx1 − x2 ) CR = 0.82

Interpretation of Results
TABLE 23-3  Test-Retest Data from the Geriatric In their study, Clifford and Cipher (2005) found the
Multidimensional Pain and   CR value was 0.82. This value is used to compute
Illness Inventory Activity lower and upper limits of agreement (Bland & Altman,
Interference Subscale 1986). The mean difference between the two assess-
ments of Activity Interference was 0.10. In other
Activity Activity words, the average difference between the first and
Interference, Interference, Difference second administrations of this subscale was 0.10. A
Time 1 Time 2 Score perfect average agreement would be 0, meaning that
2.0 2.8 0.8 on average the two sets of scores were exactly the
3.0 3.6 0.6
same. The CR value, 0.82, is added to and subtracted
3.8 4.0 0.2
3.6 3.8 0.2
from the mean difference to create lower and upper
3.2 3.2 0.0 limits of acceptable agreement: 0.10 ± 0.82. Differ-
4.4 4.6 0.2 ences within 0.10 ± 0.82 would not be deemed clini-
3.6 3.0 −0.6 cally important, according to Bland and Altman
2.8 2.6 −0.2 (2010). Differences between the two administrations
3.4 3.4 0.0 that are less than −0.72 and greater than 0.92 are
5.8 5.6 −0.2 “unacceptable for clinical purposes” (Bland & Altman,
6.8 7.2 0.4 2010). The CR is not an inferential statistic, and values
3.0 3.6 0.6 of lower and upper limits of agreement are not inter-
5.0 5.6 0.6
preted the way one would interpret a confidence inter-
4.8 4.0 −0.8
3.4 3.2 −0.2
val. Rather, they are formulas invented by Bland and
6.6 6.0 −0.6 Altman for heuristic purposes to make decisions on
3.0 3.0 0.0 the extent of agreement between two measurements.
5.8 6.2 0.4
4.8 5.2 0.4
7.0
5.8
7.2
6.0
0.2
0.2
Factor Analysis and Principal
6.0 6.0 0.0 Components Analysis
Data from Clifford, P. A., & Cipher, D. J. (2005). The Geriatric
Factor analysis and principal components analysis
Multidimensional Pain and Illness Inventory: A new instrument measuring (PCA) are statistical techniques designed to examine
pain and illness in long-term care. Clinical Gerontologist, 28(3), 49. interrelationships among large numbers of variables to

4
3
2
Upper limit of agreement
1
0 Perfect agreement
–1 Lower limit of agreement
–2
–3
One value outside limit of
–4 agreement
2 3 4 5 6 7 8

Figure 23-7  Bland and Altman plot of test-retest data.

 CHAPTER 23  Using Statistics to Examine Relationships 567

reduce them to a smaller set of variables and to iden- TABLE 23-4  Item Factor Loadings on Three
tify clusters of variables that are most closely linked
GMPI Subscales
together (factors). Factors are hypothetical constructs
created from the original variables. The difference Factor Loading GMPI Item
between factor analysis and PCA lies in the method in Scale 1: Pain and
which the variance of the data is extracted and ana- Suffering
lyzed (Tabachnick & Fidell, 2006). PCA is the proce- .88 Current level of pain
dure of choice for a researcher who is primarily .86 Level of pain in the last week
.67 Level of suffering in the past week
interested in reducing a large number of variables
Scale 2: Activity
down to a smaller number of components. Interference
A common reason for performing PCA is to assist .67 Pain’s interference with walking
with validity investigations of a new measurement .60 Pain’s interference with sitting up
method or scale, particularly subjective assessments .84 Pain’s interference with leaving
or instruments that pertain to attitudes, beliefs, values, room
or opinions. When researchers develop a new instru- .91 Pain’s interference with social
ment, PCA can serve to assist the researcher in inves- activities
tigating its content and construct validity, as described .75 Pain’s interference with
in Chapter 16. The results of PCA assist researchers satisfaction/enjoyment
Scale 3: Emotional
in understanding which questions are redundant (or
Distress
assess the same concept), which questions represent .85 Lonely because of pain
subsets of variables, and which items stand alone and .63 Irritable due to pain
reflect unique concepts. .69 Anxious due to pain
Mathematically, PCA extracts maximum variance .73 Coping with problems
(explanatory “power” to predict one variable’s value
from another’s value) from the data set with each Clifford, P. A., & Cipher, D. J. (2005). The Geriatric Multidimensional Pain
and Illness Inventory: A new instrument measuring pain and illness in
“component” (often used interchangeably with long-term care. Clinical Gerontologist, 28(3), 51.
“factor”). The first principal component is the linear
combination of the variables (or instrument items) that
maximizes the variance of their component scores.
The second component is formed from residual cor- by individuals in long-term care facilities. Of the
relations. Subsequent components are formed from 14 items, 12 items were retained to be included in
the residual correlations that have not yet been created. the final version of the GMPI scale. Using factor anal-
Once the factors have been identified mathemati- ysis, three factors were identified and are listed in
cally, the researcher attempts to explain why the vari- Table 23-4.
ables are grouped as they are. Factor analysis aids in The first factor accounted for 40% of the variance
the identification of theoretical constructs. Factor extracted from the PCA solution, followed by 16%
analysis is also used to confirm the accuracy of a theo- and 9% from the second and third factors. Table 23-4
retically developed construct. For example, a theorist lists the factor loadings of each item. Factor loadings
might state that the concept “hope” consists of the are the correlations between the item and the new
elements (1) anticipation of the future, (2) belief that factor. The first factor, named Pain and Suffering by
things will work out for the best, and (3) optimism. the researchers, was correlated with three of the GMPI
Instruments could be developed to measure these three items—current level of pain, level of pain in the last
elements, and factor analysis could be conducted to week, and level of suffering. All other GMPI items
determine whether subject responses clustered into were not highly correlated with factor 1 (the factor
these three groupings. loadings were < 0.30) and are not listed in Table 23-4.
The second factor, named Activity Interference, was
Example correlated with five of the GMPI items, all of which
The following example describes how PCA was used pertained to life interference secondary to the pain.
to investigate content and construct validity for the The factor loadings ranged from 0.60 to 0.91. The
GMPI (Clifford & Cipher, 2005). There were 14 items third factor, named Emotional Distress, was correlated
from the GMPI administered to 401 long-term care with four of the GMPI items, all of which pertained
residents. GMPI, a scale introduced earlier in this to emotional distress secondary to the pain. The factor
chapter, was developed to assess pain experienced loadings ranged from 0.63 to 0.85.
568 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
In this analysis, the items loaded highly on only
one factor, which is not always the case in PCA. For
example, the first GMPI item was “How bad is your
pain right now?” This item loaded highly on factor 1
(Pain and Suffering) but had low loadings of less than
0.30 on the other two factors. The fourth GMPI item
was “How much has your pain interfered with you
being able to walk around?” This item loaded highly
on factor 2 (Activity Interference) but had low load-
ings of less than 0.30 on the other two factors. This
was the case for all GMPI items, with the exception
of two items that did not load highly on any of the
factors. Those two items were permanently dropped
from the instrument.
“Naming” the Factor
The three factors generated from the PCA were named
according to the nature of the items that loaded on
those factors. When naming the factor, the researcher
must examine the items that cluster together in a factor
and seem to explain that clustering. Variables with
high loadings on the factor must be included, even if
they do not fit the researcher’s preconceived theoreti-
cal notions of which items “fit” together because they
reflect a similar concept. The purpose is to identify the
broad construct of meaning that has caused these par-
ticular variables to be so strongly intercorrelated.
Naming this construct is an important part of the pro-
cedure because naming of the factor provides theoreti-
cal meaning.
Factor Scores
After the initial factor analysis, additional studies are
conducted to examine changes in the phenomenon
in various situations and to determine the relation-
ships of the factors with other concepts. Factor
scores are used during data analysis in these addi-
tional studies. To obtain factor scores, the variables
included in the factor are identified, and the scores
on these variables are summed for each study par-
ticipant. Thus, each participant has a score for each
factor in the instrument. There are several methods
of computing factor scores. One of the most common
methods involves simply adding the participant’s
scores on the items that load on a factor. Using the
GMPI data as an example, to obtain a factor score
for Pain and Suffering, a study participant’s scores
on current level of pain, pain in the past week, and
level of suffering would be summed. For example,
if a participant scored an 8 on current pain, 7 on
pain the past week, and 9 on level of suffering, the
participant’s factor score for Pain and Suffering
would equal 24.
Individual’s factor score for pain
and suffering = 8 + 7 + 9 = 24
Another common method of computing a factor
score is using the factor loadings to weight each study
participant’s score. Applying the same hypothetical
scores as before, the factor loadings are multiplied by
the item scores to create the factor score:
(0.88) 8 + (0.86) 7 + (0.67) 9 = 19.09
In the first method, each item is weighted equally
in the equation because the weight is essentially “1.”
In the second method, each item is adjusted for the
extent to which that item loads on that factor. The
advantages and disadvantages of these factor score
methods, in addition to descriptions of other methods
for obtaining factor scores, are reviewed by DiSte-
fano, Zhu, & and Mîndrilă (2009).
KEY POINTS
• Correlational analyses identify relationships or
associations between or among variables.
• The purpose of the analysis is also to clarify rela-
tionships among theoretical concepts or help to
identify potentially causal relationships, which can
be tested by inferential analysis.
• All data for the analysis should have been obtained
from a single population from which values are
available on all variables to be examined.
• Correlational analysis provides two pieces of infor-
mation about the data: the nature of a linear rela-
tionship (positive or negative) between the two
variables and the magnitude (or strength) of the
linear relationship.
• Pearson’s product-moment correlation coefficient
is the preferred computation when investigating the
association among two variables measured at the
interval or ratio level and when the variables meet
the other required statistical assumptions.
• Spearman’s rank-order correlation coefficient and
Kendall’s tau are both nonparametric statistics that
are calculated when the assumptions of Pearson’s
correlation cannot be met, such as variables mea-
sured at the ordinal level or non-normal distribution
of values for a variable.
• The first clue to the possibility of a causal link is
the existence of a relationship, but a relationship
does not mean causality.
• Bland and Altman plots are a graphical display of
agreement between two administrations of an
 CHAPTER 23  Using Statistics to Examine Relationships 569
instrument or two raters of a clinician-rated
instrument.
• The coefficient of repeatability is a value that is
used to determine acceptable lower and upper
limits of interrater agreement and test-retest agree-
ment.
• Principal components analysis is a procedure that
reduces a large number of variables down to a
smaller number of components and is most often
used during the construction of a new measurement
method or scale.
• The results of principal components analysis assist
the researcher in understanding which questions
assess the same concept and are redundant, which
questions represent subsets of variables, and which
items stand alone.
REFERENCES
Aberson, C. L. (2010). Applied power analysis for the behavioral
sciences. New York, NY: Routledge Taylor & Francis Group.
Altman, D. G. (1991). Practical statistics for medical research (1st
ed.). London, UK: Chapman & Hall.
Bland, J. M., & Altman, D. M. (1986). Statistical methods for
assessing agreement between two methods of clinical measure-
ment. Lancet, 1(8476), 307–310.
Bland, J. M., & Altman, D. M. (2010). Statistical methods for
assessing agreement between two methods of clinical measure-
ment. International Journal of Nursing Studies, 47(8), 931–936.
Cipher, D. J., Clifford, P. A., & Roper, K. D. (2006). Behavioral
manifestations of pain in the demented elderly. Journal of the
American Medical Director’s Association, 7(6), 355–365.
Cipher, D. J., & Hooker, R. S. (2006). Are patients satisfied with
PAs and NPs? Journal of the American Association of Physician
Assistants, 19(1), 36–44.
Clifford, P. A., & Cipher, D. J. (2005). The Geriatric Multidimen-
sional Pain and Illness Inventory: A new instrument measuring
pain and illness in long-term care. Clinical Gerontologist, 28(3),
45–61.
Clifford, P. A., Cipher, D. J., & Roper, K. D. (2005). Assessing
dysfunctional behaviors in long-term care. Journal of the Ameri-
can Medical Director’s Association, 6(5), 300–309.
Cohen, J. (1988). Statistical power analysis for the behavioral sci-
ences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum Associates.
Cohen, J., & Cohen, P. (1983). Applied multiple regression/
correlation analysis for the behavioral sciences (2nd ed.). Hills-
dale, NJ: Erlbaum.
Corty, E. W. (2007). Using and interpreting statistics: A practical
text for the health, behavioral, and social sciences. St. Louis, MO:
Mosby.
Daniel, W. W. (2000). Applied nonparametric statistics (2nd ed.).
Pacific Grove, CA: Duxbury Press.
DiStefano, C., Zhu, M., & Mîndrilă, D. (2009). Understanding and
using factor scores: Considerations for the applied researcher.
Practical Assessment, Research & Evaluation, 14(20), 1–9.
Grove, S. K. (2007). Statistics for health care research: A practical
workbook. St. Louis, MO: Saunders.
Lupien, S. J., Maheu, F., Tu, M., Fiocco, A., & Schramek, T. E.
(2007). The effects of stress and stress hormones on human cogni-
tion: Implications for the field of brain and cognition. Brain &
Cognition, 65(3), 209–237.
Munro, B. H. (2005). Statistical methods for health care research
(5th ed.). Philadelphia, PA: Lippincott Williams, & Wilkins.
Tabachnick, B. G., & Fidell, L. S. (2006). Using multivariate sta-
tistics (5th ed.). Needham Heights, MA: Allyn & Bacon.
Yerkes, R. M., & Dodson, J. D. (1908). The relation of strength of
stimulus to rapidity of habit-formation. Journal of Comparative
Neurology & Psychology, 18(5), 459–482.
Zar, J. H. (1999). Biostatistical analysis (4th ed.). Upper Saddle
River, NJ: Prentice-Hall.

24
CHAPTER
Using Statistics to Predict
I
n nursing practice, the ability to predict future
events is crucial. Clinical researchers might inves-
tigate if length of hospital stay of patients can be
predicted by severity of illness. Health outcome
researchers want to know what factors play an impor-
tant role in responses of patients to health promotion,
illness prevention, and rehabilitation interventions.
Educators are interested in knowing what variables are
most effective in predicting scores of undergraduate
nurses on the registered nurse (RN) licensure exami-
nation. Advanced practice nurses are interested in
what variables predict their success in passing their
national certification examinations.
The statistical procedure most commonly used for
prediction is regression analysis. The purpose of a
regression analysis is to identify which factor or
factors predict or explain the value of a dependent
(outcome) variable. In some cases, the analysis is
exploratory, and the focus is prediction. In others,
selection of variables is based on a theoretical proposi-
tion, and the purpose is to develop an explanation that
confirms the theoretical proposition.
In regression analysis, the independent (predictor)
variable or variables influence variation or change in
the value of the dependent variable. The goal is to
determine how accurately one can predict the value of
an outcome (or dependent) variable based on the value
or values of one or more predictor (or independent)
variables. This chapter describes some common sta-
tistical procedures used for prediction. These proce-
dures include simple linear regression, multiple
regression, logistic regression, and Cox proportional
hazards regression.
Simple Linear Regression
Simple linear regression provides a means to esti-
mate the value of a dependent variable based on
the value of an independent variable. Simple linear
570
  http://evolve.elsevier.com/Grove/practice/
regression is an effort to explain the dynamics within
the scatter plot by drawing a straight line (the line
of best fit) through the plotted scores. This line
is drawn to explain best the linear relationship
or association between two variables. Knowing that
linear relationship, we can, with some degree of
accuracy, use regression analysis to predict the value
of one variable if we know the value of the other
variable (Cohen & Cohen, 1983). Figure 24-1 illus-
trates the linear relationship between treatment com-
pliance (independent variable) and posttreatment
functional impairment (dependent variable) among
patients with chronic pain. As shown in the scatter
plot, there is a strong inverse relationship between
the two variables. Higher levels of treatment compli-
ance are associated with lower levels of functional
impairment.
In simple linear regression, the dependent variable
is continuous, and the predictor can be any scale of
measurement. However, if the predictor is nominal, it
must be correctly coded. When the data are ready,
the parameters a and b are computed to obtain a
regression equation. To understand the mathematical
process, recall the algebraic equation for a straight
line:
y = bx + a
where
y = dependent variable (outcome)
x = independent variable (predictor)
b = slope of the line (beta, or what the increase in
value is along the x-axis for every unit of
increase in the y value)
a = y-intercept (the point where the regression line
intersects the y-axis)
A regression equation can be generated with a data
set containing participants’ x and y values. When this
equation is generated, it can be used to predict y values
 CHAPTER 24  Using Statistics to Predict 571

of future participants, given only their x values. In
simple or bivariate regression, predictions are made in
cases with two variables. The score on variable y
(dependent variable) is predicted from the same indi-
vidual’s known score on variable x (independent
variable).
No single regression line can be used to predict
with complete accuracy every y value from every x
value. You could draw an infinite number of lines
through the scattered paired values. However, the
purpose of the regression equation is to develop the
line that allows the highest degree of prediction
possible—the line of best fit. The procedure for devel-
oping the line of best fit is the method of least squares.
The formulas for the beta (b) and y-intercept (a) of the
regression equation are computed as follows. Note
6 Clifford, 2002). A subset was selected for this illustra-
5 and manageable. In actuality, studies involving linear
Functional Impairment
that when the b is calculated, that value is inserted into
the formula for a.
nΣXY − ΣXΣY ΣY − bΣX
b= a=
nΣX 2 − ( ΣX ) 2 n
Calculation of Simple Linear Regression
Table 24-1 displays how one would arrange data to
perform linear regression by hand. Regression analy-
sis is conducted by a computer for most studies, but
this calculation is provided to increase your under-
standing of the aspects of regression analysis and how
to interpret the results. This example uses actual clini-
cal data obtained from adults with a painful injury who
received multidisciplinary pain management. The cal-
culation in this chapter includes a subset of the study
data belonging to those patients (n = 12) with the
highest levels of coping skills (Cipher, Fernandez, &
tion so that the computation example would be small

regression need to be adequately powered, so they

4
employ a larger sample (Aberson, 2010; Cohen &
3 Cohen, 1983). The strength of this example is that the
data are actual, unmodified clinical data from a study.
2 The first variable in Table 24-1 is the patient’s level
of treatment compliance—the predictor or independent
1 variable (x) in this analysis. The second variable is the
patient’s posttreatment level of functional impairment—
0 the dependent variable (y). Treatment compliance was
2.0 3.0 4.0 5.0 6.0
Treatment Compliance assessed by patients’ scores on the Cognitive Psy­
chophysiological Treatment Clinical Rating Scales
Figure 24-1  Linear relationship between treatment compliance and (Clifford, Cipher, & Schumacker, 2003), which yields
posttreatment functional impairment. an overall compliance score with higher numbers

TABLE 24-1  Computation of Linear Regression Equation

Patient X (Compliance) Y (Functional Impairment) X2 XY
1 4.9 0.3 24.01 1.47
2 5.5 0.5 30.25 2.75
3 4.0 1.1 16.00 4.40
4 4.1 2.3 16.81 9.43
5 4.4 2.9 19.36 12.76
6 2.9 2.9 8.41 8.41
7 4.0 3.6 16.00 14.40
8 3.1 3.8 9.61 11.78
9 3.8 4.0 14.44 15.20
10 3.6 4.7 12.96 16.92
11 3.8 4.7 14.44 17.86
12 3.0 5.3 9.00 15.90

Σ 47.10 36.10 191.29 131.28

572 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research

representing better levels of compliance with the pain 10

Y (Functional Impairment)
management treatment regimen. Functional impair- 9
ment was represented by patients’ scores on the 8
yˆ = –1.62X + 9.37
7
Interference subscale of the Multidimensional Pain 6
Inventory (Kerns, Turk, & Rudy, 1985), with higher 5
scores representing more functional impairment. The 4
null hypothesis being tested is: Treatment compliance 3
is not a significant predictor of posttreatment func- 2
tional impairment among patients undergoing pain 1
0
management. The data in Table 24-1 are arranged in 0 1 2 3 4 5 6
columns, which correspond to the elements of the X (Treatment Compliance)
formula. The summed values in the last row of Table
24-1 are inserted into the appropriate places in the Figure 24-2  Regression line represented by regression equation.
formula.

Calculation Steps
Step 1: Calculate b. The predicted ŷ is 1.43. This procedure would be
From the values in Table 24-1, we know that n = 12, continued for the rest of the subjects, and the Pearson
ΣX = 47.10, ΣY = 36.10, ΣX2 = 191.29, and ΣXY = correlation between the actual functional impairment
131.28. These values are inserted into the formula for levels (y) and the predicted functional impairment
b, as follows: levels (ŷ) would yield the multiple R value. In this
12 (131.28) − (47.10) (36.10) example, R = 0.75. The higher R, the more likely that
b= the new regression equation accurately predicts y
12 (191.29) − 47.102
because the higher the correlation, the closer the
b = −1.62 actual y values are to the predicted ŷ values. Figure
24-2 displays the regression line where the x-axis
Step 2: Calculate a. represents possible treatment compliance values, and
From Step 1, we know that b = −1.62, and we plug the y-axis represents the predicted functional impair-
this value into the formula for a: ment levels (ŷ values).
36.10 − (−1.62) (47.10) Step 5: Determine whether the predictor variable sig-
a= nificantly predicts y.
12
To know whether the predictor significantly predicts
a = 9.37 y, the value of beta must be tested against zero because
Step 3: Write the new regression equation: zero reflects the presence of no significant association
between the variables, or null hypothesis. In simple
y = −1.62 X + 9.37 regression, this is most easily accomplished by using
the R value from Step 4:
Step 4: Calculate R
The multiple R is defined as the correlation between
the actual y values and the predicted y values using n−2
t = R
the new regression equation. The predicted y value 1 − R2
using the new equation is represented by the symbol
12 − 2
ŷ to differentiate it from y, which represents the actual t = 0.75
y values in the data set. We can use our new regression 1 − 0.56
equation from Step 3 to compute predicted functional t = 3.58
impairment levels for each participant, the outcome of
interest, using his or her compliance rating. For The t value is compared with the t probability dis­
example, the compliance rating of patient no. 1 is 4.9, tribution table (see Appendix B). The df for this t
and the predicted functional impairment for patient no. statistic is n − 2. The critical t value at alpha (α) =
1 is calculated as: 0.05, df = 10 is 2.23. Our obtained t was 3.58, which
yˆ = −1.62 (4.9) + 9.37 exceeds the critical value in the table, indicating a
significant association between the predictor (x) and
yˆ = 1.43 outcome (y).

Step 6: Calculate R2.
After establishing the statistical significance of the R
value, it must be examined for clinical importance.
This examination is accomplished by obtaining the
coefficient of determination for regression—which
simply involves squaring the R value. R2 represents
the percentage of variance in y explained by the pre-
dictor. In our example, R was 0.75, and R2 was 0.56.
Multiplying 0.56 × 100% indicates that 56% of the
variance in posttreatment functional impairment can
be explained by knowing the patients’ level of treat-
ment compliance (Cohen & Cohen, 1983).
R2 is most helpful in testing the contribution of the
predictors in explaining an outcome when more than
one predictor is included in the regression model. In
contrast to R, R2 for one regression model can be
compared with another regression model that contains
additional predictors (Cohen & Cohen, 1983). For
example, Cipher et al. (2002) added another predictor,
the patients’ coping style, to the regression model of
functional impairment. The R2 values of both models
were compared, the first with treatment compliance as
the sole predictor and the second with treatment com-
pliance and copying style as predictors. The R2 values
of the two models were statistically compared to indi-
cate whether the proportion of variance in ŷ was sig-
nificantly increased by including the second predictor
of coping style in the model.
Interpretation of Results
The b (beta or slope of the line) was −1.62, indicat­
ing an inverse relationship between treatment com­
pliance and posttreatment functional impairment.
The t statistic was significant, indicating that we can
reject our null hypothesis and conclude that: Treat­
ment compliance significantly predicted posttreat­
ment functional impairment levels among adults with
chronic pain. Higher levels of treatment compliance
significantly predicted lower levels of functional
impairment.
Multiple Regression
Multiple regression analysis is an extension of
simple linear regression in which more than one inde-
pendent variable is entered into the analysis (Munro,
2005). Because the relationships between multiple
predictors and y are tested simultaneously, the cal­
culations involved in multiple regression analysis are
very complex. Multiple regression is best conducted
using a statistical software package such as those pre-
sented in Table 21-2. However, full explanations and
examples of the matrix algebraic computations of
CHAPTER 24  Using Statistics to Predict 573
multiple regression are presented by Stevens (2009)
and Tabachnick and Fidell (2006).
Interpretations of multiple regression findings are
the same as with simple regression. The beta (b)
values of each predictor are tested for significance,
and a multiple R and R2 are computed. The only dif-
ference is that in multiple regression, when all predic-
tors are tested simultaneously, each b has been adjusted
for every other predictor in the regression model. The
b represents the independent relationship between that
predictor and y, even after controlling for (or account-
ing for) the presence of every other predictor in the
model.
Mancuso (2010) conducted a study of 102 subjects
with diabetes to develop a predictive model of glyce-
mic control, as measured by glycosylated hemoglobin
(HbA1c). The five predictors for HbA1c were health
literacy, patient trust, knowledge of diabetes, perfor-
mance of self-care activities, and depression. The five
predictors of glycemic control were tested with mul-
tiple regression analysis. The analysis yielded five b
values, each with a corresponding p value. As shown
in Table 24-2, patient trust and depression were sig-
nificant predictors of glycemic control (HbA1c), even
after adjusting for the presence or contribution of
every other predictor in the model. Health literacy,
diabetes knowledge, and performance of self-care
activities did not significantly predict HbA1c levels. R2
was 0.285, indicating that patient trust and depression
accounted for 28.5% of the variance in HbA1c (the
measure of glycemic control).
The findings from this study have potential im­
plications for the management of patients with dia-
betes. Because lower levels of patient trust were
associated with higher HbA1c values, fostering com-
munication and trusting collaboration between the
patient and the healthcare provider could directly or
indirectly improve glycemic control. Higher levels of
depression were also associated with higher HbA1c
values, and early interventions or referrals aimed at
addressing depressive symptoms could be important
in improving glycemic control. Regression analysis
is not an indication of cause and effect. However,
these results can serve as a benchmark for further re-
search aimed at investigating the influence of patient
factors such as trust and depression on glycemic
control.
Multicollinearity
Multicollinearity occurs when the independent vari-
ables in a multiple regression equation are strongly
correlated with one another. The presence of multi-
collinearity does not affect predictive power (the
 574 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research

TABLE 24-2  Predictors of Hemoglobin A1c in Patients with Diabetes

Unstandardized Coefficients Standarized Coefficient
Independent Variable B SE β t Significance (p)
Health literacy −0.063 0.080 −0.070 −0.0782 0.436
Patient trust −0.873 0.165 −0.459 −5.288 0.000*
Diabetes knowledge 0.012 0.011 0.100 1.116 0.267
Performance of self-care activities 0.005 0.135 0.003 0.040 0.968
Depression 0.036 0.014 0.226 2.589 0.011*

From Mancuso, J. M. (2010). Impact of health literacy and patient trust on glycemic control in an urban USA populations. Nursing & Health Sciences, 12(1),
94–104.
*p < 0.05, significant.

capacity of the independent variables to predict values
of the dependent variable in that specific sample);
rather, it causes problems related to generalizability
and the stability of the findings. If multicollinearity is
present, the equation lacks predictive validity, and
the amount of variance explained by each variable in
the equation is inflated. Additionally, when cross-
validation is performed, the b values do not remain
consistent across samples (Cohen & Cohen, 1983).
Multicollinearity is minimized by carefully selecting
the independent variables and thoroughly determining
their presence before the regression analysis. If highly
correlated independent variables are found, the cor-
related predictors might be combined into one score
or value yielding one predictor, or only one of the
measures (scores) might be included in the regression
equation.
The first step in identifying multicollinearity is to
examine the bivariate correlations among the indepen-
dent variables. You would perform multiple correla-
tion analyses before conducting the regression
analyses. The correlation matrix is carefully examined
for evidence of multicollinearity. Most researchers
consider multicollinearity to exist if a bivariate cor-
oasis-ebl|Rsalles|1476461084
describes the various terms used as variables in regres-
sion equations.
Dummy Variables
To use categorical variables in regression analysis, a
coding system is developed to represent group mem-
bership. Categorical variables of interest in nursing
that might be used in regression analysis include
gender, income, ethnicity, social status, level of educa-
tion, and diagnosis. If the variable is dichotomous,
such as gender, members of one category are assigned
the number 1, and all others are assigned the number
0. In this case, for gender the coding could be:
1 = female
0 = male
If the categorical variable has three values, two
dummy variables are used; for example, social class
could be classified as lower class, middle class, or
upper class. The first dummy variable (X1) would be
classified as:
1 = lower class
0 = not lower class
The second dummy variable (X2) would be classified
as:

relation is greater than 0.65. However, some research- 1 = middle class

ers use a stronger correlation of 0.80 or greater as an
indication of multicollinearity (Schroeder, 1990).
Types of Independent Variables Used in
Regression Analyses
Variables in a regression equation can take many
forms. Traditionally, as with most multivariate analy-
ses, variables are measured at the interval or ratio
level. However, researchers also use categorical or
dichotomous variables (referred to as dummy vari-
ables), multiplicative terms, and transformed terms.
A mixture of types of variables may be used in a
single regression equation. The following discussion
0 = not middle class
The three social classes would be represented in the
data set in the following manner:
Lower class X1 = 1, X2 = 0
Middle class X1 = 0, X2 = 1
Upper class X1 = 0, X2 = 0
The variables lower class and middle class would
be entered as predictors in the regression equation, in
which both are tested against the reference category,
upper class. Specifically, the b values for these two
variables would represent whether y differs by lower
class versus upper class and middle class versus upper
class. When more than three categories define the
values of the variable, increased numbers of dummy

TABLE 24-3  Association between Angiotensin-
Converting Enzyme (ACE) Inhibitor
Use and Colon Polyps
Polyps No polyps
ACE inhibitor use 251 1508
No ACE inhibitor 603 2284
ACE, Angiotensin-converting enzyme.
variables are used. The number of dummy variables is
always one less than the number of categories (Aiken
& West, 1991). An example of how one might analyze
dichotomous dummy variables is presented in the next
section.
Odds Ratio
When both the predictor and the dependent variable
are dichotomous, the odds ratio (OR) is a commonly
used statistic to obtain an indication of association.
The odds ratio is defined as the ratio of the odds of
an event occurring in one group to the odds of it occur-
ring in another group (Gordis, 2008). Put simply, the
OR is a way of comparing whether the odds of a
certain event is the same for two groups. For example,
the odds of a myocardial infarction occurring in men
versus women could be compared. This research
evidence would be valuable in predicting myocardial
infarctions in men and women in clinical practice
(Melnyk & Fineout-Overholt, 2011; Sackett, Straus,
Richardson, Rosenberg, & Haynes, 2000).
Calculation
The formula for the OR is:
ad
OR =
bc
A study examining the use of angiotensin-
converting enzyme (ACE) inhibitors in 4646 veterans
is presented as an example. The use of ACE inhibitors
was examined in relation to having advanced ade­
nomatous colon polyps in this population (Kedika
et al., 2011). These data are presented in Table 24-3.
The formula for the OR designates the predictor’s
ratios of 1’s to 0’s within the positive outcome in the
numerator and the predictor’s ratios of 1’s to 0’s
within the negative outcome in the denominator. The
values must be coded accordingly. Table 24-4 displays
the following notation to assist you in calculating the
OR:
CHAPTER 24  Using Statistics to Predict 575
TABLE 24-4  Notation in Cells of the Odds
Ratio Table
Polyps No polyps
ACE inhibitor use a b
No ACE inhibitor c d
ACE, Angiotensin-converting enzyme.
ad (251) (2284) 573, 284
OR = = = = 0.63
bc (1508) (603) 909, 324
OR = 0.63
Interpretation
An OR of ≅ −1 indicates that the probability of the
event (polyps) is the same for both groups.
An OR of >1 indicates that the probability of the event
(polyps) is higher among subjects exposed (to ACE
inhibitors).
An OR of <1 indicates that the probability of the event
(polyps) is lower among subjects exposed (to ACE
inhibitors).
The OR for the study was 0.63, indicating the odds
of polyps among veterans using ACE inhibitors is
lower than among veterans who did not use ACE
inhibitors. We can further note that veterans who used
ACE inhibitors were 37% less likely to have polyps
(Kedika et al., 2011). This value was computed by
subtracting the OR from 1.00 (1.00 − 0.63 = 0.37 ×
100% = 37%). The difference between the obtained
OR and 1.00 represents the extent of the lesser or
greater likelihood of the event occurring.
Confidence Intervals
OR values are often accompanied by a confidence
interval, which consists of a lower and upper limit
value. The 95% confidence interval is interpreted as
there is a 95% probability that the population value
lies between [lower limit] and [upper limit]. If the
confidence interval does not include the number 1.00,
this would indicate a significant association between
the medication use and the polyps. The 95% confi-
dence interval for the OR for ACE inhibitor use and
polyps was 0.54 to 0.74; this means that there is a 95%
probability that the population value lies between the
values of 0.54 and 0.74. Because this interval excludes
the value of 1.00, it indicates that there is a statistically
significant association between ACE inhibitor use and
polyps.
576 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
Logistic Regression
Logistic regression replaces linear regression when
the researcher wants to test a predictor or predictors
of a dichotomous dependent variable. Some common
examples of dependent variables that are analyzed
with logistic regression are: patient lived or died,
responded or did not respond to treatment, and
employed or unemployed. The logistic regression
model can be considered more flexible than linear
regression in the following ways:
1. Logistic regression can have continuous predictors,
nominal predictors, or a combination of the two,
with no assumptions regarding normality of the
distribution.
2. Logistic regression can test predictors with a non-
linear relationship between the predictor and the
dependent variable.
3. With a logistic regression model, you can compute
the odds of a person’s outcome. Each predictor is
associated with an OR that represents the indepen-
dent association between that predictor and the
outcome y (Tabachnick & Fidell, 2006).
Because the dependent variable is either 1 or 0,
logistic regression analysis produces a regression
equation that yields probabilities of the outcome
occurring for each person. If the predictor is continu-
ous, we can determine the probability of the outcome
occurring with a predictor score of some value X. If
the predictor is dichotomous, we can determine the
probability of the outcome occurring with a predictor
value of “1” and a predictor value of “0.”
Calculation
Logistic regression is best conducted using a statistical
software package such as those presented in Table
21-2. Full explanations and examples of the computa-
tions of logistic regression are presented by Tabach-
nick and Fidell (2006). A brief overview is provided
in this chapter, with an example of simple logistic
regression using actual clinical data.
Because the dependent variable in logistic regres-
sion is dichotomous, the predicted ŷ is always in the
range of 0 to 1, which is interpreted as a probability.
Similar to linear regression, the predicted ŷ values are
calculated from a b (or more than one b in the case of
multiple predictors) and a y-intercept. In contrast to
linear regression, the b and y-intercept are the expo-
nents of the number e (2.718). An exponent of e is
commonly referred to as the natural logarithm. In
other words, the natural logarithm of a given number
is the power to which e would have to be raised to
equal that number. When the b and y-intercept serve
as natural logarithms, it allows the result to yield a
probability (a value between 0 and 1).
Recall the example from the ACE inhibitor and
polyps data (Kedika et al., 2011). If a person used
ACE inhibitors, the probability of that patient having
colon polyps is calculated:
Given: For these data, b = −0.46, and the y-intercept
(a) is −1.33.
e −0.46 (1) + −1.33
Yˆ =
1 + e −0.46 (1) + −1.33
e −1.79 0.167
Yˆ = so Yˆ = = 0.14
1 + e −1.79 1.167
The probability of having colon polyps if the patient
used ACE inhibitors is only 0.14, or 14%. The prob-
ability of having colon polyps if the patient did not
use ACE inhibitors is 21%, as shown next. The risk of
having polyps is greater if the patient did not use ACE
inhibitors.
e −0.46 ( 0 ) + −1.33
Yˆ =
1 + e −0.46 ( 0 ) + −1.33
e −1.33 0.26
Yˆ = so Yˆ = = 0.21
1 + e −1.33 1.26
Odds Ratio (OR) in Logistic Regression
Each predictor is associated with an OR in a logistic
regression model. If the predictor is dichotomous, the
OR is interpreted as: with an X value of “yes,” the
odds of the outcome occurring is [OR value] times as
likely. The ACE inhibitor and polyps example yielded
an OR of 0.63. As was stated previously, this OR
indicates that veterans using ACE inhibitors were
0.63 times as likely to have polyps. Another way of
stating this is that veterans who used ACE inhibitors
were 37% less likely to have polyps. Thus, patients
treated with ACE inhibitors to lower their blood pres-
sure have a decreased likelihood of developing colon
polyps.
If the predictor is continuous, the OR is interpreted
as: for every 1-unit increase in X, the odds of the
outcome occurring are [OR value] times as likely. For
example, the association between years of education
and obtaining employment among persons with a
spinal cord injury was investigated (Ottomanelli,
Sippel, Cipher, & Goetz, 2011). The predictor was
age, and the dependent variable was employment (yes/
no). The OR was 1.10, indicating that for every year
older in age, the patient was 1.10 times as likely (or
10% more likely) to have obtained employment.
 CHAPTER 24  Using Statistics to Predict 577

In the same study, the association between being
male and obtaining employment among persons with
spinal cord injury was investigated (Ottomanelli et al.,
2011). The predictor was being male (yes/no), and the
dependent variable was employment (yes/no). The OR
was 1.00, indicating that patients who were male were
1.00 times as likely (or just as likely) as females to have
obtained employment. In other words, the likelihood of
employment was equal among males and females.
Cox Proportional
Hazards Regression
When testing predictors of a dependent variable that
is time-related, the appropriate statistical procedure is
Cox proportional hazards regression (or Cox regres-
sion) (Hosmer, Lemeshow, & May, 2008). The depen-
dent variable in Cox regression is called the hazard,
a neutral word intended to describe the risk of event
occurrence (e.g., risk of obtaining an illness, risk of
complications from medications, or risk of relapse).
The primary output in a Cox regression analysis rep-
resents the relationship between each predictor vari-
able and the hazard, or rate of event occurrence.
Cox regression is a type of survival analysis that
can answer questions pertaining to the amount of time
that elapses until an event occurs. Examples of the
types of questions that can be answered using Cox
regression follow. A group of nurse practitioners
begins a doctoral program. What variables predict
how long it will take the students to graduate? A group
of depressed adults completes a cognitive therapy
program. What variables predict the time elapsed from
the end of treatment until a patient’s first relapse?
The major difference between using Cox regression
as opposed to linear regression is the ability of sur-
vival analysis to handle cases where survival time is
unknown. For example, in the study of treatment for
streptococcal pharyngitis (strep throat), perhaps only
20% of cases relapse. The other 80% do not relapse
by the end of the researcher’s study. Thus, it is
unknown how long it will be until the patients relapse.
Survival times that are known only to exceed a certain
value are called censored data. Censored data can
also occur when a participant drops out of the study.
Cox regression calculations take into account cen-
sored data when estimating the relationships between
predictors and y—in contrast to linear regression anal-
yses, which would delete or exclude those cases from
analysis (Hosmer et al., 2008).
Logistic regression yields odds ratios for each pre-
dictor to represent the relationship between that pre-
dictor and y, whereas Cox regression yields hazard
ratios. A hazard ratio (HR) is interpreted almost iden-
tically to an OR with the exception that the HR repre-
sents the risk of the event occurring sooner.
An example of Cox regression used in clinical
research is presented in Table 24-5. Predictors of
major adverse cardiovascular events (MACE) in a
sample of 312 veterans with rheumatoid arthritis were
tested with Cox regression (Banerjee et al., 2008).
There were 10 predictors of cardiovascular events
tested, and the analysis yielded 10 hazard ratios, each

TABLE 24-5  Cox Proportional Hazards Regression Results of Major Adverse Cardiovascular Events in
Veterans with Rheumatoid Arthritis
Hazard Ratio Hazard Ratio
Predictor (Unadjusted)* p Value (Adjusted)† p Value
oasis-ebl|Rsalles|1476461087

DAS score 1.29 0.02 1.31 0.01

Age 1.01 0.62 0.99 0.83
Hypertension 2.55 0.03 2.43 0.08
Tobacco use 1.37 0.33 1.12 0.78
Diabetes 1.3 0.33 0.99 0.99
Hyperlipidemia 2.63 < 0.01 2.45 0.01
History of vascular disease 2.36 < 0.01 2.54 <0.01
DMARD use 0.63 0.06 0.52 <0.01
aTNF-α use 0.65 0.23 0.81 0.02
DMARD + aTNF-α use 0.68 0.34 0.82 0.83

Note: Full explanations and examples of the computations of Cox regression are presented in Hosmer, Lemeshow, & May (2008).
aTNF-α, Anti–tumor necrosis factor-α medication; DAS, disease activity score; DMARD, disease-modifying antirheumatic drug.
*Adjusted for all other model predictors.
From Banerjee, S. D., et al. (2008). Cardiovascular outcomes in male veterans with rheumatoid arthritis. American Journal of Cardiology, 101(8), 1204.
Hosmer, D. W., Lemeshow, S., & May, S. (2008). Applied survival analysis: Regression modeling of time to event data (2nd ed.). Hoboken, NJ: John Wiley &
Sons.
578 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
with a corresponding p value. As shown in Table 24-5,
the disease activity score (DAS) for extent of rheu­
matoid arthritis, hypertension, hyperlipidemia, and
history of vascular disease all were significant predic-
tors of a cardiovascular event when each predictor was
tested separately. However, when all 10 predictors
were tested simultaneously, the hazard ratios were
called adjusted hazard ratios, which means that each
HR has been adjusted for every other predictor in the
regression model. The results of the adjusted HR
values indicated that DAS, hyperlipidemia, history of
vascular disease, disease-modifying antirheumatic
drug (DMARD) use, and anti–tumor necrosis factor
(anti-TNF) medication use all were significant predic-
tors of a MACE, even after controlling for the pres-
ence of every other predictor in the model. Full
explanations and examples of the computations of
Cox regression are presented by Hosmer et al. (2008).
The findings from this study could have indications
for the treatment of rheumatoid arthritis in clinical
practice. Because higher levels of rheumatoid arthritis
disease activity were associated with a greater likeli-
hood of MACE, it could be that the successful control
of rheumatoid arthritis symptoms could directly or
indirectly reduce the risk of MACE. DMARD and
anti-TNF use were associated with a lower risk of
MACE, and so proper medication management of
these patients might be an important factor in reducing
the risk of MACE. Traditional cardiovascular risk
factors studied in other populations (e.g., age, diabe-
tes, smoking history) did not predict MACE in this
sample (D’Agostino et al., 2000; Kannel, McGee, &
Gordon, 1976). Therefore, male veterans with rheu-
matoid arthritis seem to be unique with regard to the
experience of MACE and may require tailored treat-
ment specific to their demographics to minimize car-
diovascular events.
KEY POINTS
• The purpose of a regression analysis is to predict
or explain as much of the variance in the value of
the dependent variable as possible.
• The independent (predictor) variable or variables
cause variation in the value of the dependent
(outcome) variable.
• Simple linear regression provides a means to esti-
mate the value of a dependent variable based on the
value of an independent variable.
• Multiple regression analysis is an extension of
simple linear regression in which more than one
independent variable is entered into the analysis to
predict a dependent variable.
• Multicollinearity occurs when the independent vari-
ables in a multiple regression equation are strongly
correlated and results in unstable findings.
• The odds ratio is a way of comparing whether
the odds of a certain event is the same for two
groups.
• Logistic regression replaces linear regression when
you want to test a predictor or predictors of a
dichotomous dependent variable.
• When testing predictors of a dependent variable
that is time-related, the appropriate statistical pro-
cedure is Cox proportional hazards regression (or
Cox regression).
• The hazard ratio represents the risk of the event
occurring sooner.
REFERENCES
Aberson, C. L. (2010). Applied power analysis for the behavioral
sciences. New York, NY: Routledge Taylor & Francis Group.
Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and
interpreting interactions. Newbury Park, UK: Sage.
Banerjee, S. B., Compton, A. P., Hooker, R. S., Cipher, D. J.,
Reimold, A., Brilakis, E. S., et al. (2008). Cardiovascular out-
comes in male veterans with rheumatoid arthritis. American
Journal of Cardiology, 101(8), 1201–1205.
Cipher, D. J., Fernandez, E., & Clifford, P. A. (2002). Coping style
influences compliance with multidisciplinary pain management.
Journal of Health Psychology, 7(5), 665–673.
Clifford, P. A., Cipher, D. J., & Schumacker, R. E. (2003). Compli-
ance and stages of change in multidisciplinary pain centers. The
Pain Clinic, 15(4), 355–368.
Cohen, J., & Cohen, P. (1983). Applied multiple regression/
correlation analysis for the behavioral sciences (2nd ed.). Hills-
dale, NJ: Erlbaum.
D’Agostino, R. B., Russell, M. W., Huse, D. M., Ellison, R. C.,
Silbershatz, H., Wilson, P. W., et al. (2000). Primary and
subsequent coronary risk appraisal: New results from the
Framingham Study. American Heart Journal, 139(2 Pt 1),
272–281.
Gordis, L. (2008). Epidemiolology (4th ed.). Philadelphia, PA:
Saunders.
Hosmer, D. W., Lemeshow, S., & May, S. (2008). Applied survival
analysis: Regression modeling of time to event data (2nd ed.).
Hoboken, NJ: John Wiley & Sons.
Kannel, W. B., McGee, D., & Gordon, T. (1976). A general cardio-
vascular risk profile: The Framingham Study. American Journal
of Cardiology, 38(1), 46–51.
Kedika, R., Patel, M., Pena Sahdala, H. N., Mahgoub, A., Cipher,
D. J., & Siddiqui, A. A. (2011). Long-term use of angiotensin
converting enzyme inhibitors is associated with decreased inci-
dence of advanced adenomatous colon polyps. Journal of Clinical
Gastroenterology, 45(2), e12–e16.
Kerns, R. D., Turk, D. C., & Rudy, T. E. (1985). The West-Haven
Yale Multidimensional Pain Inventory (WHYMPI). Pain, 23(4),
345–356.

Mancuso, J. M. (2010). Impact of health literacy and patient trust
on glycemic control in an urban USA populations. Nursing &
Health Sciences, 12(1), 94–104.
Melnyk, B. M., & Fineout-Overholt, E. (2011). Evidence-based
practice in nursing & healthcare: A guide to best practice (2nd
ed.). Philadelphia, PA: Lippincott Williams & Wilkins.
Munro, B. H. (2005). Statistical methods for health care research
(5th ed.). Philadelphia, PA: Lippincott Williams & Wilkins.
Ottomanelli, L., Sippel, J., Cipher, D. J., & Goetz, L. (2011). Factors
associated with employment among veterans with spinal cord
injury. Journal of Vocational Rehabilitation, 34(1), 141–150.
CHAPTER 24  Using Statistics to Predict 579
Sackett, D. L., Straus, S. E., Richardson, W. S., Rosenberg, W., &
Haynes, R. B. (2000). Evidence-based medicine: How to practice
and teach EBM. London, UK: Churchill Livingstone.
Schroeder, M. A. (1990). Diagnosing and dealing with multi­
collinearity. Western Journal of Nursing Research, 12(2),
175–187.
Stevens, J. P. (2009). Applied Multivariate Statistics for the Social
Sciences (5th ed.). London, UK: Psychology Press.
Tabachnick, B. G., & Fidell, L. S. (2006). Using Multivariate Sta-
tistics (5th ed.). Needham Heights, MA: Allyn & Bacon.

25
CHAPTER
Using Statistics to Determine
Differences
T
he statistical procedures in this chapter examine
differences between or among groups. Statisti-
cal procedures are available for nominal,
ordinal, and interval/ratio level data. The procedures
vary considerably in their power to detect differences
and in their complexity. How one interprets the results
of these statistics depends on the design of the study.
If the design is quasi-experimental or experimental
and the study is well designed and has no major issues
in regard to threats to internal and external validity,
causality can be considered, and the results can be
inferred to the associated population. If the design is
comparative descriptive, differences identified are
associated only with the sample under study. The para-
metric statistics used to determine differences that are
discussed in this chapter are the independent samples
t-test, paired or dependent samples t-test, and analysis
of variance (ANOVA). If the assumptions for paramet-
ric analyses are not achieved or if study data are at the
ordinal level, the nonparametric analyses of Mann-
Whitney U, Wilcoxon signed-rank test, and Kruskal-
Wallis H are appropriate techniques to use to test the
researcher’s hypotheses. The chapter concludes with
a discussion of the chi-square test of independence,
which is a nonparametric analysis technique for ana-
lyzing nominal level data.
Choosing Parametric versus
Nonparametric Statistics to
Determine Differences
Parametric statistics are always associated with a
certain set of assumptions that the data must meet; this
is because the formulas of parametric statistics yield
valid results only when the properties of the data are
within the confines of these assumptions (Munro,
2005). If the data do not meet the parametric
580
  http://evolve.elsevier.com/Grove/practice/
assumptions, there are nonparametric alternatives that
do not require those assumptions to be met, usually
because nonparametric statistical procedures convert
the original data to rank-ordered or ordinal level data.
Many statistical tests can assist the researcher in
determining whether his or her data meet the assump-
tions for a given parametric test. The most common
assumption (that accompanies all parametric tests) is
the assumption that the data are normally distributed.
The K2 test and the Shapiro-Wilk test are formal tests
of normality that assess whether distribution of a vari-
able is non-normal—that is, skewed or kurtotic (see
Chapter 21) (D’Agostino, Belanger, & D’Agostino,
1990). The Shapiro-Wilk test is used with samples
with less than 1000 subjects. When the sample is
larger, the Kolmogorov-Smirnov D test is more appro-
priate. All of these statistics are found in mainstream
statistical software packages and are accompanied by
a p value. Significant normality tests with p ≤ 0.05
indicate that the distribution being tested is signifi-
cantly different from the normal curve, violating the
normality assumption. The nonparametric statistical
alternative is listed in each section in the event that
the data do not meet the assumptions of each paramet-
ric test illustrated in this chapter.
t-tests
One of the most common parametric analyses used to
test for significant differences between group means
of two samples is the t-test. The independent t-test
analysis technique was developed to examine differ-
ences between two independent groups; the paired or
dependent t-test analysis technique was developed
to examine differences between two matched or paired
groups, or a comparison of pretest and posttest mea-
surements. The details of the independent and paired
t-tests are described in this section.
 CHAPTER 25  Using Statistics to Determine Differences 581

t-test for Independent Samples s1 = group 1 variance

The most common parametric analysis technique used s2 = group 2 variance
in nursing studies to test for significant differences If the two groups have the same number of subjects
between two independent samples is the independent in each group, one can use this simplified formula:
samples t-test. The samples are independent if the
study participants in one group are unrelated to or
s12 + s22
different from the participants in the second group. sX1 − X1 =
Use of the t-test for independent samples involves the n
following assumptions:
1. Sample means from the population are normally where
distributed. n = sample size in each group and not the total
2. The dependent or outcome variable is measured at sample of both groups
the interval/ratio level. Using an example from a study examining the
3. The two samples have equal variance. levels of depression among 16 elderly long-term care
4. All observations within each sample are indepen- residents, differences between residents with and
dent. without dementia were investigated (Cipher & Clif-
The t-test is robust to moderate violation of its ford, 2004). A subset of data for these patients was
assumptions. Robustness means that the results of selected for this example so that the computation
analysis can still be relied on to be accurate when an would be small and manageable (Table 25-1). In actu-
assumption has been violated. The t-test is not robust ality, studies involving t-tests need to be adequately
with respect to the between-samples or within-samples powered to identify significant differences between
independence assumptions, and it is not robust with groups accurately (Aberson 2010; Cohen & Cohen,
respect to an extreme violation of the normality 1983). All data presented in this chapter are actual,
assumption unless the sample sizes are extremely unmodified clinical data for a small number of study
large. Sample groups do not have to be equal for this participants.
analysis—instead, the concern is for equal variance. A The independent variable in this example was level
variety of t-tests have been developed for various of dementia and included two levels—a “no demen-
types of samples. The formula and calculation of the tia” group and a “severe dementia” group. The level
independent samples t-test is presented next. of dementia was based on clinical ratings of neuropsy-
chologists using the Functional Assessment Staging
Calculation Tool (Reisberg, Ferris, Deleon, & Crook, 1982). The
The formula for the t-test is: dependent variable was the score of the long-term care
resident on the Geriatric Depression Scale (GDS)
X1 − X 2 (Yesavage, Brink, & Rose, 1983). The GDS assesses
t =
sX1 − X2
where
X1 = mean of group 1 TABLE 25-1  Depression Scores by Dementia
X 2 = mean of group 2
Level among Elderly Long-Term
sX1 − X2 = to the standard error of the difference
Care Residents
between the two groups.
To compute the t-test, one must compute the No Dementia
denominator in the formula, which is the standard Patient Group GDS Patient Severe Dementia
error of the difference between the means. If the two No. Score No. Group GDS Score
groups have different sample sizes, one must use this 1 5 9 3
formula: 2 5 10 6
3 10 11 5
(n1 − 1)s12 + (n2 − 1)s22  1 1 4 11 12 4
sX1 − X2  +  5 8 13 9
n1 + n2 − 2 n1 n2
6 8 14 7
7 10 15 4
where 8 10 16 7
n1 = group 1 sample size
n2 = group 2 sample size GDS, Geriatric depression scale.
582 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research

the level of depression in elderly adults, with higher Step 5: Compute degrees of freedom (df):
numbers indicative of more depressive symptoms. df = n1 + n2 − 2
The null hypothesis is: There are no significant
differences between elderly adults with dementia df = 8 + 8 − 2
and elderly adults without dementia on depression
scores. df = 14
The computations for the t-test are as follows: Step 6: Locate the critical t value in the t distribution
Step 1: Compute means for both groups, which table in Appendix B at the back of your textbook
involves the sum of scores for each group divided and compare the critical t value with the obtained
by the number in the group. t value.
The mean for Group 1, No Dementia: X1 = 8.38 The critical t value for 14 degrees of freedom at
The mean for Group 2, Severe Dementia: X 2 = 5.63 alpha (α) = 0.05 is 2.15. This means that if we
Step 2: Compute the numerator of the t-test: viewed the t distribution for df = 14, the middle
95% of the distribution would be marked by −2.15
X1 − X 2 = 2.75 and 2.15, as shown in Figure 25-1.
Step 3: Compute the standard error of the difference. Interpretation of Results
a. Compute the variances for each group: Our obtained t is 2.55, exceeding the critical value,
s2 for Group 1 = 5.41 which means that our t-test is significant and repre-
s2 for Group 2 = 3.98 sents a real difference between the two groups. We can
b. Plug into the standard error of the difference reject the null hypothesis and state: An independent
formula: samples t-test computed on GDS scores revealed long-
term residents with no dementia had significantly
s12 + s22 higher depression scores than long-term residents who
sX1 − X2 = had severe dementia, t (14) = 2.55, p < 0.05; X = 8.4
n
versus 5.6. Prior research suggests that elderly resi-
5.41 + 3.98 dents with dementia do not experience less depression,
sX1 − X2 = but rather they have difficulty communicating their
8
distress (Ott & Fogel, 2004; Scherder et al., 2005).
sX1 − X2 = 1.08 With additional research in this area, this knowledge
might be used to facilitate improvements in methods
Step 4: Compute t value: used by healthcare professionals to assess emotional
X1 − X 2 distress accurately among elderly adults with demen-
t = tia (Cipher, Clifford, & Roper, 2006; Thakur & Blazer,
sX1 − X2
2008).
2.75
t = Nonparametric Alternative
1.08 If the data do not meet the assumptions involving
t = 2.55 normality or equal variances for an independent

2½% of t distribution; 2½% of t distribution;

Improbable values of t Improbable values of t
Middle 95% of t
distribution; Probable
values of t
–2.15 2.15

Figure 25-1  Probability distribution of t at df = 14.

 CHAPTER 25  Using Statistics to Determine Differences 583

samples t-test, the nonparametric alternative is the Calculation

Mann-Whitney U test. Mann-Whitney U calcula-
tions involve converting the data to ranks, discarding
any variance or normality issues associated with the
original values. In some studies, the data collected are
ordinal level, and the Mann-Whitney U test is appro-
priate for analysis of the data. The Mann-Whitney U
test is 95% as powerful as the t-test in determining
differences between two groups. For a more detailed
description of the Mann-Whitney U test, see the sta-
tistical textbooks by Daniel (2000) and Munro (2005).
The statistical workbook for healthcare research by
Grove (2007) has exercises for expanding your under-
standing of t-tests and Mann-Whitney U results from
published studies.
t-tests for Paired Samples
When samples are related, the formula used to calcu-
late the t statistic is different from the formula previ-
ously described for independent groups. One type of
paired samples refers to a research design that repeat-
edly assesses the same group of people, a design com-
monly referred to as a repeated measures design.
Another research design for which a paired samples
t-test is appropriate is the case-control research design.
Case-control designs involve a matching procedure
whereby a control subject is matched to each case, in
which the cases and controls are different people but
matched demographically (Gordis, 2008). Paired or
dependent samples t-tests can also be applied to a
crossover study design, in which subjects receive one
kind of treatment and subsequently receive a compari-
son treatment (Gordis, 2008). However, similar to the
independent samples t-test, this t-test requires that dif-
ferences between the paired scores be independent and
normally or approximately normally distributed.
The formula for the paired samples t-test is:
D
t=
SD
where
D = mean difference of the paired data
SD = standard error of the difference
To compute the t-test, one must compute the
denominator in the formula, the standard error of the
difference:
sD
sD =
N
where
sD = standard deviation of the differences between
the paired data
N = number of subjects in the sample
Using an example from a study examining the level
of functional impairment among 10 adults receiving
rehabilitation for a painful injury, changes over time
were investigated (Cipher, Kurian, Fulda, Snider, &
Van Beest, 2007). These data are presented in Table
25-2. A subset was selected for this example so that
the computations would be small and manageable. In
actuality, studies involving both independent and
dependent samples t-tests need to be adequately
powered (Aberson 2010; Cohen & Cohen, 1983).
The independent variable in this example was treat-
ment over time, meaning that the whole sample
received rehabilitation for their injury for 3 weeks.
The dependent variable was functional impairment,
which was represented by patients’ scores on the
Interference subscale of the Multidimensional Pain

TABLE 25-2  Functional Impairment Levels at Baseline and after Treatment

Baseline Functional Post-Treatment Functional
Subject No. Impairment Scores Impairment Scores Difference
1 2.9 1.7 1.2
2 5.7 2.9 2.8
3 2.3 2.9 −0.6
4 3.9 3 0.9
5 3.8 3.1 0.7
6 3.3 3.2 0.1
7 2.9 3.2 −0.3
8 4.7 3.2 1.5
9 3.2 2.1 1.1
10 4.9 3.4 1.5
584 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
Inventory (MPI) (Kerns, Turk, & Rudy, 1985), with
higher scores representing more functional impair-
ment. The null hypothesis is: There is no significant
reduction in functional impairment from baseline
to posttreatment for patients in a rehabilitation
program.
The computations for the t-test are as follows:
Step 1: Compute the difference between each subject’s
pair of data (see last column of Table 25-2).
Step 2: Compute the mean of the difference scores,
which becomes the numerator of the t-test:
D = 0.89
Step 3: Compute the standard error of the difference.
a.  Compute the standard deviation of the differ­
ence scores:
s = 0.99
b.  Plug into the standard error of the difference
formula:
s test. The Wilcoxon signed-rank test calculations
sD = D
N involve converting the data to ranks, discarding any
0.99
sD =
10
sD = 0.313
Step 4: Compute t value:
D an exercise for expanding your understanding of the
t =
sD
0.89
t =
0.313
t = 2.84
Step 5: Compute degrees of freedom:
df = n − 1
df = 10 − 1
df = 9
Step 6: Locate the critical t value on the t distribution
table and compare with our obtained t value.
The critical t value for 9 degrees of freedom at
alpha (α) = 0.05 is 2.26. Our obtained t is 2.84,
exceeding the critical value (see t—Table in
Appendix B). This means that if we viewed
the t distribution for df = 9, the middle 95%
of the distribution would be marked by −2.26
and 2.26.
Interpretation of Results
Our obtained t = 2.84 exceeds the critical t value in the
table, which means that our t-test is statistically signifi-
cant and represents a real difference between partici-
pant’s preintervention and postintervention functional
impairment scores. We can reject our null hypothesis
and state: A paired samples t-test computed on MPI
functional impairment scores revealed that the patients
undergoing rehabilitation had significantly lower func-
tional impairment levels from baseline to posttreat-
ment, t (9) = 2.84, p < 0.05; X = 3.8 versus 2.9. During
the 3-week rehabilitation program, patients success-
fully reduced their functional impairment levels. With
additional research in this area, this knowledge might
be used to facilitate evidence-based practice interven-
tions in rehabilitation facilities to improve patients’
functional status (Melnyk & Fineout-Overholt, 2011).
Nonparametric Alternative
If the interval/ratio level data do not meet the normal-
ity assumptions for a paired samples t-test, the non-
parametric alternative is the Wilcoxon signed-rank
variance or normality issues associated with the origi-
nal values. This analysis technique is also appropriate
when the study data are ordinal level, such as self-care
abilities identified as low, moderate, and high based
on the Orem Self-Care Model (Orem, 2001). This test
is thoroughly addressed by Daniel (2000) and Munro
(2005) in their statistical textbooks. Grove (2007) has
Wilcoxon signed-rank results from published studies.
One-Way Analysis of Variance
Analysis of variance (ANOVA) is a statistical proce-
dure that compares data between two or more groups
or conditions to investigate the presence of differences
between those groups on some continuous dependent
or outcome variable. The formulas for ANOVA
compute two estimates of variance: (1) differences
among (within) the data and (2) differences between
the groups or conditions.
Why perform ANOVA and not a t-test? A t-test is
formulated to compare two sets of data or two groups
at one time. Data generated from a clinical trial that
has four experimental groups, treatment 1, treatment
2, treatments 1 and 2 combined, and a control, would
require six t-tests. As a result, the chance of making a
Type I error (alpha error) increases substantially (or is
inflated) because so many computations are being per-
formed. Specifically, the chance of making a Type I
 CHAPTER 25  Using Statistics to Determine Differences 585

TABLE 25-3  Monthly Medical Costs of Three Treatment Groups*

Multidisciplinary Group Costs Standard Care Group Costs Pharmacotherapy Group Costs
74 168 603
748 328 707
433 186 868
422 199 286
297 154 919
Σ = Sum = 1974 1035 3382
Grand total = 6392

*Costs were U.S. dollars averaged monthly.

error is the number of comparisons multiplied by the
alpha level. ANOVA is the better statistical technique
for examining differences among more than two
groups.
ANOVA is a procedure that culminates in a statistic
called the F statistic. This value is compared against
an F distribution (see Appendix D) to determine
whether the groups significantly differ from one
another on the dependent variable studied. The basic
formula for the F is:
Mean square between groups
F =
Mean square within groups
accurately among study groups (Aberson 2010; Cohen
& Cohen, 1983).
The independent variable in this example was the
type of study treatment or intervention. Patients
received multidisciplinary treatment, standard care
(primary care physician only), or pharmacotherapy by
an anesthesiologist. The dependent variable was the
medical costs incurred per month during the year after
treatment. The null hypothesis was: There is no sig-
nificant difference between the treatment groups in
posttreatment monthly medical costs.
The steps to perform an ANOVA are as follows:
Step 1: Compute correction term, C.
Square the grand sum (G), and divide by total N:

The term mean square (MS) is used interchange- 63922

ably with the word “variance.” The formulas for C = = 2, 723, 844.3
15
ANOVA compute two estimates of variance: the
between-groups variance and the within-groups vari- Step 2: Compute total sum of squares.
ance. The between-groups variance represents differ- Square every value in data set, sum, and subtract
ences between the groups or conditions being C:
compared, and the within-groups variance represents
differences among (within) each group’s data. The (742 + 7482 + 4332 + 4222 + 2972 + 1682 + 3282 …
formula is: + 9192 ) − 2, 723, 844.3 = 3, 795, 722
MS between − 2, 723, 844.3 = 1, 071, 877.7
F =
MS within
Step 3: Compute between groups sum of squares.
Calculation Square the sum of each column and divide by N.
Using an example from a study examining medical Add each, and then subtract C.
costs among 15 women receiving treatment for a
19742 10352 33822
chronic pain condition, monthly medical costs for 1 + + − 2723, 844.3
year incurred after treatment ended were examined 5 5 5
(Cipher, Fernandez, & Clifford, 2001). The data from (779, 335.2 + 214, 245 + 2, 288, 937.8)
this study are presented in Table 25-3. A subset of data − 2, 723, 844.3 = 558, 673.7
was selected from the unmodified clinical data of this
study for this example so that the computation would Step 4: Compute within groups sum of squares.
be manageable. In actuality, studies involving ANOVA Subtract the between groups sum of squares (Step
need to be adequately powered to detect differences 3) from total sum of squares (Step 2).
586 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
TABLE 25-4  Analysis of Variance Summary Table
Source of
Variation SS df MS F required to complete the ANOVA process and deter-
Between groups 558,673.73 2 279,336.9 6.53
Within groups 513,204 12 42,767
Total 1,071,877.7 14
1, 071, 877.7 − 558, 673.7 = 513, 204
Step 5: Create an ANOVA summary table (Table
25-4).
a.  Insert the sum of squares values in the first
column.
b.  The degrees of freedom are in the second
column. Because the F is a ratio of two separate
statistics (mean square between groups and
mean square within groups) both have different
df formulas—one for the “numerator” and one
for the denominator:
Mean square between groups df
= number of groups − 1
Mean square within groups df
= (number of groups − 1) (n − 1)
For this example, the df for the numerator is 3
− 1 = 2. The df for the denominator is (3 − 1)
(5 − 1) = 12.
c.  The mean square between groups and mean
square within groups are in the third column.
These values are computed by dividing the
SS by the df. Therefore, the MS between =
558,673.73 ÷ 2 = 279,336.9. The MS within =
513,204 ÷ 12 = 42,767.
d.  The F is the final column and is computed by
dividing the MS between by the MS within.
Therefore, F = 279,336.9 ÷ 42,767 = 6.53.
Step 6: Locate the critical F value on the F distribution
table (see Appendix D) and compare our obtained
F value with it. The critical F value for 2 and 12
df at alpha (α) = 0.05 is 3.88. Our obtained F is
6.53, which exceeds the critical value.
Interpretation of Results
Our obtained F = 6.53 exceeds the critical value in the
table, which means that our F is statistically signifi-
cant and that the population means are not equal. We
can reject our null hypothesis that the three groups
have the same monthly posttreatment medical costs.
However, the F does not tell us which treatment
groups differ from one another. Further testing, termed
multiple comparison tests or post hoc tests, are
mine all the significant differences among the study
groups.
Post hoc tests have been developed specifically to
determine the location of group differences after
ANOVA is performed on data from more than two
groups. These tests were developed to reduce the inci-
dence of a Type I error. Frequently used post hoc tests
are the Newman-Keuls test, the Tukey Honestly Sig-
nificant Difference (HSD) test, the Scheffe test, and
the Dunnett test (Munro, 2005). When these tests are
calculated, the alpha level is reduced in proportion to
the number of additional tests required to locate sta-
tistically significant differences. For example, for
several of the aforementioned post hoc tests, if many
groups’ mean values are being compared, the magni-
tude of the difference is set higher than if only two
groups are being compared. Post hoc tests are tedious
to perform by hand and are best handled with statisti-
cal computer software programs.
After computing post hoc tests for our example, the
final interpretation is: Analysis of variance performed
on monthly medical costs revealed significant differ-
ences between the three treatment groups, F (2, 12) =
6.53, p < 0.05. Post hoc comparisons using the Tukey
HSD comparison test indicated that the patients in the
pharmacotherapy group incurred significantly higher
medical costs after treatment than the patients in the
standard care group ($676.60 versus $207.00).
However, there were no significant differences between
the monthly medical costs of the multidisciplinary
group and the standard care group or between the
multidisciplinary group and the pharmacotherapy
group. Grove’s (2007) statistical workbook for health-
care research has exercises for expanding your inter-
pretation and understanding of ANOVA and post hoc
procedure results from published studies.
Nonparametric Alternative
If the data do not meet the normality assumptions for
an ANOVA, the nonparametric alternative is the
Kruskal-Wallis test. Calculations for the Kruskal-
Wallis test involve converting the data to ranks, dis-
carding any variance or normality issues associated
with the original values. Similar to the ANOVA, the
Kruskal-Wallis test is a nonparametric analysis tech-
nique that can accommodate the comparisons of more
than two groups. The Kruskal-Wallis test is also the
appropriate analysis technique to use if the study
data being analyzed are ordinal level. This test is
 CHAPTER 25  Using Statistics to Determine Differences 587
thoroughly addressed in textbooks by Daniel (2000)
and Munro (2005).
Other ANOVA Procedures
The ANOVA example presented earlier used equal
sample sizes in each group. The calculations for an
ANOVA with unequal sample sizes are slightly
different. Moreover, there are other kinds of ANOVA
that accommodate other research designs involving
various numbers of independent and dependent
variables, such as factorial ANOVA, repeated mea-
sures ANOVA, and mixed factorial ANOVA. These
ANOVA procedures are presented and explained
in comprehensive statistics textbooks such as Zar’s
text (2007).
Chi-Square Test
of Independence
The chi-square (χ2) test compares differences in pro-
portions of nominal level variables. When a study
requires that researchers compare proportions (per-
centages) in one category versus another category, the
χ2 is a statistic that reveals if the difference in propor-
tion is statistically improbable. The χ2 has its own
theoretical distribution and associated χ2 table (see
Appendix E).
A one-way chi-square is a statistic that compares
different levels of one variable only. For example, a
researcher may collect information on gender and
compare the proportions of males to females. If the
one-way chi-square is statistically significant, it would
indicate that proportions of one gender are signifi-
cantly higher than proportions of the other gender
than what would be expected by chance (Daniel,
2000). A two-way chi-square is a statistic that tests
whether proportions in levels of one variable are
significantly different from proportions of the second
variable. For example, the presence of advanced
colon polyps was studied in three groups of patients:
patients having a normal body mass index (BMI),
patients who were overweight, and patients who
were obese (Siddiqui et al., 2009). The research ques-
tion tested was: Is there a significant difference
between the three groups (normal, overweight, and
obese) in the presence of advanced colon polyps? The
results of the chi-square analysis indicated that a larger
proportion of obese patients fell into the category of
having advanced colon polyps compared with normal-
weight and overweight patients, suggesting that
obesity may be a risk factor for developing advanced
colon polyps.
Assumptions
One assumption of the chi-square test is that only one
datum entry is made for each subject in the sample. If
repeated measures from the same subject are being
used for analysis, such as pretests and posttests, chi-
square is not an appropriate test (the McNemar test
is the appropriate test) (Munro, 2005). Another
assumption is that for each variable, the categories are
mutually exclusive and exhaustive. No cells may have
an expected frequency of zero. However, in the actual
data, the observed cell frequency may be zero. Until
more recently, each cell was expected to have a fre-
quency of at least five, but this requirement has been
mathematically shown to be unnecessary. However,
no more than 20% of the cells should have fewer than
five (Conover, 1971). The test is distribution-free, or
nonparametric, which means that no assumption has
been made for a normal distribution of values in the
population from which the sample was taken.
Calculation
The computations for a two-way chi-square test are
presented in this section. A study examining the pres-
ence of candiduria (presence of Candida species in the
urine) among 97 adults with a spinal cord injury is
presented as an example. The differences in the use of
antibiotics were investigated using chi-square analysis
(Goetz, Howard, Cipher & Revankar, 2010).The data
for this study are presented in Table 25-5, as a contin-
gency table. A contingency table is a table that dis-
plays the relationship between two or more categorical
variables (Daniel, 2000).
The formula for a two-way chi-square is:
n[(A)(D) − (B)(C)]2
χ2 =
(A + B)(C + D)(A + C)(B + D)
where the contingency table is labeled as such:
A B
C D
Fitting the cells into the formula would appear as:
TABLE 25-5  Candiduria and Antibiotic Use in
Adults with Spinal Cord Injuries
Candiduria No candiduria
Antibiotic use 15 43
No antibiotic use 0 39
588 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
97 [(15)(39) − (43)(0)]2
χ2 =
(15 + 43)(0 + 39)(15 + 0)(43 + 39)
33,195, 825
χ2 =
2, 782, 260
χ2 = 11.93
With any chi-square analysis, the degrees of
freedom (df) must be calculated to determine the sig-
nificance of the value of the statistic. The following
formula is used for this calculation:
df = ( R − 1)(C − 1)
where
R = number of rows
C = number of columns
In the example, the chi-square value was 11.93, and
df was 1, which was calculated as follows:
df = (2 − 1)(2 − 1) = 1
Interpretation of Results
The chi-square statistic is compared with the chi-
square values in the table in Appendix E. The table
includes the critical values of chi-square for specific
degrees of freedom at selected levels of significance.
If the value of the statistic is equal to or greater than
the value identified in the chi-square table, the differ-
ence between the two variables is statistically signifi-
cant. The critical χ2 for df = 1 is 3.84, and our obtained
χ2 is 11.93, exceeding the critical value and indicating
a significant difference between antibiotic users and
nonusers in the presence of candiduria. Subjects who
used antibiotics had significantly higher rates of can-
diduria than subjects who did not use antibiotics (26%
versus 0%). This finding suggests that antibiotic use
may be a risk factor for developing candiduria, and
further research is needed to investigate candiduria as
a direct effect of antibiotics.
KEY POINTS
• Parametric statistics used to determine differences
are accompanied by certain assumptions, and the
data must be tested for whether they meet those
assumptions before computing the statistic.
• Many tests of normality can assist the researcher in
determining the suitability of the data for the use
of parametric statistics.
• In the event that the data do not meet the
assumptions of the parametric statistic, there are
nonparametric alternatives that do not adhere to the
assumptions of the parametric test.
• The t-test is one of the most commonly used para-
metric analyses to test for significant differences
between statistical measures of two samples or
groups.
• The independent samples t-test indicates a differ-
ence between two groups of subjects, whereas the
paired samples t-test indicates a difference in two
assessments of the same subjects or two groups
matched on selected variables.
• The Mann-Whitney U test is the nonparametric
alternative to the independent samples t-test when
the study data violate one or more of the indepen-
dent samples t-test assumptions.
• The Wilcoxon signed-rank test is the nonparamet-
ric alternative to the paired or dependent samples
t-test when the study data violate one or more of
the paired samples t-test assumptions.
• ANOVA can be used to examine data from two or
more groups and compares the variance within
each group with the variance between groups.
• ANOVA conducted on three or more groups that is
significant requires the use of post hoc analysis
procedures to determine the location of group
differences.
• The Kruskal-Wallis test is the nonparametric alter-
native to the ANOVA when the study data violate
one or more of the ANOVA assumptions.
• The chi-square test compares proportions (percent-
ages) in one category of a variable of interest with
proportions in another category.
• McNemar test is the appropriate statistical test to
use when analyzing nominal level data obtained
from repeated measures from the same subject,
such as pretests and posttests.
REFERENCES
Aberson, C. L. (2010). Applied power analysis for the behavioral
sciences. New York, NY: Routledge Taylor & Francis Group.
Cipher, D. J., & Clifford, P. A. (2004). Dementia, pain, behavioral
disturbances and ADLs: Toward a comprehensive conceptualiza-
tion of quality of life in long-term care. International Journal of
Geriatric Psychiatry, 19(8), 741–748.
Cipher, D. J., Clifford, P. A., & Roper, K. D. (2006). Behavioral
manifestations of pain in the demented elderly. Journal of the
American Medical Director’s Association, 7(6), 355–365.
Cipher, D. J., Fernandez, E., & Clifford, P. A. (2001). Cost effective-
ness of multidisciplinary pain management: Comparison of three
treatment groups. Journal of Clinical Psychology in Medical Set-
tings, 8(4), 237–244.
Cipher, D. J., Kurian, A. K., Fulda, K. G., Snider, R., & Van Beest,
J. (2007). Using the MBMD to delineate treatment outcomes in
 CHAPTER 25  Using Statistics to Determine Differences 589
rehabilitation. Journal of Clinical Psychology in Medical Settings,
14(2), 102–112.
Cohen, J., & Cohen, P. (1983). Applied multiple regression/
correlation analysis for the behavioral sciences. Hillsdale, NJ:
Lawrence Erlbaum.
Conover, W. J. (1971). Practical nonparametric statistics. New
York, NY: Wiley.
D’Agostino, R. B., Belanger, A., & D’Agostino, R. B., Jr. (1990).
A suggestion for using powerful and informative tests of normal-
ity. American Statistician, 44(4), 316–321.
Daniel, W. W. (2000). Applied nonparametric statistics (2nd ed.).
Pacific Grove, CA: Duxbury Press.
Goetz, L., Howard, M., Cipher, D., & Revankar, S. G. (2010).
Occurrence of candiduria in a population of chronically catheter-
ized patients with spinal-cord injury. Spinal Cord, 48(1), 51–54.
Gordis, L. (2008). Epidemiolology (4th ed.). Philadelphia, PA:
Saunders.
Grove, S. K. (2007). Statistics for health care research: A practical
workbook. St. Louis, MO: Saunders.
Kerns, R. D., Turk, D. C., & Rudy, T. E. (1985). The West-Haven
Yale Multidimensional Pain Inventory (WHYMPI). Pain, 23(4),
345–356.
Melnyk, B. M., & Fineout-Overholt, E. (2011). Evidence-based
practice in nursing & healthcare: A guide to best practice (2nd
ed.). Philadelphia, PA: Lippincott Williams & Wilkins.
Munro, B. H. (2005). Statistical methods for health care research.
Philadelphia: PA: Lippincott Williams & Wilkins.
Orem, D. E. (2001). Nursing concepts of practice (6th ed.). St.
Louis, MO: Mosby.
Ott, B. R. & Fogel, B. S. (2004). Measurement of depression in
dementia: Self vs clinician rating. International Journal of Geri-
atric Psychiatry, 7(12), 899–904.
Reisberg, B., Ferris, B., Deleon, M. J., & Crook, T. (1982).
The global deterioration scale for assessment of primary degen-
erative dementia. American Journal of Psychiatry, 139(9),
1136–1139.
Scherder, E., Oosterman, J., Swaab, D., Herr, K., Ooms, M., Ribbe,
M., et al (2005). Recent developments in pain in dementia. British
Medical Journal, 330(7489), 461–464.
Siddiqui, A. A., Nazario, H., Mahgoub, A., Pandove, S., Cipher, D.
J., & Spechler, S. J. (2009). Obesity is associated with an
increased prevalence of advanced adenomatous colon polyps in a
male veteran population. Digestive Disease & Sciences, 54(7),
1560–1564.
Thakur, M., & Blazer, D. G. (2008). Depression in long-term care.
Journal of the American Medical Directors Association, 9(2),
82–87.
Yesavage, J., Brink, T., & Rose, T. (1983). Development and valida-
tion of a geriatric depression scale: A preliminary report. Journal
of Psychiatry Reserves, 17(1), 37–49.
Zar, J. H. (2007). Biostatistical Analysis (5th ed.). Upper Saddle
River, NJ: Prentice-Hall.

26
CHAPTER
Interpreting Research Outcomes
W
hen data analysis is complete, there is a
feeling that the answers are known and the
study is finished. However, the results of
statistical analysis alone are inadequate to complete a
study. The researcher may know the results, but without
careful intellectual examination, these results are of
little use to others or to the body of nursing knowledge.
To be useful, the evidence from data analysis needs to
be carefully examined, organized, and given meaning,
and the statistical significance and clinical importance
need to be assessed with implications for practice and
directions for further research. This process is referred
to as the interpretation of research outcomes.
Interpretation is one of the most important parts of
a study because some of the most profound insights
of the entire research process occur during this step of
the study. There is a tendency to rush this important
step to finish the study, but it is not a step to be mini-
mized or hurried. The process takes time for reflection.
Often, a researcher becomes too close to the details to
be able to see the big picture. At this point, dialogue
with colleagues or mentors can add clarity and expand
meaning to the interpretation of study findings and
formation of conclusions.
Data collection and analysis are action-oriented
activities that require concrete thinking. However,
when interpreting the results of the study, one tends to
implement abstract thinking, including the creative
use of introspection, reasoning, and intuition. In some
ways, these last steps in the research process are the
most difficult. They require you to synthesize the logic
used to develop the research plan, implement the strat-
egies used in the data collection phase, and examine
the mathematical logic or insight and pattern forma-
tion obtained in data analysis. Evaluating the research
process used in the study, producing meaning from the
results, and forecasting the usefulness of the findings
are all part of interpretation and require high-level
intellectual processes.
590
  http://evolve.elsevier.com/Grove/practice/
The interpretation of research outcomes is usually
the final chapter of theses or dissertations and the
final section of research articles, which is often enti-
tled “Discussion.” Presentations of studies often con-
clude with an interpretation of the research outcomes.
This chapter focuses on the interpretation of research
outcomes from quantitative, outcomes, and inter­
vention research. The interpretations of research
outcomes for qualitative research are presented in
Chapter 12. The process of interpreting research out-
comes for quantitative, outcomes, and intervention
studies includes the following: (1) examining study
evidence, (2) determining findings, (3) forming con-
clusions, (4) identifying limitations, (5) generalizing
the findings, (6) considering implications for practice,
and (7) suggesting further studies. Each of these
activities is discussed in this chapter with examples
from a quantita­tive descriptive, predictive correla-
tional study by Rutkowski and Connelly (2011).
Examining Evidence
The first step in interpretation involves considering
the evidence available that supports or contradicts the
validity of study results related to the research purpose
and objectives, questions, or hypotheses. The process
is similar to conducting a critical appraisal of your
own study. Your temptation is to ignore flaws—
certainly not to point them out. However, an honest
completion of this process is essential to build a body
of knowledge. It is not a time for confession, remorse,
and apology but rather for thoughtful reflection. You
need to identify the limitations of your study and con-
sider how these might affect the study findings and
conclusions (Fawcett & Garity, 2009).
Evidence from the Research Plan
The initial evidence regarding the validity of the study
results is derived from reexamining the research plan.
 CHAPTER 26  Interpreting Research Outcomes 591
Reexamination requires that you reexplore the logic
of the methodology. Analyze the logical links among
the problem statement, purpose, research questions,
variables, framework, design, sample, measurement
methods, and types of analyses. These elements of the
study should logically link together and be consistent
with the research problem. Remember the old adage:
a chain is only as strong as its weakest link. This saying
is also true of research because all studies have limita-
tions. This examination of the study needs to identify
its weakest links or limitations as well as its strengths.
You need to examine these limitations in terms of
their impact on the results. Could the results, or some
of the results, be a consequence of a weakness in the
methodology rather than a true test of the hypotheses?
Can the research objectives, questions, or hypotheses
be answered from the methodology used in the study?
Could the results be a consequence of an inappropriate
conceptual or operational definition of a variable? Do
the research questions clearly emerge from the frame-
work? Can the results be related back to the frame-
work? Are the analyses logically planned to address
the research objectives, questions, or hypotheses?
If the types of analyses are inappropriate for exam-
ining the research questions, what do the results of
analyses mean? For example, if the design failed to
control extraneous variables, could some of these vari-
ables explain the results, rather than the results being
explained by the variables measured and examined
through statistical analysis? Was the sample studied a
logical group on which to test the hypotheses? The
researcher must carefully evaluate each link in the
design to determine potential weaknesses. Every link
is clearly related to the meaning given to the study
results. If the researcher is reviewing a newly com-
pleted study and determines that some of the analyses
were inappropriate, then these analyses need to be
redone. If the study has several weaknesses or breaks
in logical links, the findings may need to be seriously
questioned.
Evidence from Measurement
One assumption often made in interpreting study
results is that the study variables were adequately
measured. This adequacy is determined by examining
the fit of operational definitions with the framework
and through validity and reliability information.
Although you should determine the reliability and
validity of measurement strategies before using them
in your study, you need to reexamine the measures at
this point to determine the strength of evidence avail-
able from the results (Fawcett & Garity, 2009; Waltz,
Strickland, & Lenz, 2010). For example, did the scale
used to measure anxiety truly reflect the anxiety expe-
rienced in the study population? What was the effect
size? Were the validity and reliability of instruments
examined in the present study? Can this information
be used to interpret the results? The validity and reli-
ability of measurement methods are critical to the
validity of results
A descriptive, predictive correlational study by
Rutkowski and Connelly (2011) is presented as an
example throughout this chapter. The purpose of this
study was “to examine the relationships between
parent physical activity, parent-adolescent obesity
risk knowledge, and adolescent physical activity”
(Rutkowski & Connelly, 2011, p. 52). These research-
ers based their study on the significant problem of
childhood obesity in the United States and the need to
determine knowledge of adolescents and their parents
of obesity risk and their physical activity. The study
included three mea­surement methods, Obesity Risk
Knowledge Scale (ORK-10) (Swift, Glazebrook, &
Mcdonald, 2006), International Physical Activity
Questionnaire (IPAQ, 2001), and Patient Centered
Assessments and Counseling for Exercise Plus Nutri-
tion + Moderate Vigorous Physical Activity (PACE+
MVPA) (Prochaska, Sallis, & Long, 2001). The
researchers clearly linked the study variables (obesity
risk knowledge, physical activity of adults, and physi-
cal activity of adolescents) to the three measurement
methods in Table 26-1. The measurement methods,
validity, and reliability were addressed in the follow-
ing study excerpt.
“The ORK-10 (Swift et al., 2006) is a 10-item self-
report instrument that measures knowledge regard-
ing the health risks associated with obesity.… Internal
consistency reliability estimates have been reported
with Cronbach’s alpha coefficient [r] of .83 (Swift
et al., 2006). The measure has not been tested with
children, but Swift encourages its use with children
12 years and older (J. Swift, personal communication,
July 10, 2007). In this study, the coefficient alphas for
the adolescent scale scores are .53 and .59 for
parents.…
The PACE+ Adolescent Physical Activity Measure
(MVPA; Prochaska et al., 2001) was originally devel-
oped as a screening tool for use by clinical staff to
measure physical activity levels in adolescents seeking
treatment in primary care.… Although brief, this
instrument is practical and assesses targeted behavior
that offers clinical information to practitioners.…
The IPAQ has been developed to provide a set
of well-developed instruments that are used
592 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research

TABLE 26-1  Summary of Study Measures and Scoring

Variable Measure
Obesity Risk Knowledge
Adults and adolescents ORK-10 (data from Swift et al., 2006): includes 10 items that yield responses of either 1 or 0 points.
Scores range from 0-10 with higher scores indicating higher levels of knowledge.
Physical Activity
Adults IPAQ (http://www.ipaq.ki.se/IPAQ.asp?mnu_sel=EEF&pg_sel=DDE): Data analyzed in minutes/day and
sorted into the following three categories: (1) low—no activity is reported or some activity is reported
but not enough to meet categories 2 or 3; (2) moderate—≥3 days of vigorous activity of at least 20
minutes or ≥5 days of moderate-intensity activity or walking of at least 30 minutes per day or ≥5 days
of any combination of walking, moderate-intensity, or vigorous-intensity activities achieving a
minimum of at least 600 MET-minutes/week; (3) high—either vigorous-intensity activity on at least 3
days and accumulating at least 1500 MET-minutes/week or ≥7 days of any combination of walking,
moderate-intensity, or vigorous-intensity activities accumulating at least 3000 MET-minutes/week.
Adolescents PACE+ Adolescent Physical Activity Measure (data from Prochaska et al., 2001): each response has a
scale of 0-7 days. The scores for each question are added together, and the total is divided by 2. A score
of <5 indicates that the Healthy People 2010 guideline for adolescent physical activity is not being met.

MET, Metabolic equivalent of the task.

Rutkowski, E. M., & Connelly, C. D. (2011). Obesity risk knowledge and physical activity in families of adolescents. Journal of Pediatric Nursing, 26(1), 53.

reliability of the ORK-10 is a study limitation that

internationally to obtain comparable estimates of
[adult] physical activity (IPAQ, 2001). This instrument
is designed in both a short version including four
generic items and a long version that reflects five
activity domains (IPAQ). The study presented here
incorporates the short version because it is able to
capture the general conditions of parental activity
without being burdensome to complete.… Validity
and reliability evaluation in 14 countries and across
six continents has been reported (IPAQ). In this study,
coefficient alpha [r] is .80.” (Rutkowski & Connelly,
2011, pp. 53-54)
Rutkowski and Connelly (2011) provided a detailed
description of their measurement methods and how
they were scored (see Table 26-1). The validity and
reliability of the IPAQ are very strong because the
scale has been used internationally in several countries
and had strong reliability in this study (r = 0.80). The
PACE+ MVPA is recognized to have test-retest reli-
ability and concurrent validity, but no specific infor-
mation was provided. The ORK-10 has evidence of
strong internal consistency reliability with adults in
previous studies but has not been used with adoles-
cents and had low internal reliability alpha coefficients
in this study (r = 0.053 for adolescents and r = 0.59
for parents). The discussion of the measurement
methods would have been strengthened by an expanded
description of the ORK-10 scale validity and the reli-
ability and validity for the PACE+ MVPA. The
might have influenced the study findings. The research-
ers recognized the limitation of the ORK-10 scale
reliability for both adolescents and adults, which is
discussed later in this chapter.
Evidence from the Data Collection Process
Many activities that occur during data collection
affect the meaning of study results. Did the study
have a high refusal rate for subject participation, or
was the attrition high? Was the sample size sufficient
(Aberson, 2010; Fawcett & Garity, 2009; Thompson,
2002)? Did strategies for acquiring a sample eliminate
important groups whose data would have influenced
the results? Did the research team achieve interven-
tion fidelity when the treatment was implemented
(Stein, Sargent, & Rafaels, 2007)? Did unforeseen
events occur during the study that might have changed
or had an impact on the data? Were measurement
techniques consistent? What impact do inconsisten-
cies have on interpreting results? Sometimes data col-
lection does not proceed as planned. Unforeseen
situations alter the collection of data. What were these
variations in the study? What impact do they have on
interpreting the results? Sometimes someone other
than the subject completes data collection forms.
Also, variations may occur when scales are adminis-
tered. For example, an anxiety scale may be given to
one subject immediately before a painful procedure
and to another subject on awakening in the morning.
Values on these measures cannot be considered com-
parable. Data integrity also depends on the responses
 CHAPTER 26  Interpreting Research Outcomes 593
of the research participants, which could be compro-
mised by anxiety, time constraints, denial, or other
factors not in the direct control of the researcher. The
researcher must be on the alert for these subject
factors that could compromise the integrity of the
data. Values on these measures cannot be considered
comparable. These types of differences are seldom
reported and sometimes not even recorded. To some
extent, only the researcher knows how consistently
the measurements were taken. Reporting of this infor-
mation depends on the integrity of the researcher
(Fawcett & Garity, 2009; Kerlinger & Lee, 2000;
Pyrczak & Bruce, 2005; Stein et al., 2007).
Rutkowski and Connelly (2011) clearly identified
their sampling method as purposive, convenience and
noted appropriate inclusion sampling criteria. A power
analysis was conducted to determine that the sample
size of 94 adolescent/parent dyads was adequate for
this study. There was no sample attrition noted in the
study, but it would have been helpful if the researchers
had discussed the refusal rate of potential participants.
The procedures for collecting data in this study were
clearly described as indicated in the following excerpt.
However, the researchers might have provided a little
more detail about the self-administration of the scales
by the adults and adolescents.
“All study procedures including protocols for recruit-
ing participants and obtaining informed consent had
been reviewed and approved by the appropriate
administrative and university institutional review
boards for the protection of human subjects prior to
study initiation.… The study protocol was presented,
questions addressed, and informed consent was
obtained from the parent as well as parents’ permis-
sion to approach their child regarding participation.
Next, parents self-administered a survey containing
demographic questions and several standardized
instruments including the Obesity Risk Knowledge
Scale (ORK-10; Swift et al, 2006) and the IPAQ.
Surveys were collected by the PI [principal investiga-
tor] immediately.
Following the collection of parent surveys, a
follow-up meeting was held with those adolescents
who met the criteria and whose parents signed con-
sents for their participation. Adolescents who signed
assent forms were then asked to self-administer both
the demographic survey and several standardized
measures including the ORK-10 and PACE+ MVPA
(Prochaska et al., 2001).” (Rutkowski & Connelly,
2011, pp. 52-53)
Evidence from the Data Analysis Process
The process of data analysis is an important factor in
evaluating the meaning of results. One important part
of this examination is to summarize the study weak-
nesses related to the data analysis process. Ask your-
self these questions concerning the meaning of your
results: Were the data checked to ensure that limited
or no errors occurred during data entry into the com-
puter? How many subjects have missing data, and
how was missing data managed to decrease the effects
on the study results? Were the analyses accurately
implemented and calculated? Were statistical assump-
tions violated? Were the statistics used appropriate for
the data? It is best to address these issues initially
before analyses are performed and again when com-
pleting the analyses and preparing the final report.
Researchers should consult with a biostatistician
during the planning of a study and during data analy-
sis to ensure the appropriateness of the statistical tests
selected. The biostatistician could also be helpful in
interpreting the results. Before submitting a study for
publication, we recommend rechecking each analysis
reported in the paper. We also encourage reexamining
the analysis statements in the article for accuracy and
clarity. Are you correctly interpreting the results of
the analysis? Documentation on each statistical value
or analysis statement reported in the paper is filed
with a copy of the article. The documentation includes
the date of the analysis, the page number of the com-
puter printout showing the results (the printout is
stored in a file by date of analysis), the sample size
for the analysis, and the number of missing values
(Corty, 2007; Grove, 2007; Fawcett & Garity, 2009).
Rutkowski and Connelly (2011) identified their data
analysis with a separate heading that is presented in
the following excerpt.
“Data Analysis
The sample size for the analysis is 94 dyads, which is
sufficient to detect a moderate standardized effect
size (ES) (d = 0.32) using a two-tailed significance test
with a power of 0.80 and a significance level of 0.05.
Data have been analyzed using the software package
Statistical Package for Social Sciences, Version 15 (SPSS,
Inc, 2008). Descriptive and multivariate statistics have
been used in this study.” (Rutkowski & Connelly,
2011, p. 54)
The data analysis section clearly addressed the
sample size of the study, analysis software package
used, and the types of analyses conducted. This section
594 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
would have been strengthened by including a power
analysis that discussed the effect size for relationships
because the study examined both differences between
the adolescents and their parents and relationships of
obesity risk knowledge with physical activities for
adolescents and adults (Aberson, 2010). This section
mentioned the inferential statistic conducted for group
differences, the two-tailed t-test, but did not mention
that relationships were examined with the Pearson
product-moment correlation and regression analysis
techniques.
Evidence from Data Analysis Results
The outcomes of data analysis are the most direct
evidence of the results. The researcher has intimate
knowledge of the research and needs to evaluate its
strengths and limitations carefully when judging the
validity of the results. In descriptive and correlational
studies, the validity of the results depends on how
accurately the variables were measured in selected
samples and settings. Interpretation of results from
quasi-experimental and experimental studies is often
based on decision theory, with five possible results:
(1) significant results that are in keeping with the
results predicted by the researcher, (2) nonsignificant
results, (3) significant results that oppose the results
predicted by the researcher, (4) mixed results, and
(5) unexpected results (Shadish, Cook, & Campbell,
2002).
Significant and Predicted Results
Significant results that coincide with the researcher’s
predictions are the easiest to explain and, unless
weaknesses are present, validate the proposed logical
links among the elements of the study. These results
support the logical links developed by the researcher
among the purpose, framework, questions, variables,
and measurement methods (Shadish et al., 2002).
This outcome is very satisfying to the researcher.
However, the researcher needs to consider alternative
explanations for the positive findings. What other ele-
ments could possibly have led to the significant
results?
Nonsignificant Results
Unpredicted nonsignificant or inconclusive results
are the most difficult to explain. These results are
often referred to as negative results. The negative
results could be a true reflection of reality. In this
case, the reasoning of the researcher or the theory
used by the researcher to develop the hypothesis is
in error. If so, the negative findings are an
important addition to the body of knowledge. With
nonsignificant results, it is important to determine if
adequate power of 0.8 or higher was achieved for the
data analysis. The researcher needs to conduct a
power analysis to determine if the sample size was
adequate to prevent the risk of a Type II error
(Aberson, 2010; Shadish et al., 2002). A Type II error
means that in reality the findings are significant, but
because of weaknesses in the methodology, the sig-
nificance was not detected.
Negative results could also be due to inappropriate
methodology, a deviant sample, problems with inter-
nal validity, inadequate measurement, use of weak
statistical techniques, or faulty analysis. Unless these
weak links are detected, the reported results could
lead to faulty information in the body of knowledge
(Angell, 1989). It is easier for the researcher to blame
faulty methodology for nonsignificant findings than to
find failures in theoretical or logical reasoning. If
faulty methodology is blamed, the researcher needs to
explain exactly how the breakdown in methodology
led to the negative results. Negative results, in any
case, do not mean that there are no relationships
among the variables or differences between groups;
they indicate that the study failed to find any relation-
ships or differences.
Significant and Not Predicted Results
Significant results that are the opposite of those pre-
dicted, if the results are valid, are an important addi-
tion to the body of knowledge. An example would be
a study in which the researchers proposed that social
support and ego strength were positively related. If the
study showed that high social support was related to
low ego strength, the result would be the opposite of
that predicted. Such results, when verified by other
studies, indicate that the theory being tested needs
modification and refinement. Because these types of
studies can affect nursing practice, this information is
important.
Mixed Results
Mixed results are probably the most common
outcome of studies. In this case, one variable may
uphold the characteristics predicted, whereas another
does not, or two dependent measures of the same
variable may show opposite results. These results
might be an accurate reflection of reality but also
might be due to study limitations (Shadish et al.,
2002). The study limitations might include differing
reliability or validity of two methods of measuring
variables, varying data collection processes, small
sample size, or missing subject data. Additional
research is indicated to examine mixed study results.
 CHAPTER 26  Interpreting Research Outcomes 595

Mixed results may also indicate a need to modify TABLE 26-2  Characteristics of the Adolescent-
existing theory. Parent Dyads
Unexpected Results Adolescent Parent
Unexpected results are relationships found between Age, M (SD), range 12.8 (1.0), 44.1 (5.2),
variables that were not hypothesized and not pre- 11.8-13.8 31-60
dicted from the framework guiding the study. These Ethnicity, n (%)
Caucasian 70 (74.5) 69 (74.2)
unexpected results are also called serendipitous
Hispanic 14 (14.9) 16 (17.2)
results. Most researchers examine as many elements Asian 5 (5.3) 6 (6.5)
of data as possible in addition to the elements directed Other 5 (5.3) 2 (2.2)
by the research objectives, questions, or hypotheses. Gender, n (%)
The researchers can use these findings to develop or Male 56 (59.6) 19 (20.2)
refine theories and to formulate later studies. In addi- Female 38 (40.4) 75 (79.8)
tion, serendipitous results are as important as evi- BMI, M (SD), range 19.4 (2.7), 24.1 (4.9),
dence in developing the implications of the study. 13.6-32.9 24.1-43.2
However, researchers must deal carefully with seren- Household size, M (SD), 4.41 (0.9), 4.41 (0.9),
dipitous results when considering their meaning range 2-6 2-6
Grade in school, n (%)
because the study was not designed to examine these
6th 17 (18.3)
results. 7th 28 (30.1)
8th 35 (37.6)
Evidence from Previous Studies 9th 8 (8.6)
The results of the present study should always be 10th 5 (5.4)
examined in light of previous findings. It is important Grade point average, n (%)
for the researcher to know whether the results are 2.0-2.9 10 (11.6)
consistent with past research. Consistency in findings 3.0-3.9 53 (61.6)
across studies is important for developing theories and 4.0 or higher 23 (24.5)
refining scientific knowledge for the nursing profes- Marital status, n (%)
Married 85 (91.4)
sion. Therefore, any inconsistencies need to be
Single 2 (2.2)
explored to determine reasons for the differences. Divorced 5 (5.4)
Replication of studies and synthesis of findings from Widowed 1 (1.1)
existing studies using meta-analyses and systematic Number of years with 17.5 (4.9),
reviews are critical for the development of empirical partner, M (SD), range 2-31
knowledge for an evidence-based practice (Brown, Education, n (%)
2009; Craig & Smyth, 2012; Melnyk & Fineout- Less than high school 2 (2.3)
Overholt, 2011). High school diploma 14 (16.1)
College 46 (51.7)
Graduate school 26 (29.9)
Determining Findings Work outside the home, 31 (33.3)
yes, n (%)
Findings are developed by evaluating evidence (dis-
cussed previously in this chapter) and translating and BMI, Body mass index; M (SD), mean (standard deviation).
interpreting study results. Although much of the From Rutkowski, E. M., & Connelly, C. D. (2011). Obesity risk knowledge
process of developing findings from results occurs in and physical activity in families of adolescents. Journal of Pediatric
Nursing, 26(1), 54.
the mind of the researcher, evidence of such thinking
can be found in published research reports (Pyrczak
& Bruce, 2005). It is important during this process to
talk with colleagues or mentors to clarify meanings or and presented the sample characteristics of the adoles-
expand implications of the research findings. The cents and their parents in Table 26-2. Rutkowski
study results and findings are presented for Rutkowski and Connelly (2011, p. 55) noted that “This sample is
and Connelly’s (2011) research examining the rela- not representative of families living in southern Cali-
tionships among the variables parent and adolescent fornia; however, it is representative of families living
physical activities and parent and adolescent obesity in the community sampled.” The study objectives
risk knowledge. As part of the study results, the (aims), results, and findings are presented in the fol-
researchers described the participants of the study lowing excerpt.
596 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
Objective 1 was to describe and compare the obesity
risk knowledge and physical activity level of the parents
and adolescents.
“Results. In this sample of adolescent/parent dyads,
the mean score for adolescents’ obesity risk knowledge
(ORK-10) is 4.69 (SD = 1.63), and the mean score for
the parents is 5.54 (SD = 1.84 [Table 26-3]). These
scores are based upon 10 being the score that reflects
the most knowledge and 0 reflecting the least knowl-
edge. t-tests have been conducted to determine
whether youth differ significantly from parents on the
ORK-10. The results indicated that adolescent obesity
risk knowledge is statistically significantly lower than
their parents t(1,92) = 3.45, p < .001.
Both the parents and adolescents are active at levels
measured to meet standardized guidelines (U.S. Depart-
ment of Health and Human Services [USDHHS], 2000).
The level of physical activity as measured by the PACE+
indicates that most adolescent participants are meeting
the recommended level of daily activity (M = 5.33, SD
= 1.23 [see Table 26-3]). Among the parent participants,
IPAQ scores range from 0 to 12,558 (M = 3741.07, SD
= 3135.4); categories of activity levels indicate 61.7%
report physical activity at a vigorous level (M = 1960,
SD = 2,164), 18.1% report a moderate level (M = 872,
TABLE 26-3  Obesity Risk Knowledge and
Physical Activity of Family Dyads
(N = 94)
Obesity Risk Knowledge, Adolescent Parent
4.69 (1.63) 5.54 (1.84)
  M (SD), range 0-10 0-10
Physical Activity
PACE+, M (SD) 5.33(1.23)
IPAQ, M (SD) 3741.07
(3135.4)
  Vigorous level, n (%) 58 (61)
  Moderate level, n (%) 17 (18.1)
  Low walking, n (%) 14 (14.9)
From Rutkowski, E. M., & Connelly, C. D. (2011). Obesity risk knowledge
and physical activity in families of adolescents. Journal of Pediatric
Nursing, 26(1), 55.
Objective 2 was to examine the relationships of
obesity risk knowledge and physical activities for the
parents and adolescents.
SD = 1,285), and 14.9% report physical activity at a low
level [see Table 26-3].” (Rutkowski & Connelly, 2011,
p. 55)
“Findings. The ORK-10 scores of both the adoles-
cents and parents are low, with adolescent scores sta-
tistically lower than parental scores. This is not
unexpected because parents have acquired more expe-
riences that would provide insight into some of the
questions included in this measure. When selecting the
ORK-10 for use in this study, most of the scale items
seemed to be relevant.… Despite this perceived rele-
vance, several of the participating youth questioned the
measure’s content and stated unfamiliarity with some
of the questions, particularly those items posed in the
negative, for example, ‘Obesity does not increase
the risk of developing high blood pressure.’ These
same youth identified words that, if substituted, would
increase the level of comprehension of the scale items,
which may consequently impact overall scores. An addi-
tional finding of interest indicates that in this group of
experts, both adolescent and parent mean scores are
higher than for the nonexpert group in the Swift et al.
(2006) original research.” (Rutkowski & Connelly, 2011,
pp. 55-56)
“Results. A statistically significant inverse relation-
ship has been found between parental ‘sitting’ activi-
ties (including commute time) and the adult ORK
scores (r = −.23, p < .05) as well as parent activity
levels and the PACE+ score for activity levels of ado-
lescents (r = −.23, p < .05). There is no significant
relationship between adolescent obesity risk knowl-
edge and adolescent physical activity (r = .15, p >
.05).” (Rutkowski & Connelly, 2011, p. 55)
“Findings. Among our sample of adolescent/parent
dyads, our findings indicate obesity risk knowledge is
not associated with adolescent physical activity. This
study does however provide important information
on the association between parent physical activity
and adolescent physical activity. Especially notable is
the inverse relationship between parent activity level
and the PACE+ scores of adolescents. This is note-
worthy because it is consistent with findings reported
a decade earlier by researchers from the School of
Public Health at the University of Minnesota (Luepker,
1999). They noted that seriously committed parents
who follow exercise routines that are vigorous in
 CHAPTER 26  Interpreting Research Outcomes 597
nature are frequently engaged in fitness activities that
involve only themselves.… Typically these parents do
not share their physical activity time with family
members, especially younger age children (Luepker,
1999). Many parents participate in their routines outside
of ‘family time’ and in essence exclude children pur-
posefully because they may impede the intensity and
duration goals of the parental activity. Luepker con-
cluded that parents who are moderate and less ‘serious’
are better role models because their children actually
see them being physical.… These parents are more apt
Objective 3 includes the variables of “adolescent and
parental obesity risk knowledge, parent physical activity,
and selected adolescent characteristics examined to
determine their prediction of the adolescent’s physical
activity” (Rutkowski & Connelly, 2001, p. 52).
“Results. Simultaneous multiple regression was con-
ducted to determine the accuracy of the independent
variables of parent physical activity, parent ORK-10
score, and adolescent ORK-10 score in predicting ado-
lescent physical activity [dependent variable] while con-
trolling for adolescent gender, ethnicity, and age.
Regression results indicate the overall model does not
significantly predict adolescent physical activity, R2 =
0.116, R2adj = 0.048, F(6,78) = 1.70, p = 0.13.” (Rutkowski
& Connelly, 2011, p. 55)
“Findings. The researchers did not discuss the non-
significant regression analysis results but did address the
low Cronbach alpha reliability coefficients for the
Forming Conclusions
Conclusions are derived from the study findings and
are a synthesis of findings. Forming conclusions for
a study requires a combination of logical reasoning,
creative formation of a meaningful whole from pieces
of information obtained through data analysis and
findings from previous studies, receptivity to subtle
clues in the data, and use of an open context in con-
sidering alternative explanations of the data.
When forming conclusions, it is important to
remember that research never proves anything; rather,
research offers support for a position. Proof is a logical
part of deductive reasoning, but not of the research
process. Therefore, formulation of causal statements
is risky. For example, the causal statement that A
causes B (absolutely, in all situations) cannot be sci-
entifically proved. It is more credible to state conclu-
sions in the form of conditional probabilities that are
to include their children in their physical activities. Seri-
ously active parents often work out early in the morning
before work, on lunch breaks, after work before return-
ing home, or at night after the children go to bed. Their
children never benefit from a role modeling relationship
in the area of physical activity (Bandura, 1997). As
explained by Luepker, the paradoxical situation of having
a very fit, athletic parent and a less than adequately
active child is often the result of this absent role model.”
(Rutkowski & Connelly, 2011, p. 55)
ORK-10 scale for the adolescents and their parents. The
low reliability of this scale for both groups might have
had an impact on the study results.
“It should be noted that the data from our sample of
family dyads have produced lower internal reliabilities
than the Swift et al. (2006) original study. This is not
unexpected because inconsistent internal reliabilities are
often found for instruments in ‘early development’
(Frank-Stromborg & Olsen, 2004). In our study, the
measure has been used with participants from a differ-
ent culture and age groups than those of the original
research. Based upon our findings, further refinement
of the measure is needed to avoid contextual confusion
and to increase cultural and international application.
Use of the ORK-10 is an initial step in assessing obesity
risk knowledge. Further research into the scale refine-
ment, reliability, and validity is warranted.” (Rutkowski
& Connelly, 2011, p. 56)
qualified. It would be more appropriate to state in the
study that if A occurred, then B occurred under con­
ditions x, y, and z (Kerlinger & Lee, 2000; Shadish
et al., 2002), or that if A occurred, then B had an 80%
probability of occurring. Because this was a descrip-
tive, predictive correlational study and some of the
results were nonsignificant, Rutkowski and Connelly
(2011) presented only tentative conclusions that are
included in the statement of findings. One conclusion
is the adults’ obesity risk knowledge is related to their
physical activity but the adolescents’ obesity risk
knowledge is not related to their physical activity. The
main study conclusion was that parents’ activity levels
(vigorous, moderate, or low) are inversely related to
the adolescents’ activity levels. Thus, children’s activ-
ity levels are most positively impacted when parents
role model physical activity and exercise with their
children.
598 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
One of the risks in developing conclusions in
research is going beyond the data—specifically,
forming conclusions that the data do not warrant. The
most common example is a study that examines rela-
tionships between A and B by correlational analysis
and then concludes that A causes B. Going beyond the
data is due to faulty logic and occurs more frequently
in published studies than one would like to believe.
Be sure to check the validity of arguments related to
conclusions before revealing findings.
Identifying Limitations
Limitations are restrictions or problems in a study
that may decrease the generalizability of the findings.
All studies have limitations, and these might be theo-
retical or methodological in nature. Theoretical limi-
tations are weaknesses in a study framework and
conceptual and operational definitions of variables
that restrict the abstract generalization of the findings.
At this stage of the research process, theoretical and
methodological limitations often overlap. Concerns
are raised about the theoretical basis of the study
because of unexpected findings. Some of the follow-
ing theoretical limitations might be identified by
researchers:
1. A concept might not be as clearly defined as needed
within the theory used as the study framework.
2. The relationships among some concepts from the
theory are not presented as clearly as needed in the
study framework.
3. A study variable might not be as clearly linked to
a concept in the framework as needed.
4. An objective, question, or hypothesis might not be
clearly linked to a relationship or proposition in the
study framework.
Theoretical limitations are one explanation for
study results that are nonsignificant or unexpected. For
example, the study framework included a relationship
between self-efficacy and health decisions. However,
the data analysis did not reveal a statistically signifi-
cant relationship between these two concepts. The
underlying reasons may be that the theory was unclear
related to the definitions of concepts or identification
of relationships or that the measurements of the con-
cepts were not valid or reliable.
Methodological limitations are weaknesses in
the study design that can limit the credibility of the
findings and restrict the population to which the find-
ings can be generalized. Methodological limitations
result from factors such as nonrepresentative samples,
weak designs, single setting, limited control over
treatment (intervention) implementation, instruments
with limited reliability and validity, limited control
over data collection, and improper use of statistical
analyses (Fawcett & Garity, 2009; Kerlinger & Lee,
2000; Shadish et al., 2002).
Ideally, the theoretical and methodological limita-
tions are identified before the conduct of the study, and
the researchers minimize these limitations as much as
possible. However, some limitations are not identified
until the study is conducted and are usually identified
in the discussion section of the study report with
implications of how they might have influenced the
study findings. Despite a researcher’s higher motives
to be objective, subjective judgments and biases some-
times creep into the study conclusions. Re­searchers
need to control subjectivity and biases and recognize
any theoretical or methodological limitations of their
study before forming conclusions. Limitations and
their impact on the study findings must be included in
the research report. Rutkowski and Connelly (2011)
clearly identified their study limitations, which did
limit the conclusions drawn in this study. The reli-
ability limitation of the ORK-10 scale was previously
discussed in the findings for Objective 3. In addition,
the study would have been strengthened by the conduct
of a power analysis for each of the nonsignificant
results to determine the power for detecting relation-
ships and the potential for a type II error. The other
study limitations are described in the following study
excerpt.
“Limitations
The findings of this study must be considered in rela-
tion to the study’s limitations. First, the sample is a
purposive convenience sample that is relatively
homogenous with respect to ethnicity, geographic
location, and social economic status (SES), not ran-
domly selected or matched. This nonrandom proce-
dure may influence the findings through self-selection
bias. Second, the cross-sectional design disallows for
changes over time and may not capture the complex
phenomena under study.” (Rutkowski & Connelly,
2011, p. 56)
Generalizing the Findings
Generalization extends the implications of the find-
ings from the sample studied to a larger population or
from the situation studied to a larger situation. For
example, if the study were conducted on diabetic
 CHAPTER 26  Interpreting Research Outcomes 599
patients, it may be possible to generalize the findings
to persons with other chronic illnesses or to well indi-
viduals. Studies such as randomized clinical trials
usually have large heterogeneous samples, which
increase the ability to generalize the findings (Shadish
et al., 2002).
How far can generalizations be made? The answer
to this question is debatable. From a narrow perspec-
tive, one cannot really generalize from the sample on
which the study was conducted. Any other sample is
likely to be different in some way. The conservative
position, represented by Kerlinger and Lee (2000),
recommends caution in considering the extent of gen-
eralization. Conservatives consider generalization par-
ticularly risky if the sample was small, homogeneous,
and not randomly selected. However, as discussed in
Chapter 15, generalizations are often made to abstract
or theoretical populations. Thus, conclusions need to
address applications to theory. Judgments about the
reasonableness of generalizing need to address issues
related to external validity, as discussed in Chapter 10.
Generalizations based on accumulated evidence
from many studies are called empirical generaliza-
tions. These generalizations are important for verify-
ing theoretical statements or developing new theories.
Empirical generalizations are the base of a science and
contribute to scientific conceptualization, which
provide a basis for generating evidence-based guide-
lines to manage specific practice problems (Brown,
2009; Craig & Smyth, 2012; Melnyk & Fineout-
Overholt, 2011). Chapter 19 provides a detailed dis-
cussion of research synthesis processes and strategies
for promoting evidence-based nursing practice.
Rutkowski and Connelly (2011) made no general-
izations in their study, and their findings were reflec-
tive only of the sample studied. Because this is a
relatively new area of research that was examined with
a descriptive, predictive correlational design, the lack
of generalization of findings seems appropriate. In
addition, the limitations of the ORK-10 scale reliabil-
ity, the nonrandom homogeneous sample, and the
cross-sectional design would not support generaliza-
tion of the findings from the study sample to the target
population (see Chapter 15 for more details on
generalization).
Considering Implications
Implications of research findings for nursing are
the meanings of the conclusions for the body of
nursing knowledge, theory, and practice. Implications
for practices are often based, in part, on whether treat-
ment decisions or outcomes would be different in
view of the study findings. A clinically important
study should result in altered decisions or actions by
nurses that improve patient and family outcomes
(Gatchel & Mayer, 2010). The study implications
provide specific suggestions for implementing the
findings in practice. You need to consider the areas
of nursing for which your study findings would be
useful. In presenting their implications for practice,
Rutkowski and Connelly (2011) discussed the role of
nurses in educating parents to be role models to their
children regarding physical activity.
“Clinical Implications
“Pediatric nurses, community health nurses, pediatric
nurse practitioners, school nurses, and clinical nurse
specialists all have opportunities to interface with
families of adolescents. Nurses practicing in primary
care settings are the first-line defenders against the
pandemic of childhood obesity.…
“Because prevention is the focus of primary care
settings, nurses have opportunities to implement
promising strategies with families of adolescents
(Larson & Story, 2007). Such encounters are ideal for
dissemination of the findings of this study in the area
of role modeling by parents for adolescents as a
means of impacting behavior change. Discussions
regarding the need for parental role modeling should
be encouraged.
“It is imperative for parents to acknowledge that
taking the time to be physically active with their chil-
dren may have a positive outcome in increasing the
activity levels of their children, which will reduce
future chronic health problems.… Children need
exemplars to establish lifestyle routines that include
physical activity. Parents are in an ideal position to be
such role models. The patterns of physical activity
developed in adolescents will inoculate them against
becoming sedentary adults (Luepker, 1999).” (Rut-
kowski & Connelly, 2011, p. 56)
Recommending Further
Research
Completing a study and examining its implications
should culminate in recommendations for further
research that emerge from the present study and from
previous studies in the same area of interest.
600 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
Suggested studies or recommendations for further
study may include replications or repeating the design
with a different or larger sample or different popula-
tion. In every study, the researcher gains knowledge
and experience that can be used to design “a better
study next time.” Formulating recommendations for
future studies will stimulate you to define more clearly
how to improve your study. From a logical or theoreti-
cal point of view, the findings should lead you directly
to more hypotheses to further test the framework you
are using. Improvements could involve an alternative
methodology, a refined measurement tool, changes in
sampling criteria, or a different setting.
Rutkowski and Connelly (2011) provided the fol-
lowing suggestions for future research:
“Directions for Future Studies
Future research examining obesity risk knowledge in
relationship to physical activity is needed to stem the
tide of the epidemic of childhood obesity. There is a
critical need for a reliable and valid measure to assess
the level of knowledge that both children and adults
have regarding the consequences of being overweight
(Swift et al., 2006). Behavior changes are the remedy
for this catastrophic health issue and will not be pos-
sible without the development of interventions that
are effective in reducing the weight-gaining lifestyles
of many Americans, especially children. In addition,
research is needed to learn what the public does not
know about obesity’s risks and its long-term effects.
To this end, research is currently being conducted in
the refinement of the ORK-10 to enhance its applica-
tion with multicultural and international populations.
A reliable and valid measure could become the
impetus in guiding discussion in primary care settings
regarding the choices in lifestyle to avoid obesity and
the concomitant illnesses that are undermining the
healthy futures of adolescents and their families.”
(Rutkowski & Connelly, 2011, p. 56)
KEY POINTS
• To be useful, evidence from data analysis needs to
be carefully examined, organized, and given
meaning; and this process is referred to as
interpretation.
• Interpretation includes several intellectual activi-
ties, such as examining evidence, forming conclu-
sions, identifying study limitations, generalizing
the findings, considering implications, and suggest-
ing further research.
• The first step in interpretation is considering all of
the evidence available that supports or contradicts
the validity of the results. Evidence is obtained
from various sources, including the research plan,
measurement reliability and validity (or precision
and accuracy), data collection process, data analy-
sis process, data analysis results, and previous
studies.
• The outcomes of data analysis are the most direct
evidence available of the results related to the
research purpose and the objectives, questions, or
hypotheses.
• Five possible results are (1) significant results that
are in keeping with those predicted by the
researcher, (2) nonsignificant results, (3) significant
results that are opposite those predicted by the
researcher, (4) mixed results, and (5) unexpected
results.
• Findings are a consequence of evaluating evidence,
which includes the findings from previous studies.
• Conclusions are derived from the findings and are
a synthesis of the findings.
• The limitations of a study might be theoretical or
methodological and need to be clearly identified
and discussed in relation to the study conclusions.
• Generalization extends the implications of the find-
ings from the sample studied to a larger target
population.
• Implications of the study for nursing are the mean-
ings of study conclusions for the body of nursing
knowledge, theory, and practice.
• Completion of a study and examination of implica-
tions should culminate in recommending future
studies that emerge from the present study and pre-
vious studies.
REFERENCES
Aberson, C. L. (2010). Applied power analysis for the behavioral
sciences. New York, NY: Routledge Taylor & Francis Group.
Angell, M. (1989). Negative studies. New England Journal of Medi-
cine, 321(7), 464–466.
Bandura, A. (1997). Self-efficacy: The exercise of control. New
York, NY: W.H. Freeman & Company.
Brown, S. J. (2009). Evidence-based nursing: The research-practice
connection. Sudbury, MA: Jones & Bartlett Publishers.
Corty, E. W. (2007). Using and interpreting statistics: A practical
text for the health, behavioral, and social sciences. St. Louis, MO:
Mosby.
Craig, J. V., & Smyth, R. L. (2012). The evidence-based practice
manual for nurses (3rd ed.). Edinburgh, UK: Churchill
Livingstone.
 CHAPTER 26  Interpreting Research Outcomes 601
Fawcett, J., & Garity, J. (2009). Evaluating research for evidence-
based nursing practice. Philadelphia, PA: F. A. Davis.
Frank-Stromborg, M., & Olsen, S. (2004). Instruments for clinical
healthcare research (3rd ed.). Sudbury, MA: Jones & Bartlett.
Gatchel, R. J., & Mayer, T. G. (2010). Testing minimal clinically
important difference: Consensus or conundrum? The Spine
Journal, 35(19), 1739–1743.
Grove, S. K. (2007). Statistics for health care research: A practical
workbook. St. Louis, MO: Saunders.
International Physical Activity Questionnaires (IPAQ). (2001).
International Physical Activity Questionnaire. Retrieved from
http://www.ipaq.ki.se/ipaq.htm.
Kerlinger, F. N., & Lee, H. P. (2000). Foundations of behavioral
research (4th ed.). Fort Worth, TX: Harcourt College.
Larson, N., & Story, M. (2007). The pandemic of obesity among
children and adolescents: What actions are needed to reverse
current trends? [Editorial]. Journal of Adolescent Health, 41,
521–522.
Luepker, R. (1999). How physically active are American children
and what can we do about it? International Journal of Obesity,
23(Suppl. 2), S12–S17.
Melnyk, B. M., & Fineout-Overholt, E. (2011). Evidence-based
practice in nursing & healthcare: A guide to best practice (2nd
ed.). Philadelphia, PA: Lippincott Williams & Wilkins.
Prochaska, J., Sallis, J., & Long, B. (2001). A physical activity
screening measure for use with adolescents in primary care.
Archives of Pediatric & Adolescent Medicine, 155(5), 554–559.
Pyrczak, F., & Bruce, R. R. (2005). Writing empirical research
reports (5th ed.). Glendale, CA: Pyrczak.
Rutkowski, E. M., & Connelly, C. D. (2011). Obesity risk knowl-
edge and physical activity in families of adolescents. Journal of
Pediatric Nursing, 26(1), 51–57.
Shadish, W. R., Cook, T. D., & Campbell, D. T. (2002). Experimen-
tal and quasi-experimental designs for generalization causal
inference. Chicago, IL: Rand McNally.
Statistical Package for the Social Sciences (SPSS), Inc. (2008).
SPSS for windows. Chicago, IL: SPSS.
Stein, K. F., Sargent, J. T., & Rafaels, N. (2007). Intervention
research: Establishing fidelity of the independent variable in
nursing clinical trials. Nursing Research, 56(1), 54–62.
Swift, J., Glazebrook, C., & Mcdonald, I. (2006). Validation of a
brief, reliable scale to measure knowledge about the health risks
associated with obesity. International Journal of Obesity, 30(4),
661–668.
Thompson, S. K. (2002). Sampling (2nd ed.). New York, NY: John
Wiley & Sons.
U.S. Department of Health and Human Services (USDHHS).
(2000). Health people 2010. Washington, D.C.: U.S. Government
Printing Office.
Waltz, C. F., Strickland, O. L., & Lenz, E. R. (2010). Measurement
in nursing and health research (4th ed.). New York, NY: Springer
Publishing Company.

27
CHAPTER
Disseminating Research Findings
I
magine that as a nurse researcher you are con­
ducting a study in which you describe a unique
phenomenon, detect a previously unrecognized
relationship, or determine the effectiveness of an inter­
vention. This information might make a difference in
nursing practice; however, you feel unskilled in pre­
senting the information and overwhelmed by the idea
of publishing. You place the study documents in a
drawer with the intent to communicate the findings
someday. Because many nurses have these types of
feelings, many valuable nursing studies are not com­
municated, and the information is lost. Failure to com­
municate research findings is considered a violation
of ethics and a form of scientific misconduct. After
involving members of an institutional review board
committee to approve your study, and after subjects
consented and participated in your study, you have an
ethical obligation to complete your end of the process.
Without disseminating the findings, funding is wasted,
subject time and data are wasted, and knowledge for
nursing practice is not advanced.
Communicating research findings, the final step
in the research process, involves developing a research
report and disseminating it through presentations and
publications to audiences of nurses, healthcare profes­
sionals, policy makers, and healthcare consumers.
Disseminating study findings provides many advan­
tages for the researcher, the nursing profession, and
the consumer of nursing services. By presenting and
publishing findings, researchers advance the knowl­
edge of a discipline, which is essential for providing
evidence-based practice. For individual researchers,
communicating study findings often leads to pro­
fessional advancement, personal recognition, and
other psychological and financial compensations.
These rewards are extremely important for the con­
tinuation of research in a discipline. By communicat­
ing research findings, the researcher also promotes the
critical analysis of previous studies, encourages the
602
  http://evolve.elsevier.com/Grove/practice/
replication of studies, and identifies additional research
problems. Over time, the findings from many studies
are synthesized with the ultimate goal of providing
evidence-based health care to patients, families, and
communities (Craig & Smyth, 2012; Melnyk &
Fineout-Overholt, 2010).
To facilitate the communication of research find­
ings for nurse clinicians and researchers, regardless of
whether the research is a primary study, a secondary
analysis, or a meta-analysis, this chapter describes the
basic content of a research report, the audiences for
communicating study findings, and the processes for
presenting and publishing research reports.
Content of a Research Report
Both quantitative and qualitative research reports
include four major sections or content areas: (1) intro­
duction, (2) methods, (3) results, and (4) discussion
of the findings (Pyrczak & Bruce, 2007). The type
and depth of information included in these sections
depend on the study, the intended audiences, and
the mechanisms for disseminating the report. For
example, theses and dissertations are research reports,
written to demonstrate the student’s depth in under­
standing of the research problem and process to
faculty members. Research reports developed for
publication in journals are written to communicate
study findings efficiently and effectively to nurses
and other healthcare professionals. Some journals
limit the introduction to two or three brief paragraphs
that include a statement about the theoretical frame­
work for the study, a sufficient review of the literature
to state what the gap in knowledge is, and the clear
purpose of the study. The methods, results, and dis­
cussion sections of qualitative studies are usually
more detailed than those sections of quantitative
studies because of the complex data collection and
analysis procedures. Finally, the discussion section
 CHAPTER 27  Disseminating Research Findings 603
briefly acknowledges the limitations of the study,
presents the findings in relation to other literature,
and discusses the implications of the findings for the
intended journal audience.
Quantitative Research Report
This section provides direction to novice researchers
writing their first quantitative research report. To
begin, the title of your research report needs to indi­
cate what you have studied and attract the attention of
interested readers. The title should be concise and
consistent with the study purpose and the research
objectives, questions, or hypotheses. Often a title
includes the major study variables and population and
indicates the type of study conducted, but it should not
include the results or conclusions of a study (Pyrczak
& Bruce, 2007). Heo, Moser, Lennie, Zambroski, and
Chung (2007, p. 16) provided the following title for
their study: “A Comparison of Health-Related Quality
of Life between Older Adults with Heart Failure and
Healthy Older Adults.” This title is concise, states the
focus of the study (comparative descriptive), identifies
a key study variable (health-related quality of life
[HRQOL]), and includes the populations studied
(older adults with heart failure and healthy older
adults). However, this study is also predictive, and this
is not indicated in the study title. The researchers
studied additional independent variables (health per­
ception, functional status, physical symptom status,
emotional symptom status, and social support) to
predict the dependent variable HRQOL in older adults
with and without heart failure. An alternative title
would be “Predictors of Health-Related Quality of
Life in Older Adults with and without Heart Failure.”
Some journals limit the length for titles of manu­
scripts, whereas other journals discourage use of
colons.
Most research reports also include an abstract that
summarizes the key aspects of the study. An abstract
is usually about 200 to 300 words and describes the
problem, purpose, framework, methods, sample size,
key results, and conclusions (Pyrczak & Bruce, 2007).
We provide details for developing an abstract of a
study later in this chapter. Heo et al. (2007) included
the following abstract for their study.
“Background: Health-related quality of life (HRQOL)
in older adults with heart failure may be affected by
a variety of variables, including aging. It is important
to determine the unique impact of heart failure to
more effectively improve HRQOL in this population
[problem].
Objective: The purpose of this study was to compare
HRQOL and physical, psychologic, clinical, and
sociodemographic status in older adults with and
without heart failure.
Methods: The HRQOL of 90 older adults with heart
failure and 116 healthy older adults was compared.
The factors best associated with HRQOL in each
group were determined using multiple regression
model.
Results: HRQOL was substantially worse among
older adults with heart failure than among healthy
older adults. Older adults with heart failure had more
severe physical and emotional symptoms, poorer
functional status, and worse health perceptions. Phys-
ical symptom status was the strongest predictor of
HRQOL in both groups. In addition, in older adults
with heart failure, physical symptom status, age, and
anxiety were related to HRQOL.
Conclusions: The poor HRQOL seen in patients
with heart failure is not just a reflection of aging.
Comprehensive interventions targeted toward the
factors that specifically negatively impact HRQOL are
essential in older adults with heart failure.” (Heo
et al., 2007, p. 16)
Heo et al. (2007) concisely organized their abstract
with headings and clearly indicated the problem
(background), sample size (90 older adults with heart
failure and 116 healthy older adults), results, and con­
clusions. The purpose of the study was clearly stated
under the heading “Objective” as required in the jour­
nal’s instructions. The “Methods” section was less
clear on the type of study design, variables studied,
measurement methods, and data collection process.
This abstract is a strong abstract but could have been
improved if the authors had been allowed to expand
the “Methods” section.
Following the abstract are the four major sections
of a research report: introduction, methods, results,
and discussion. Table 27-1 provides an outline of
the content covered in each of these sections for a
quantitative research report. The research report by
Heo et al. (2007) is used as an example when discuss­
ing these sections. The complete research article
can be accessed through the Cumulative Index to
Nursing and Allied Health Literature (CINAHL).
Also, because Heo et al. (2007) received funding from
the National Institutes of Health to conduct the study,
604 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
TABLE 27-1  Outline for a Quantitative
Research Report
Introduction
Background and significance of the problem
Purpose of study
Brief review of relevant literature (may include theoretical
framework and conceptual definitions)
Gap in knowledge study will address
Research objectives, questions, or hypotheses
Methods
Research design and intervention if applicable
Setting
Sampling method, consent process
Measurement methods (instrument descriptions and scoring)
Data collection process
Data analysis
Results
Description of sample (may use tables or figures)
Results organized by objectives, questions, or hypotheses
Use narrative, tables, and figures to present results
Discussion
Major findings compared with previous research
Limitations of study
Conclusions
Implications
Future studies that are needed
References
Include references cited in paper, using format specified by
journal
the publication is also available at no cost from
PubMed Central (http://www.ncbi.nlm.nih.gov/pmc/
articles/).
Introduction
The introduction of a research report discusses the
background and significance of the problem; identi­
fies the problem statement and purpose, reviews the
relevant empirical and theoretical literature, describes
the study framework, and identifies the research
purpose (aims, objectives, questions, or hypotheses
if applicable). You will have developed this content
for the research proposal; you summarize it in the
final report. Depending on the type of research report,
the review of literature and framework might be
separate sections or separate chapters as in a thesis
or dissertation. Key content from the introduction of
the study by Heo et  al. (2007) is presented as an
example.
“Despite advances in treatment and care, approxi-
mately five million people have heart failure in the
United States. The number is increasing each year
despite high mortality [significance].… In patients with
heart failure, health-related quality of life (HRQOL) is
an important outcome that is closely related to clinical
outcomes including rehospitalization and even mor-
tality [background].… It is essential to compare these
variables and their effects on HRQOL between older
adults with and without heart failure to determine the
unique impact of heart failure and more effectively
improve HRQOL in older adults with heart failure
[problem statement].
The purpose of this study was to compare HRQOL
between older adults with and without heart failure.
The specific aims [objectives] were to (1) compare
physical, psychologic, and social variables in older
adults with heart failure with those in healthy older
adults, and (2) determine the best model predicting
HRQOL in each group from among the physical, psy-
chologic, and social variables.” (Heo et al., 2007, pp.
16-17)
Review of Literature
The review of literature section of a research report
documents the current knowledge of the problem
investigated. The sources included in the literature
review are the sources that you used to develop your
study and interpret the findings. A review of literature
can be two or three paragraphs or several pages long.
In journal articles, the review of literature is concise
and usually includes 15 to 20 sources. Theses and
dissertations frequently include an extensive literature
review to document the student’s knowledge of the
research problem. The summary of the literature
review clearly identifies what is known, what is not
known or the gap in knowledge, and the contribution
of this study to the current knowledge base. The objec­
tives, questions, or hypotheses that were used to direct
the study often are stated at the end of the literature
review. Heo et al. (2007) provided a brief summary of
the relevant literature and included what is known and
not known about the effects of heart failure on
HRQOL. See Chapter 6 for more information on
writing a review of the literature.
Framework
A research report needs to include an explicitly identi­
fied framework. In this section, you identify and define
the major concepts in the framework and describe the
relationships among the concepts (see Chapter 7). You

Biological/
physiological
status
Physical and
emotional
Individual
character-
symptom Health-
status perception
istics
Functional
status
Environ-
mental
character-
istics
(social
support)
Figure 27-1  Conceptual framework. Factors affecting health-related
quality of life. From Heo, S., Moser, D. K., Lennie, T. A., Zambroski, C.
H., & Chung, M. L. (2007). A comparison of health-related quality of life
between older adults with heart failure and healthy older adults. Heart
& Lung, 36(1), 107.
can develop a schematic map or model to clarify the
logic within the framework. If a particular proposition
or relationship is being tested, that proposition should
be clearly stated. Developing a framework and iden­
tifying the proposition or propositions examined in a
study connect the framework and the research purpose
to the objectives, questions, or hypotheses. The con­
cepts in the framework need to be linked to the study
variables and are used to define the variables concep­
tually. Heo et al. (2007, p. 17) clearly identified the
framework of their study as the Wilson and Cleary
(1995) HRQOL model (see Figure 27-1). The vari­
ables were conceptually and operationally defined and
presented in a table that is included later in this chapter.
If the journal allowed more space, the researchers
might have expanded the literature review section
based on this framework to include effects of age,
health perception, functional status, physical symptom
status, emotional symptom status, and social support
on HRQOL in individuals with and without heart
failure.
Methods
The methods section of a research report describes
how the study was conducted. This section needs to
be concise yet provide sufficient detail for nurses to
appraise critically or replicate the study procedures. In
this section you will describe the study design, sample,
setting, methods of measurement, data collection
process, and plan for data analysis. If the research
project included a pilot study, the planning, implemen­
tation, and results obtained from the pilot study are
briefly described. You will also describe any changes
made in the research project based on the pilot study
(Pyrczak & Bruce, 2007).
CHAPTER 27  Disseminating Research Findings 605
Design
The study design should be explicitly stated. Heo et al.
(2007) implied a descriptive comparison study by indi­
Health- cating that their purpose was to compare two groups,
related and this was reinforced with their first aim—to
quality “compare physical, psychologic, and social variables
of life
in older adults with heart failure with those in healthy
older adults” (Heo et al., 2007, p. 17). The design
should match with the stated purpose and link to the
planned analysis. For example, it is clear that one
group is older adults with heart failure and they will
be compared with a group of older adults who are
healthy. The two groups will be compared on three
types of variables (physical, psychologic, and social).
However, the abstract also includes HRQOL as a vari­
able. The second aim—to “determine the best model
predicting HRQOL in each group from among the
physical, psychologic, and social variables” (Heo
et al., 2007, p. 17)—is actually a different study design.
It is still a descriptive design rather than an intervention
or experiment. However, the word “predictive”
assumes a longitudinal design with predictor variables
measured at an initial point in time and an outcome
variable measured at a later point in time. The type of
design is most accurately described as a longitudinal
descriptive correlational study. From the way in which
the second aim is written, the reader can expect to see
two predictive models tested—one for the healthy
group and one for the group with heart failure.
A descriptive correlational study design could be
longitudinal, as in the case just described, or it could
be a cross-sectional descriptive correlational study. A
cross-sectional study design includes the same vari­
ables and the same type of analysis (correlations and
model testing), but data are collected at only one point
in time to describe correlates of HRQOL. In both
longitudinal and cross-sectional descriptive study
designs, the researchers test the correlations or rela­
tionships between variables and test models with a
particular outcome of interest, but conclusions can be
interpreted only as associations or relationships
between variables rather than causal.
If your design includes an intervention or treatment,
your report needs to describe the treatment, including
the protocol for implementing the treatment, training
of people (interventionalists) to implement the proto­
col, and a discussion of the consistency of administra­
tion of the treatment (Santacroce, Maccarelli, & Grey,
2004). The reliable and competent implementation of
an experimental treatment is referred to as intervention
fidelity. As discussed in Chapter 14, intervention
fidelity includes two core components: (1) adherence
to the delivery of the prescribed treatment behaviors,
606 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
session, or course and (2) competence of the researcher
or interventionalist in delivering the intervention
(Stein, Sargent, & Rafaels, 2007). Santacroce et al.
(2004) and Stein et al. (2007) provide detailed direc­
tions to promote intervention fidelity in various quasi-
experimental studies and clinical trials.
Heo et al. (2007, p. 17) implemented a combined
comparative descriptive and predictive design to
accomplish the two aims of their study. A more
complex study design, such as a randomized clinical
trial, might be presented in a table or figure, such as
the examples provided in Chapter 11. Most journals
require that these more complex studies be depicted
in a flow chart using the CONSORT guidelines
(Bennett, 2006) as discussed in Chapter 20.
Sample and Setting
This section of the research report should describe the
sampling method, criteria for selecting the sample,
sample size, and sample characteristics. When a clini­
cal trial or experiment is involved, the report must also
address the statistical power analysis used to deter­
mine how many subjects per group would be needed
to find a statistically significant difference if signifi­
cance is set at α ≤ 0.05 or another alpha level. If fewer
subjects enroll or complete the study than what
was indicated in the original power analysis, the dif­
ference may not be statistically significant even if
the same group difference is clinically relevant. This
is known as a Type 2 statistical error if the researchers
conclude that there was no difference between the
two groups when the same difference would be statis­
tically significant if the sample had been larger (see
Chapter 15).
Details about subject recruitment, including refusal
or acceptance rates, should be reported. The number
of subjects completing the study should be provided
if it differs from the initial sample size, and attrition
or retention rate needs to be addressed. If your subjects
were divided into groups (experimental and compari­
son or control groups), identify the method for assign­
ing subjects to groups and the number of subjects in
each group. The protection of subjects’ rights and the
process of informed consent should be explicitly
stated. In a published study, the setting is often
described in one or two sentences, and agencies are
not identified by name unless permission has been
obtained. Researchers can present the sample charac­
teristics in narrative format; however, most research­
ers present the characteristics of their sample in a table
(see Chapter 8). Although some researchers report the
subject characteristics in this section, these are often
presented in the first part of the results section. Heo
et al. (2007) used the heading “Sample and Setting”
to introduce the following description of their sam­
pling process.
“Patients hospitalized with an exacerbation of heart
failure at one of the three hospitals in a large Mid-
western city [setting] were screened during admis-
sion for inclusion in the study [convenience sample].
Patients aged more than 55 years were included. The
other inclusion criteria were as follows: (1) a primary
diagnosis of heart failure with either preserved or
nonpreserved left ventricular systolic function; (2)
New York Heart Association (NYHA) functional clas-
sification II to IV; (3) discharged home; and (4) living
within the greater metropolitan area. Exclusion crite-
ria were as follows: (1) discharged to an extended
care facility; (2) referred for hospice or home care;
(3) referred to cardiac rehabilitation; or (4) cognitive
or psychiatric problems [inclusion and exclusion
criteria].
“Healthy older adults were recruited from three
local senior centers [setting] by using flyers and word
of mouth [sample of convenience]. Individuals aged
more than 55 years were included. Those with the
following characteristics were excluded from partici-
pation: (1) cognitive or psychiatric problems that pre-
cluded informed consent; (2) diagnosis of heart
disease or heart failure; and (3) other serious chronic
illnesses including stroke, chronic lung diseases, and
cancer. Older adults with the common comorbidities
of hypertension and diabetes were not excluded so
individuals with characteristics typical of older adults
that are also seen in older adults with heart failure
could be included [inclusion and exclusion criteria].”
(Heo et al., 2007, p. 17)
The researchers identified inclusion and exclusion
sample criteria used to determine the target popula­
tions for the study (older adults with and without heart
failure). They also clearly indicated the sample size of
90 older adults with heart failure and 116 healthy older
adults. This sample may be sufficient for a descriptive
study, but the adequacy of the sample size would have
been better justified for the group comparisons identi­
fied in aim 1 if the researchers had conducted a power
analysis. By identifying the refusal and attrition rates
for the study, the researchers would have increased the
readers’ understanding about the representativeness of
the sample and ability to generalize to the larger popu­
lation. The more people who refuse or drop out of the

study, the less one can generalize to the target popula­
tion. The sampling method appeared to be a sample of
convenience, but it was not clearly stated. A conve­
nience sampling method increases the potential for
sampling error and decreases the ability of the
researchers to generalize the study findings to the
larger population of interest. Recruitment from three
local senior centers and three hospital sites increases
the likelihood that findings can be generalized to
the larger population of older adults, and at least the
samples are representative of the local area in the
Midwest United States.
Measures
This section of the report describes the measurement
methods, which is how variables were operationalized
and measured in the study. Details about the measures
or instruments used in the data collection process are
crucial if nurses are to appraise critically and replicate
a study. The details include the measure’s scaling and
range of scores and the frequency with which the
instrument was used; any reliability and validity infor­
mation previously published on the instrument should
also be provided. In addition, the report needs to
include the instrument’s reliability in the current study
and any further support of validity for the current
study’s sample. If you have used physiological mea­
sures, be sure to address their accuracy, precision,
selectivity, sensitivity, and sources of error (Pyrczak
TABLE 27-2  Variables and Instruments
Variable (Conceptual Definition)
Health-related quality of life (perception of
effects of a clinical condition or its
treatment on daily life)
Health perception (perception of overall health)
Functional status (physical functional
impairments in daily activities)
Physical symptom status (dyspnea and fatigue)
Emotional symptom status (anxiety and
depression)
Biological/physiological status (number of
comorbidities)
Individual characteristic (age and gender)
Environmental characteristics (social support)
SF-36, 36-item short-form health survey.
From Heo, S., Moser, D. K., Lennie, T. A., Zambroski, C. H., & Chung, M. L. (2007). A comparison of health-related quality of life between older adults
with heart failure and healthy older adults. Heart & Lung, 36(1), 18.
CHAPTER 27  Disseminating Research Findings 607
& Bruce, 2007). In a section titled “Measures,” Heo
et al. (2007, pp. 17-19) described all of their measures
in detail. Each variable was clearly identified, concep­
tually defined and linked to the framework, and
operationally defined to indicate how it was measured.
The conceptual and operational definitions (instru­
ments used) were clearly presented in a table (Table
27-2). These researchers provided an excellent link
between the framework and the methodology of their
study.
Data Collection Process and Procedures
The description of the data collection process in the
research report details who collected the data, the pro­
cedure for collecting data, and the type and frequency
of measurements obtained. In describing who col­
lected the data, your report needs to specify the experi­
ence of the data collector and any training provided.
If more than one person collected data, describe the
precautions taken to ensure consistency and interrater
reliability (Pyrczak & Bruce, 2007).
Heo et al. (2007) detailed their data collection
process in a section of their report titled “Procedure,”
which is presented in the following excerpt. The data
collection process was clear and concise, institutional
review board approval was indicated, subjects’
informed consent was described, use of trained data
collectors was addressed, and timing and setting were
discussed.
Instrument
Older Adults with Heart Failure Healthy Older Adults
Minnesota Living with Heart Failure
One item from SF-36
Composite measure from New York Composite measure from Duke
Heart Association functional class Activity Status Index
Composite measure from Dyspnea Composite measure from two items
and Fatigue Index (dyspnea and fatigue) of Memorial
Symptom Assessment Scale
Subscales (depression and anxiety)
of Brief Symptom Inventory
Clinical characteristics questionnaire
Demographic questionnaire
Marital status and having a confidant
by demographic questionnaire
608 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
“Procedure
Institutional review board approval was obtained for the
conduct of this study. Eligible older adults with heart
failure were identified by trained nurse research assis-
tants, and those who gave written, informed consent to
participate in the study were included. All data on
HRQOL, health perception, functional status, physical
symptom status, emotional symptom status, biologic/
physiologic status, individual characteristics, and envi-
ronment characteristics (social support) were collected
by nurse research assistants after hospital discharge [see
Table 27-2]. Eligible healthy older adults were identified
by investigators among older adults who appeared at
Analysis Plan
Heo et al. (2007) provided an excellent description of
the analysis techniques used to obtain their results.
“Data were analyzed using SPSS (Statistical Packages
for the Social Sciences) for Windows (version 12.0,
SPSS Inc., Chicago, IL). Descriptive statistics including
mean, standard deviation, frequency, and percentage
were used to present demographic and clinical char-
acteristics. Mann-Whitney U test, t-test, or chi-square
test was used to examine differences in individual
characteristics and biologic/physiologic status in older
adults with heart failure. To address specific aim 1,
the Mann-Whitney U test was used to determine the
difference in HRQOL, health perception, functional
status, physical and emotional symptom status, and
environmental characteristics (social support) because
the distribution of each variable in healthy older adults
did not show normality. To address specific aim 2,
stepwise multiple regression was used to identify
variables as a group best associated with HRQOL in
each group. A p value of less than 0.05 was consid-
ered statistically significant.” (Heo et al., 2007, p. 20)
Results
The results section reveals what was learned from
the study and includes the results generated from
your statistical analyses. The results section is best
organized by the research objectives, questions, or
hypotheses that are linked to the conceptual frame­
work. Heo et al. (2007) organized their results by
characteristics of the sample and the two research
aims. Research results can be presented in narrative
three senior centers. Investigators explained the
research purpose and procedures, obtained informed
consent, and answered participants’ questions. All gave
written, informed consent. Questionnaires were com-
pleted at the senior centers [see Table 27-2].
Completion of instruments required 30 to 45 minutes,
and trained research nurses were available to assist both
older adults with heart failure and healthy older adults.
In addition, the instruments were checked to make sure
participants did not inadvertently leave any items unan-
swered. Participants were free to leave items unan-
swered.” (Heo et al., 2007, p. 20)
format and organized into figures and tables. Table 1
is typically the descriptive characteristics of the
sample. Because Aim 1 was to compare the healthy
group with the heart failure group, the reader would
expect to see sample demographic and clinical char­
acteristics described in two columns (healthy group
and heart failure group samples). The reader would
also expect to see the statistics showing any group
differences in these characteristics. The presentation
of results depends on the end product of the data
analysis and your own preference and any journal
instructions. Generally, what is presented in a table is
not restated in the text of the narrative. When reporting
results in a narrative format, the value of the calculated
statistic (t, F, r, or χ2), the degrees of freedom (df), and
probability (p value) should be included. When report­
ing any nonsignificant results, it is important to include
the effect size and power level for that analysis so that
readers would be able to evaluate the risk of a Type II
error (see Chapter 15).
Students often have difficulty putting all of these
Greek-letter statistical findings back into words for the
text of the results section. Pallant (2007) provides
useful and specific examples of how to translate the
Greek symbol results back into English for the text of
your report. The Publication Manual of the American
Psychological Association (American Psychological
Association [APA], 2010) provides direction for how
to present various statistical results in a research
report. The format for reporting chi-square results is a
symbol χ2 with degrees of freedom and sample size in
parentheses followed by the χ2 statistic value and p
value. For example, when APA format is required, a
chi-square value should be presented in the following
format in the text of a research report: χ2 (4, N = 90)
= 11.14, p = 0.025. Statistical values need to be
 CHAPTER 27  Disseminating Research Findings 609
reported with two decimal digits of accuracy. Although
a computer output of data may include results reported
to several decimal places, this is unnecessary for the
report. For example, reporting the χ2 value as 11.14 is
sufficient even if the computer output says 11.13965
(APA, 2010). The p value should be reported as the
exact value, but when the computer output says p =
0.0000, it should be reported as p < 0.001 because the
computer calculates the value only out so far before it
rounds off to zero.
Heo et al. (2007) presented their results in narrative
and table formats. Some of these results are repre­
sented in the next section as example tables. Because
journals have space limitations, results are best
depicted in tables and should not duplicate what is in
the text.
Presentation of Results in Figures and Tables
Figures and tables are used to present a large
amount of detailed information concisely and clearly.
Researchers use figures and tables to demonstrate rela­
tionships, document change over time, and reduce the
number of words in the text of the report (APA, 2010;
Saver, 2006). However, figures and tables are useful
only if they are appropriate for the results you have
generated and if they are well constructed (Saver,
2006). Table 27-3 provides guidelines for developing
accurate and clear figures and tables for a research
report. More extensive guidelines and examples for
developing tables and figures for research reports can
be found in the Publication Manual of the American
TABLE 27-3  Guidelines for Developing Tables
and Figures in Research Reports
Select results to include in report
Identify a few key tables and figures that explain or support
major points
Develop simple tables and figures
Ensure tables and figures are complete and clear without
referring to the narrative
Give each table or figure a brief title
Tables and figures are numbered separately in the report (e.g.,
Table 1, 2; Figure 1, 2)
Use descriptive headings, labels, and symbols—may need to
provide a key for abbreviations or symbols
Include probability values as actual values or indicate whether
less or more stringent by asterisks
Refer to each table and figure in the narrative (e.g., Table 1
presents…)
Use narrative to summarize main ideas, but do not repeat
specifics that are in figures and tables
Compiled from APA, 2010; Pallant, 2007; and Pyrczak & Bruce, 2007.
Psychological Association (APA, 2010). For meta-
analysis reports that synthesize the results of many
studies, particular figures, called forest plots, are very
important in the presentation of results (Floyd, Galvin,
Roop, Oermann, & Nordstom, 2010).
Figures
Figures are diagrams or pictures that illustrate
the results. Researchers often use computer programs
to generate sophisticated black-and-white or color
figures. Common figures included in nursing research
reports are bar graphs and line graphs. Journals often
require high-resolution images for reproduction. The
APA manual (APA, 2010, p. 167) has a figure checklist
for you to review when deciding whether or not to
include a figure. Generally, figures require specific
formatting and may have less detail than readers want,
so potential authors should carefully check with the
journal guidelines (Saver, 2006).
Bar graphs typically have horizontal or vertical
bars that represent the size or amount of the group or
variable studied. The bar graph is also a means of
comparing one group with another. Goyal, Gay, and
Lee (2010) conducted a study to describe how socio­
economic status influences the risk of postpartum
depressive symptoms in first-time mothers. As part of
a larger study, the researchers asked women to com­
plete a depression measure, the 20-item Center for
Epidemiologic Studies–Depression [CES-D] scale,
(see Chapter 17) during their third month of pregnancy
and again at 1 month, 2 months, and 3 months post­
partum. The researchers compared groups based on
prenatal depression risk and presented their results in
a vertical bar graph (see Figure 27-2). This figure
indicates that the women with high prenatal risk scores
(CES-D ≥16) were more likely to have high postpar­
tum scores at all three time points. The text indicates
that this was significant (χ2 [1] = 19.9 to 32.5; all
p < 0.001).
A line graph is developed by joining a series of
points with a line and shows how a variable varies
over time. In this type of graph, the horizontal scale
(x-axis) is used to measure time, and the vertical scale
(y-axis) is used to measure number or quantity. A line
graph figure needs at least three data points over time
on the horizontal axis to show a trend or pattern.
However, complexity does not enhance the ability to
convey the data in a meaningful way, so it is recom­
mended that no more than 10 time points should be
included on a single line graph, and there should be
no more than four lines or groups per graph. Figure
27-3 is a line graph developed by Goyal et al. (2010,
p. 100) to depict the pattern of depressive symptoms
over time by contrasting two groups of childbearing
 610 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
women from their third trimester to their 3-month
postpartum assessment. Figure 27-3 is well con­
structed because it includes four data points per line
along the x-axis (mean depressive symptom score at
four visits) and two lines representing the study groups
(low income and high income) being examined for
differences in depressive symptoms after the birth of
their first infant. The third dashed line at the value of
16 on the y-axis is helpful for readers to orient the
scores for the two groups in context to the score that
indicates clinical depression. The discussion of this
graph in the text indicated that the pattern of depres­
sive symptoms over the four time points was different
Low prenatal depression risk
High prenatal depression risk
Depressive Risk (CES-D≥16)
60%
% with High Postpartum
54%
38%
40%
21%
20%
9%
0%
1mo PP 2mo PP
Postpartum depression risk by
prenatal depression risk (n = 198).
Figure 27-2  Postpartum (PP) depression risk by prenatal depression
risk (n = 198). CES-D, Center for Epidemiologic Studies–Depression.
(From Goyal, D., Gay, C., & Lee, K. A. (2010). How much does low
socioeconomic status increase the risk of prenatal and postpartum
depressive symptoms in first-time mothers? Women’s Health Issues,
20(2), 101.)
Depressive Symptoms
Figure 27-3  Depressive symptoms over time by
(CES-D score)
income group (n = 198). CES-D, Center for
Epidemiologic Studies–Depression; PP,
postpartum. (From Goyal, D., Gay, C., & Lee, K. A.
(2010). How much does low socioeconomic status
increase the risk of prenatal and postpartum
depressive symptoms in first-time mothers?
Women’s Health Issues, 20(2), 101.)
for the two groups (F [3192] = 2.76; p = 0.044), with
the low-income group improving at 1 month and 2
months postpartum. The figures could be improved by
having the sample sizes (n) for each group depicted
within the figure or in the legend.
Tables
Tables are used more frequently in research reports
than figures and can be developed to present results
from numerous statistical analyses without additional
software programs. In tables, the results are presented
in columns and rows so that the reader can review
them easily. The sample tables included in this section
present means ( X ), standard deviations (SD), t values,
chi-square (χ2) values, Mann-Whitney U values (Z),
correlations (r), and regression analysis (R2) results.
Means and standard deviations of the study variables
should be included in the published study because of
their importance to future research. A variable’s mean
and standard deviation are essential for (1) providing
42% a basis for comparison across studies, (2) calculating
the effect size to determine sample size for future
studies, and (3) conducting future meta-analyses
(Conn & Rantz, 2003; Sandelowski, 2008). The
7% sample size for each column should be included if the
N varies from the total sample.
Heo et al. (2007) presented the results of their
3mo PP
study about HRQOL for older adults with and without
heart failure in four tables; three are included in this
chapter. Table 27-4 compares individual characteris­
tics (demographic variables) and biological and physi­
ological status of the two groups. This table includes
means and standard deviations (SDs) for variables
measured at the interval or ratio level (age, education
in years, and number of comorbidities) and frequen­
cies and percentages (%) for variables measured at the
Depressive symptoms over time by income group (n = 198).
Cut-off for depression risk
16
Low income group
12
8
High income group
4
0
Pregnancy 1mo PP 2mo PP 3mo PP
Time
 CHAPTER 27  Disseminating Research Findings 611

TABLE 27-4  Comparison between Older Adults with Heart Failure and Healthy Older Adults:
Individual Characteristics and Biological and Physiological Status
Older Adults with Heart Healthy Older Adults
Failure (N = 90), no. (%) (N = 116), no. (%) or Total (N = 206),
Characteristics or M ± SD (Range) M ± SD (Range) no. (%) or M ± SD Statistic p Value
Age 74.8 ± 8.1 (57-93) 74.2 ± 6.3 (58-92) 74.5 ± 7.1 t = 0.56 0.57
Education (yr) 11.7 ± 2.6 (3-20) 13.0 ± 2.7 (4-20) 12.4 ± 2.7 Z = −3.69 <0.001
No. comorbidities 2.3 ± 1.2 (0-5) 1.0 ± 0.9 (0-4) 1.6 ± 1.2 Z = −7.18 <0.001
Sex (male) 48 (53.3) 27 (23.3) 75 (36.4) χ2 = 18.26 <0.001
Marital status
Married 42 (46.7) 46 (39.7) 88 (42.7) χ2 = 1.27 0.74
Divorced/separated 8 (8.9) 13 (11.2) 21 (10.2)
Widowed 33 (36.6) 45 (38.8) 78 (37.9)
Other 7 (7.8) 12 (10.3) 19 (9.2)
Living arrangements 48 (53.3) 61 (52.6) 109 (52.9) χ2 = 0.01 0.92
(live alone)

M, Mean; SD, standard deviation.

From Heo, S., Moser, D. K., Lennie, T. A., Zambroski, C. H., & Chung, M. L. (2007). A comparison of health-related quality of life between older adults
with heart failure and healthy older adults. Heart & Lung, 36(1), 19.

nominal level (sex, marital status, and living arrange­
ments). The two groups were compared for differ­
ences on selected variables using t-test (t) (variables
measured at interval or ratio level), Mann-Whitney
U (Z) (variables measured at ordinal level), and chi-
square (χ2) (variables measured at nominal level).
Nonsignificant results (p > 0.05) between the two
groups in regard to age, marital status, and living
arrangements indicate the groups are comparable in
these areas. It was expected that the older adults with
heart failure would have significantly more comor­
bidities (p < 0.001) than the healthy adults; however,
the significant differences for education (p < 0.001)
and sex (p < 0.001) were unexpected. Researchers
desire their groups to be comparable on demographic
variables, so any differences found between the groups
on study outcome variables of interest are less likely
to be due to basic demographic differences. Any dif­
ferences on baseline characteristics need to be con­
trolled in the remaining statistical analyses.
Heo et al. (2007) conducted a Mann-Whitney U (Z)
analysis to address aim 1 to determine differences in
two groups for HRQOL, health perception, functional
status, physical status, emotional symptoms status
(anxiety and depression), and social support (Table
27-5). Older adults with heart failure reported signifi­
cantly (p < 0.001) poorer HRQOL, health perceptions,
functional status, physical symptoms status, anxiety,
and depression than healthy older adults. Only social
support did not significantly differ between the two
groups (p = 0.66). The results indicate the increased
needs of individuals with heart failure and the need to
develop evidence-based interventions to manage these
needs.
Tables are used to identify correlations among vari­
ables, and often the table presents a correlation matrix
generated from the data analysis. The correlation
matrix indicates the correlation values (coefficients)
obtained when examining the relationships between
variables, two variables at a time (bivariate correla­
tions). The table also includes information about the
significance (p value) of each correlation coefficient,
but the significance will be sample-size dependent.
Borge, Wahl, and Moum (2010) conducted a study
entitled, “Association of Breathlessness with Multiple
Symptoms in Chronic Obstructive Pulmonary
Disease.” One of the questions in their study was
“What are the relationships between … multiple
symptoms (breathlessness, depression, anxiety, sleep­
ing difficulties, fatigue and pain)?” (Borge et al.,
2010, p. 2690). The researchers presented their results
in a correlation matrix, shown in Table 27-6. The six
symptom variables are listed horizontally across the
top of the table and vertically down the left side of the
table. The relationship of Breathlessness with itself
(Breathlessness) is always a perfect positive relation­
ship of 1.00. Breathlessness is correlated with Fatigue
(r = 0.51***). The asterisks (***) indicate that this
correlation is significant at p ≤ 0.001 (some journals
would require the exact p value). This table identifies
the correlation coefficients (Pearson r value) between
the variables, and the significance of each of these
coefficients. Borge et al. (2010) described this table in
the text of their article in the following way.
 612 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research

TABLE 27-5  Comparison between Older Adults with Heart Failure and Healthy Older Adults: Health-
Related Quality of Life, Health Perception, Functional Status, Physical Symptom Status,
Emotional Symptoms Status, and Social Support.
Mean ± SD
Older Adults with Healthy Older Adults
Scale Heart Failure (N = 90) (N = 116) Statistic p Value
LHFQ total* 50.8 ± 22.1 14.4 ± 15.9 Z = −9.9 <0.001
LHFQ physical* 23.1 ± 10.4 7.1 ± 8.1 Z = −9.3 <0.001
LHFQ emotional* 11.4 ± 7.2 2.3 ± 3.7 Z = −9.1 <0.001
Health perception* 2.6 ± 0.8 2.1 ± 0.6 Z = −4.9 <0.001
Functional status* 2.6 ± 0.5 1.7 ± 0.6 Z = −8.6 <0.001
Physical symptom status† 4.9 ± 2.2 11.0 ± 1.9 Z = −11.8 <0.001
Emotional symptom status
Anxiety* 0.9 ± 0.7 0.3 ± 0.5 Z = −6.7 <0.001
Depression* 1.0 ± 0.8 0.4 ± 0.6 Z = −5.7 <0.001
Social support† 2.9 ± 0.8 2.8 ± 1.1 Z = −0.4 0.661

*Higher scores indicate poor status.

†
Higher scores mean better status.
LHFQ, Minnesota Living with Heart Failure Questionnaire; SD, standard deviation.
From Heo, S., Moser, D. K., Lennie, T. A., Zambroski, C. H., & Chung, M. L. (2007). A comparison of health-related quality of life between older adults
with heart failure and healthy older adults. Heart & Lung, 36(1), 21.

TABLE 27-6  Correlation Matrix between Symptoms of Patients with Chronic Obstructive Pulmonary
Disease (N = 154)
Breathlessness Fatigue Anxiety Depression Sleeping difficulties Pain intensity
Breathlessness 1
Fatigue 0.51*** 1
Anxiety 0.34*** 0.45*** 1
Depression 0.38*** 0.53*** 0.65*** 1
Sleeping difficulties 0.42*** 0.58*** 0.56*** 0.50*** 1
Pain intensity 0.26** 0.48*** 0.41*** 0.32*** 0.51*** 1

**Statistically significant at 0.01 level.

***Statistically significant at 0.001 level.
From Borge, C. R., Wahl, A. K., & Moum, T. (2010). Association of breathlessness with multiple symptoms in chronic obstructive pulmonary disease.
Journal of Advanced Nursing, 66(12), 2695.
oasis-ebl|Rsalles|1476726447

higher all other symptom scores will be for that

“[Table 27-6] shows a significant positive (p < 0.001–p
< 0.01) bivariate relationship between all symp-
toms… Pearson’s r ranged from 0.26 (breathlessness-
pain intensity) to 0.65 (depression-anxiety)” (Borge
et al., 2010, p. 2693).
For many reasons, it would be important to exam­
ine this type of correlation matrix before proceeding
with other, more complex statistical analyses. Readers
should first examine the matrix for the direction of a
relationship: is it positive or negative? All of the rela­
tionships in this matrix were positive, indicating that
the higher a patient’s self-reported breathlessness, the
patient. These relationships make sense; a nurse
would not see one of these symptoms improve if
another symptom worsened (a negative or inverse
relationship). Readers should also look for the strength
of the relationship (is it > 0.20 or > −0.20 regardless
of the direction or the p value?). All correlations were
greater than 0.20, and all symptoms would likely be
considered in the next, more complex analysis using
multiple regression models. Readers should also look
for multicolinearity between variables (r > 0.70),
which would indicate that two variables are measur­
ing the same concept. There was no multicolinearity
between symptoms, but anxiety and depression were
highly correlated (r = 0.65) in this sample. If these

TABLE 27-7  Models Best Associated with Health-Related Quality of Life in Older Adults with Heart
Failure (N = 90)
Scale Variables Beta*
LHFQ Physical symptom status −0.501 ‡
Total Age −0.221†
Anxiety 0.218†
*Standardized slope coefficient.
†
p < 0.05.
‡
p < 0.001.
LHFQ, Minnesota Living with Heart Failure Questionnaire.
From Heo, S., Moser, D. K., Lennie, T. A., Zambroski, C. H., & Misook, L. C. (2007). A comparison of health-related quality of life between older adults
with heart failure and healthy older adults. Heart & Lung, 36(1), 21.
two variables were correlated (r > 0.70), researchers
would need to consider their overlapping variance
when testing a multiple regression model. See Chap­
ters 21-25 for additional explanation of specific statis­
tical tests.
Heo et al. (2007, p. 17) conducted a stepwise
regression (R2) to address aim 2 in their study “to
determine the best model predicting HRQOL in each
group from among the physical, psychologic, and
social variables.” The researchers reported, “In older
adults with heart failure, physical symptom status,
age, and one of the emotional symptoms status,
anxiety, were associated with HRQOL (F [3, 86] =
25.05, p < 0.001)” (Heo et al., 2007, pp. 20-21). The
results from the regression analysis are presented in
Table 27-7. The table clearly presents the variables
included in the regression analysis, the beta result for
each variable (physical symptoms status, age, and
anxiety), the unique R2 (percentage of variance in
HRQOL explained by each variable), and the accumu­
lated R2 (0.466 or 46.6% variance in HRQOL explained
by the three variables). An analysis of variance
(ANOVA) was conducted to determine significance,
and the result (F test with df) and significance (p
value) are provided (see Table 27-7). To be useful for
calculating effect sizes and doing meta-analyses, exact
p values are more useful, but some journal formats
require this type of asterisk notation. However, when
the p value in a computer output is listed as “0.0000,”
it stops at this point, and the author should report this
as p < 0.001.
Discussion
The discussion section ties the other sections of your
research report together and gives them meaning. It
includes your major findings, limitations of the study,
conclusions drawn from the findings, implications of
the findings for nursing, and recommendations for
CHAPTER 27  Disseminating Research Findings 613
Unique R2 Accumulated R2 F p Value
0.359 0.359 25.050 <0.001
0.071 0.430
0.037 0.466
further research. Your major findings, which are gen­
erated through an interpretation of the results, should
be discussed in relation to the overriding conceptual
framework as well as the research problem, purpose,
and questions or hypotheses. Researchers should
compare their findings with the findings from previous
research and describe how the new findings extend
existing knowledge. Discussion of the findings also
includes the limitations that were identified while
conducting the study. A study might have limitations
related to sample (e.g., size, response rate, attrition),
design (e.g., convenience sample, only one clinical
site, lack of randomization), or instruments (e.g.,
measure self-report alcohol intake rather than blood
alcohol level). These limitations influence the gener­
alizability of the findings (Pyrczak & Bruce, 2007).
The research report includes the conclusions or the
knowledge generated from the findings. Conclusions
are frequently stated in tentative or speculative terms
because one study by itself does not produce conclu­
sive findings that can be generalized to the larger
population. You might provide a brief rationale for
accepting certain conclusions and rejecting others. The
conclusions need to be discussed in light of their impli­
cations for knowledge, theory, and practice. Describe
how the findings and conclusions might be imple­
mented in specific practice areas. Conclude your
research report with recommendations for further
research. Identify specific problems that require inves­
tigation, and describe procedures for replicating the
study. The discussion section of the report demon­
strates the value of conducting the study and stimu­
lates the reader to conduct additional research that is
needed to provide evidence-based practice (Craig &
Smyth, 2012; Melnyk & Fineout-Overholt, 2010). As
is often the case, by the time the study is published,
the researchers are conducting the next study to address
their own recommendations for future research.
614 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
Heo et al. (2007) linked their study findings with
findings from previous research, identified the limita­
tions of their study, formed conclusions, indicated
areas for future research, and made recommendations
for practice. The researchers concluded:
“Despite marked differences in physical, psychologic,
and HRQOL status between the two groups [healthy
older adults and older adults with heart failure] physi-
cal symptom status was the variable most strongly
related to HRQOL in both groups. Both health care
providers and patients who often dismiss negative
symptomatology as normal signs of aging need to be
aware that they are not so appropriate intervention
can be undertaken” (Heo et al., 2007, p. 23).
The complete research report for this study can be
accessed through CINAHL or PubMed. We recom­
mend you review the study to increase your under­
standing of the content and organization of a
quantitative research report.
Reference Citations
The final section of the research report is the reference
list, which includes all sources that were cited in the
report. Most of the sources in the reference list are
relevant studies that provided a knowledge base for
conducting the study. These sources need to have
complete citations recorded in a consistent manner.
The APA (2010) format is required by the editors of
many nursing and psychology journals. Sources need
to be cited in the text of the report using a consistent
format. It is very important to follow the format guide­
lines for the journal to which you plan to submit your
manuscript for publication. Some journals request that
reference citations include only the past 5 years,
whereas other journals may limit the number of refer­
ences to less than 50 (Nursing Research limits the
number of references to 40 at this time).
Qualitative Research Report
Reports for qualitative research are as diverse as the
different types of qualitative studies. The different
types of qualitative research are presented in Chapter
4, and methods from specific qualitative studies are
presented in Chapter 12. The intent of a qualitative
research report is to describe the dynamic implemen­
tation of the research project and the unique, creative
findings obtained (Marshall & Rossman, 2010). Table
27-8 provides guidelines for developing a qualitative
TABLE 27-8  Outline for a Qualitative Research
Report
Introduction
Background & Significance of the phenomenon to be
studied
Aims or purpose of the study
Brief review of relevant literature (may include theoretical
framework and conceptual definitions)
Gap in knowledge study will address
Study questions and qualitative approach to be used
Methods
Describe qualitative method (phenomenology, historical,
grounded theory, ethnographical or exploratory
descriptive qualitative)
Rationale for and basic steps involved in using this method
Sampling method, consent process
Setting and researcher access to setting
Data collection process
Data analysis plan
Results
Description of sample (may use table or figure)
Results organized by question or themes, depending on the
method
Use narrative and quotations to present findings
Discussion
Major findings compared to previous research
Limitations of the study
Conclusions
Implications
Future studies that are needed
References
Include those cited in paper, using format specified by
journal
research report. Similar to a quantitative report, a
qualitative research report needs a clear, concise title
that identifies the focus of the study. A study by Chun,
Chesla, and Kwan (2011), titled, “So We Adapt ‘Step
by Step’: Acculturation Experiences Affecting Diabe­
tes Management and Perceived Health for Chinese
American Immigrants,” is used as an example to
present aspects of a qualitative research report.
The abstract for a qualitative research report briefly
summarizes the key parts of the study and usually
includes the following: (1) aim of the study; (2) quali­
tative approach (e.g., phenomenology, grounded
theory, ethnography, or historical); (3)methods includ­
ing sample, setting, and methods of data collection;
(4) brief synopsis of findings; and (5) implications of
the findings (Munhall, 2012). Detailed guidelines for
developing an abstract for a qualitative study are pre­
sented later in this chapter. The abstract for the study
by Chun et al. (2011) follows.
 CHAPTER 27  Disseminating Research Findings 615
“This study examines how acculturation affects type 2
diabetes management and perceived health for Chinese
American immigrants in the U.S. Acculturation experi-
ences or cultural adaptation experiences affecting diabe-
tes management and health were solicited from an
informant group of immigrant patients and their spouses
(N = 40) during group, couple and individual interviews
conducted from 2005 to 2008. A separate respondent
group of immigrant patients and their spouses (N = 19)
meeting inclusion criteria reviewed and confirmed
themes generated by the informant group. Using inter-
pretive phenomenology, three key themes in patients’
and spouses’ acculturation experiences were identified:
a) utilizing health care, b) maintaining family relations and
roles, and c) establishing community ties and grounded-
ness in the U.S. Acculturation experiences reflecting
these themes were broad in scope and not fully captured
by current self-report and proxy acculturation mea-
sures. In the current study, shifting family roles and
evaluations of diabetes care and physical environment in
the U.S. significantly affected diabetes management and
health, yet are overlooked in acculturation and health
Chun et al.’s (2011) abstract clearly identifies the
aim of the study, the type of study conducted (inter­
pretive phenomenology), the population (Chinese
immigrants to the United States with type 2 diabetes
and their spouses), and sample size (40). The abstract
met the word limit for this particular journal but
might have been clearer if authors had indicated that
their sample was Cantonese-speaking and sample
size was 20 couples or dyads. It is not until the
methods section that the researchers describe partici­
pants as a convenience sample residing in one area of
California.
Following the abstract, qualitative research reports
usually include four major sections: introduction,
methods, results, and discussion. The content included
in each of these sections is identified in Table 27-8 and
presented next.
Introduction
The introduction section of the report identifies the
research topic or life experience under investigation
and specifies the type of qualitative study that was
conducted. The study aim or purpose and specific
research questions flow from the phenomenon, clarify
the study focus, and identify expected outcomes of the
investigation. The introduction is used to describe
the significance of the study topic for nursing knowl­
edge and practice and to provide documentation of
investigations. Furthermore, the salience and impact of
specific acculturation experiences respective to diabetes
management and perceived health varied across partici-
pants due to individual, family, developmental, and envi-
ronmental factors. In regards to salience, maintaining
filial and interdependent family relations in the U.S. was
of particular concern for older participants and coping
with inadequate health insurance in the U.S. was espe-
cially distressing for self-described lower-middle to
middle-class participants. In terms of impact, family
separation and relocating to ethnically similar neighbor-
hoods in the U.S. differentially affected diabetes manage-
ment and health due to participants’ varied family
relations and pre-migration family support levels and
diverse cultural and linguistic backgrounds, respectively.
Implications for expanding current conceptualizations
and measures of acculturation to better comprehend its
dynamic and multidimensional properties and complex
effects on health are discussed. Additionally, implications
for developing culturally-appropriate diabetes manage-
ment recommendations for Chinese immigrants and
their families are outlined” (Chun et al., 2011, p. 256).
the significance by citing current and relevant litera­
ture (Sandelowski, 2010). The evolution of the
study is described to provide a context for the phe­
nomenon being studied. Provide a rationale for con­
ducting the study, and place the study within both a
historical and experiential context. The historical
context provides a basis for the study and situates
the study in a period of time. The experiential context
presents your involvement in the research experience
and your understanding of the phenomenon under
study.
Some qualitative reports include a brief literature
review in the introduction or in a separate section of
the report after a brief introduction. Other reports
include literature only near the end of the report in the
discussion of the findings. The following text was
abstracted from the introduction of Chun et al.’s
(2011) interpretive phenomenology study of accul­
turation experiences affecting diabetes management
and perceived health for Chinese American immi­
grants. This study is an example of a multidisciplinary
study funded by the National Institute of Nursing
Research. The complete article can be accessed
through PubMed. Because this research was federally
funded by the National Institutes of Health, the article
can be accessed at no cost from PubMed Central by
searching on the author’s name (http://www.ncbi.nlm.
nih.gov/pmc/).
 616 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
Lastly, both proxy and self-report acculturation mea-
sures do not directly evaluate the context in which
acculturation unfolds, thus important family, cultural and
sociopolitical contexts shaping cultural adaptation and
health are overlooked. Consequently, fundamental
questions concerning the nature of acculturation and its
relationship to diabetes management and health—
namely, how do Chinese immigrants actually experience
acculturation, and how do these experiences complicate
or support their diabetes management and health?—
largely remain a matter of conjecture.
A primary goal of this study was to articulate the
complex ways in which acculturation affects diabetes
management and perceived health for first-generation
Methods
The methods section includes three parts: general
method of inquiry, specific methods of inquiry used
in the study, and data analysis plan (Table 27-8). The
general methods section includes the specific qualita­
tive approach (e.g., phenomenology, grounded theory,
or ethnography) used to conduct the study and pro­
vides a brief background or reference to that method.
The philosophical basis for and the assumptions of the
qualitative method are provided and documented from
the current literature. The methodology steps used to
conduct a qualitative study vary based on the philo­
sophical basis of the study, the topic studied, and the
experiences of the researcher conducting the study.
When developing a report of a qualitative study, you
need to discuss the steps of the qualitative method, the
procedures, and the outcomes and provide a rationale
for selecting this qualitative method to guide your
investigation (Denzin & Lincoln 2005; Marshall &
Rossman, 2010). Chun et al. (2011) described the
qualitative method used to conduct their investigation
oasis-ebl|Rsalles|1476726464
in the following way.
“Method
This interpretive comparative interview study was
conducted from 2005 to 2008 with an informant
group of 20 foreign-born Chinese American couples
in which at least one member was diagnosed with
type 2 diabetes.…
Narrative and thematic analyses were conducted
by a multicultural and multidisciplinary team of
Chinese American and European American nurses
and psychologists (Benner, 1994; Cohen, Kahn, &
Steeves, 2000).” (Chun et al., 2011, pp. 257-258)
Chinese immigrants in the U.S. Special attention was
given in identifying key acculturation experiences,
including instances of stressful or successful cultural
adaptation, which respectively complicated or facilitated
their diabetes management. Qualitative research
methods were utilized to comprehend the multidimen-
sional and dynamic properties of acculturation based on
immigrant participants’ extended narratives of their cul-
tural adaptation experiences. Qualitative methods also
allowed for findings and interpretations that were
culturally-anchored in immigrant participants’ daily dia-
betes management and health practices.” (Chun et al.,
2011, p. 257)
In the methods section of a qualitative report,
it is helpful to describe your relevant educational and
clinical background for conducting the study. This
documentation helps evaluate the worth of the
study because the researcher serves as a primary data-
gathering instrument and analyses occur within the
reasoning processes of the researcher (Munhall, 2012).
In the report of your qualitative study, you need to
describe the site and participants selected for the
study. Your unique role as the researcher should also
be detailed, including training of project staff, entry
into the setting, selection of participants, and ethical
considerations extended to the participants during data
collection and analysis (see Table 27-8).
The research report includes a description of your
data collection tools, such as observation guides, open-
ended interviews, direct participation, documents, life
histories, audiovisual media (photographs, DVDs,
videos, or audiotapes), biographies, and diaries. The
flexible, dynamic way in which the researcher collects
data is described, including the time spent collecting
data, how data were recorded, and amount of data col­
lected. For example, if your data collection involved
participant observation, you should describe the
number, length, structure, and focus of the observation
and participation periods. In addition, you should iden­
tify the tools (e.g., audiotapes) for recording the data
from these periods of observation and participation.
Chun et al. (2011) described the methods applied in
their study in the following way.
“A convenience sample was recruited from commu-
nity clinics, community service organizations and via
public notices in the San Francisco Bay Area in
 CHAPTER 27  Disseminating Research Findings 617
California. Six semi-structured interviews with couples
in individual, couple and group contexts focused on
illness understandings, perceptions of diabetes care,
acculturation histories and concrete positive, negative
and memorable narratives of diabetes care. Couples
narrated in each other’s presence (2 couple interviews)
and in group interviews with those who shared their
experience as a patient (2 group interviews) or spouse
(2 group interviews). A subset of these informants (n =
13) was also interviewed individually if extra time was
needed to complete interview questions or if nondisclo-
sure in shared interview settings suggested a private
interview would yield more complete data. Interviews
were conducted in Cantonese and audiotaped text was
simultaneously translated from Cantonese to English
and transcribed verbatim by skilled bilingual bicultural
staff. Each audiofile was then reviewed and checked for
accuracy by a separate bilingual bicultural staff member
who had conducted the interview [procedures and data
collection].
Analysis Plan
The plan for analyzing the data should include who
will be coding the data, how they were trained, any
issues regarding interrater reliability for coding, and
the software product used, if any. Chun et al. (2011)
described their analysis plan as follows.
“After all text was coded for thematic codes in Atlas-ti,
complete text for two codes—‘acculturation positive’
and ‘acculturation negative’—were reviewed. ‘Accul-
turation positive’ included examples of positive cul-
tural adaptation experiences in the U.S. that supported
diabetes management, including acculturation buffers
or protective factors that mitigated acculturation
stress and enhanced positive health perceptions.
‘Acculturation negative’ included examples of negative
cultural adaptation experiences that made diabetes
management difficult, including acculturation stress-
ors and challenges that compromised health percep-
tions. The analyzed text comprised over 465 pages
of extracted text, and represented a broad inclusive
portion of narratives drawn from interviews con-
ducted in all three contexts (couple, group and indi-
vidual interviews). Simultaneous to this thematic
analysis, summaries for each couple were also con-
structed. Thus text for this manuscript was analyzed
Inclusion criteria included a diabetes diagnosis for at
least one year, aged 35 to 75, married for a minimum
of 1 year, self-identified as Chinese American or Chinese,
immigrated to the U.S. from mainland China or Hong
Kong and having a spouse who would agree to partici-
pate. Exclusion criteria for patients included major
diabetes complications (proliferative retinopathy, cere-
brovascular accident or myocardial infarction within the
last 12 months, renal insufficiency, or amputations)
because our intent was to study patients who were
early enough in the disease progression to benefit from
behavioral and family interventions [participant selec-
tion].” (Chun et al., 2011, p. 257)
“… Ethical approval for this study was obtained from
the human subjects review boards at University of Cali-
fornia, San Francisco and the University of San Francisco
[ethical considerations].” (Chun et al., 2011, p. 258)
in the context of the holistic analysis of each couple’s
acculturation experiences.
To address generalizability, findings were presented
to a separate respondent group of patients and
spouses who met the same inclusion criteria as the
informant group, a process known as member check-
ing. Respondents (13 patients and 6 spouses) repre-
sented 16 separate families.… Respondents met in
separate patient groups and spouse groups for two
2-hour interviews to review themes presented in this
manuscript for adequacy, and to add personal varia-
tions to the presented themes.” (Chun et al., 2011,
p. 258)
Results
The results section of the research report includes data
analysis procedures and presentation of the findings.
The data, in the form of notes, tapes, and other materi­
als from observations and interviews, must be synthe­
sized into meaningful categories or organized into
common themes by the researcher. The data analysis
procedures are performed during and after the data
collection process (Marshall & Rossman, 2010;
Munhall, 2012). These analysis procedures and the
process for implementing them are described in Chun
et al.’s (2011) research report.
618 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
“Results
“Acculturation experiences affecting type 2 diabetes
management and perceived health were identified
from the narratives of Chinese immigrant patients (P)
and spouses (S) in the current study. These accultura-
tion experiences centered on three key themes: a)
utilizing health care, b) maintaining family relations
and roles, and c) establishing community ties and
groundedness in the U.S. Participants reported vari-
able acculturative stress levels across these themes
that differentially affected their diabetes management
practices and perceived health.” (Chun et al., 2011,
p. 258)
Present your results in a manner that clarifies for
the reader the phenomenon under investigation.
These results include descriptions, themes, social
processes, and theories that emerged from the study
of life experiences, cultures, or historical events.
Sometimes, these theoretical ideas are organized
into conceptual maps, models, or tables. Researchers
often gather additional data or reexamine existing
data to verify their theoretical conclusions, and this
process is described in the report (Denzin & Lincoln,
2005; Marshall & Rossman, 2010). Some qualitative
study findings lack clarity and quality, which makes
it difficult for practitioners to understand and apply
them. Some of the problems with qualitative study
findings are misuse of quotes and theory, lack of
clarity in identifying patterns and themes in the
data, and misrepresentation of data and data analysis
procedures in the report (Sandelowski, 2010).
Researchers must clearly and accurately develop
their findings and present them in a way that a
diverse audience of practitioners and researchers can
understand.
Chun et al. (2011) presented their results in narra­
tive form. Each of these themes had supporting
themes that were extrapolated from the interviews.
Each theme was presented with supporting data ver­
batim to substantiate the researchers’ process in
identifying themes for the reader to grasp a full
understanding of the participant’s experience. In the
final major theme, “establishing community ties and
sense of groundedness,” Chun et al. (2011) identified
three subthemes substantiated with eight rich quotes
from their participant couples. The following excerpt
from their report is particularly meaningful, as it
served to provide the authors with the title for their
manuscript.
“Relocating to ethnically-matched
neighborhoods
Participants reported strong community ties and
groundedness when relocating to ethnically-matched
neighborhoods like San Francisco Chinatown because
it diffused language stressors, made them feel at ease
and at home, and permitted gradual cultural adjust-
ment to the U.S. It also supported cultural mainte-
nance of valued Chinese dietary practices to manage
their diabetes.
Mr. P22 (70-years old): We don’t have any big prob-
lems … there are a lot of Chinese here.… People can
help and communicate with each other.… So we
adapt step by step.… It’s as if I’m in my home country
and village.… The foods we eat, the living customs
are more or less the same….” (Chun et al., 2011, p.
261)
Discussion
The discussion section includes conclusions, study
limitations, implications for nursing, and recommen­
dations for further research. The conclusions are a
synthesis of the study findings and the relevant theo­
retical and empirical literature. Limitations are identi­
fied and their influence on the formulation of the
conclusions is addressed. Conclusions include the
study aim and the research questions, which were
used to guide the conduct of the study. Implications of
the findings for nursing practice and theory develop­
ment are explored, and suggestions are provided for
further research (Denzin & Lincoln, 2005). Chun
et al. (2011) formed conclusions with support from
the relevant literature, identified study limitations, and
provided implications for practice. They also pro­
vided direction for further research in the following
excerpt.
“The current study’s findings reveal a complex and
multifaceted relationship between acculturation and
diabetes management and perceived health. This was
illustrated by the broad scope, varying salience, and
differential impact of participants’ acculturation expe-
riences. In regard to scope, participants reported a
wide range of acculturation experiences covering
three key themes.… Such a broad scope of accultura-
tion experiences highlights the need to expand the
parameters in which acculturation is conceptualized
and measured in health research.
 CHAPTER 27  Disseminating Research Findings 619
Most studies employ proxy or self-report accul-
turation measures that do not directly or comprehen-
sively assess such broad acculturation influences on
diabetes and health (Chun, 2006; Perez-Escamilla &
Putnik, 2007; Salant & Lauderdale, 2003).
Limitations to the current study include the focus
on Cantonese-speaking Chinese immigrants. This lin-
guistic dialect is predominant among Chinese immi-
grants in the San Francisco region where this study
was conducted. Future studies should include the
growing numbers of Mandarin-speaking Chinese
immigrants in the U.S. with distinct sociocultural
backgrounds and life experiences. Additionally, future
investigations should include the perspectives of
other family members such as children, siblings, and
extended family in multigenerational immigrant
households and the perspectives of widowed or
unmarried persons with diabetes. Expanded analyses
of different family relationships and structures can
deepen our understanding of Chinese immigrant
family acculturation processes and their significance
to diabetes management and health. Also, potential
age-related differences in acculturation experiences
were not fully examined in this study and should thus
be included in future research. Lastly, future investiga-
tions should identify culturally-appropriate diabetes
interventions, including bicultural competencies and
skills development, that facilitate diabetes manage-
ment in different cultural settings” (Chun et al., 2011,
pp. 262-263).
Theses and Dissertations
Theses and dissertations are research reports that stu­
dents develop in depth as part of the requirements for a
degree. The content included in a thesis or dissertation
depends on the university requirements, the research
guidelines of the nursing college or school, and the
members of the student’s research committee. Most
theses and dissertations are organized by chapters, with
the college or university specifying the content of each
chapter. The content included in a thesis follows the
general outline of quantitative and qualitative research
reports (Tables 27-7 and 27-8). Chapter 28 also pro­
vides guidelines for the content of theses and disserta­
tion proposals. Baggs (2011) discussed the option of
publishable papers as chapters for a dissertation and
issues to consider regarding copyright and intellectual
property. The APA (2010) website has a chapter on
converting a thesis or dissertation to a journal article
(http://supp.apa.org/style/pubman-ch08.pdf ).
Audiences for Communication
of Research Findings
Before developing a research report, you need to
determine who will benefit from knowing the findings.
The greatest impact on nursing practice can be
achieved by communicating nursing research findings
to a variety of audiences, including nurses, other
health professionals, healthcare consumers, and policy
makers. Nurses, including administrators, educators,
practitioners, and researchers, must be aware of
research findings for use in practice and as a basis for
conducting additional studies. Other health profes­
sionals need to be aware of the knowledge generated
by nurse researchers and facilitate the use of that
knowledge in the healthcare system as part of the
delivery of evidence-based practice (Craig & Smyth,
2012). Consumers are interested in research findings
about illnesses that they or family members have.
Policy makers at the local, state, and federal levels use
research findings to generate health policy that has an
impact on consumers, individual practitioners, and the
healthcare system. Rather than the more common
research with individuals as the source of data,
Chapman, Wides, and Spetz (2010) provided an
excellent example of communicating policy-related
research findings using the Medicare Claims Process-
ing Manual, reports from the National Council of
State Boards of Nursing, and congressional reports as
their source of data. They tabulated their data and
concluded that more data are needed in these docu­
ments about the type of care provided. They also con­
cluded from their analysis that the payment system for
advanced practice nurses needs to be remodeled
(Chapman et al., 2010).
Strategies for Communicating Research
to Different Audiences
Research findings can be communicated or dissemi­
nated as oral podium presentations, visual posters, and
written reports. Each type of report can reach addi­
tional audiences in electronic format. Table 27-9 out­
lines various strategies for disseminating findings to
nurses, healthcare professionals, policy makers, and
consumers.
Audience of Nurses
The most common mechanisms nurses use to com­
municate research findings to their peers are presen­
tations at conferences and meetings. Sigma Theta
Tau, the international honor society for nursing,
sponsors international, national, regional, and local
research conferences. Specialty organizations, such
620 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research

TABLE 27-9  Audience and Strategies for Communicating Research

Audience
Nurses—administrators,
educators, clinicians,
practitioners, researchers
Other healthcare
professionals
Policy makers
Healthcare consumers
Strategies for Communicating Research
Oral and Visual Presentation Written Report Electronic Media
Nursing research conferences Nursing-refereed journals DVD
Professional nursing meetings Nursing books Podcast
and conferences Monographs Video and audio recordings
Collaborative nursing groups Research newsletters Websites
Thesis and dissertation defenses Theses and dissertations
Foundation reports
Electronic databases
Professional conferences and Professional journals and books DVD
meetings Newsletters Podcast
Interdisciplinary collaboration Foundation reports Video and audio recordings
Electronic databases Websites
Testimony on health problems to Written testimony Video recorded testimony
state and federal legislators Reports to legislators
Reports to funding agencies
AHRQ and NINR reports and
presentations to policy makers
Television and radio interviews Newspaper Websites
Community meetings News and popular magazines Podcast
Patient and family teaching Electronic databases

AHRQ, Agency for Healthcare Research and Quality; NINR, National Institute for Nursing Research.

as the American Association of Critical Care Nurses,
Oncology Nurses’ Society, and Association of
Women’s Health, Obstetrics, and Neonatal Nursing,
sponsor research conferences. Many universities and
some healthcare agencies sponsor or cosponsor
research conferences. For various reasons, nurses are
not always able to attend these research conferences.
To increase the communication of research findings
and disseminate the new knowledge more diffusely,
conference sponsors often provide podcasts or DVDs
of the research presentations. Some sponsors publish
abstracts of studies with the conference proceedings,
publish the abstracts in a research journal supple­
ment, or provide materials electronically on their
websites.
The publishing opportunities in nursing continue
to increase, with 74 journals currently listed and eval­
uated by Journal Citation Reports (Polit & Northam,
2011). Opportunities to publish research have grown
with the expansion of research journals. When decid­
ing on a potential journal for your work, four criteria
should be considered: (1) the intended readers you
want to reach with your research findings, (2) the fit
of your topic to the journal’s focus, (3) the journal’s
track record for publishing in a timely manner, and
(4) the impact factor for the journal. The content for
your work may be most applicable to a small spe­
cialty group, or perhaps a broader audience would
find it interesting and useful to their practice. You
should not limit your options to a nursing journal if a
wider audience of health professionals needs to know
the findings of your research. You can look at your
list of reference citations to get an idea of journals
that have an interest in publishing on your topic. If
it is important for your findings to be reported as
soon as possible, consider an online journal or a
journal that has monthly issues rather than quarterly
issues. Finally, the impact factor for a journal is
important to consider (see Publishing Research Find­
ings section of this chapter). The 74 nursing journals
listed by Polit and Northam (2011) give you a place
to start.
Many researchers disseminate their findings by
publishing books or chapters in books. Foundations
and federal agencies that sponsor a research project
often provide reports of studies that have been con­
ducted or are in progress. The American Nurses’
Foundation publishes a newsletter, Nursing Research
 CHAPTER 27  Disseminating Research Findings 621
Report, which identifies studies they funded and
includes abstracts of these studies. The National Insti­
tute for Nursing Research provides reports on its
funded grants, including research project titles, names
and addresses of researchers, period of support, a
brief description of each project, and any publication
citations that result from the research (www.nih.gov/
ninr).
Audience of Healthcare Professionals
and Policy Makers
Nurse researchers communicate their research to
other health professionals at meetings and conferences
sponsored by organizations such as the American
Heart Association, American Public Health Associa­
tion, American Cancer Society, American Lung Asso­
ciation, National Hospice Organization, and National
Rural Health Association. Nurses must believe in the
value of their research and present their findings at
conferences that attract a variety of healthcare profes­
sionals. Nurse researchers and other health profession­
als conducting research on the same problem might
collaborate to publish an article, a series of articles, a
book chapter, or a book. This type of interdisciplinary
collaboration increases communication of research
findings and facilitates synthesis of research knowl­
edge to promote evidence-based practice.
The Agency for Healthcare Research and Quality
(2011) was one of the first national organizations to
develop and distribute evidence-based guidelines for
practice. The Agency for Healthcare Research and
Quality implemented the “Evidence-based Practice
Centers (EPC) Program,” which includes the award­
ing of contracts to institutions in the United States
and Canada to serve as an EPC. EPCs review all
relevant scientific literature on clinical, behavioral,
and organization and financing topics to produce
evidence reports and technology assessments. These
reports are used for informing and developing cover­
age decisions, quality measures, educational materials
and tools, guidelines, and research agendas (Donald­
son, Rutledge, & Geiser, 2008). The institutions
with EPCs and the activities of these centers can
be viewed online at www.ahrq.gov/clinic/epc (see
Chapter 19).
Audience of Healthcare Consumers
Healthcare consumers as an audience are frequently
neglected by nurse researchers. The findings from
nursing studies can be communicated rapidly to the
public through news releases. A nursing research
article published in a local paper has the potential of
being picked up by a national wire service and
published in other papers across the United States.
Nurse researchers also need to make their findings
available through electronic means. An increasing
amount of healthcare information is available elec­
tronically on various websites, but often this informa­
tion is not based on research. There is a need to provide
consumers with evidence-based guidelines and educa­
tional materials to assist them in making quality
healthcare decisions.
Nursing research findings should be communicated
to consumers by being published in news magazines,
such as Time and Newsweek, or popular health maga­
zines, such as American Baby and Health. Health
articles published for consumer magazines and online
distribution reach millions of readers at a time (e.g.,
webmd.com or WebMD the Magazine). Television and
radio are other valuable media for communicating
research findings to consumers and other healthcare
providers. Another important method of communicat­
ing research findings to consumers is through patient
and family teaching. Nursing interventions and prac­
tice protocols based on research are more credible to
consumers than unresearched actions, and the ultimate
goal is evidence-based practice (Craig & Smyth,
2012). Freelance journalists often contact authors of
scientific articles, and these writers have the skills to
translate research findings into language for consum­
ers. Lee and Gay (2011) published a study entitled,
“Can Modifications to the Bedroom Environment
Improve the Sleep of New Parents? Two Randomized
Controlled Trials.” A skilled journalist writing for Par-
enting Magazine subsequently was able to catch con­
sumers’ attention with the title, “Desperately Seeking
Sleep,” to disseminate the same data contained in the
research publication but for a targeted public audience
(Bernstein, 2011).
Presenting Research Findings
Research findings are communicated at conferences
and meetings through verbal and poster presentations.
With presentations, researchers have an opportunity
to share their preliminary findings, answer questions
about their studies, interact with other interested
researchers, and receive immediate feedback on their
study. Research project findings are frequently pre­
sented at conferences as preliminary findings based
on the quality of the abstract submitted by the
researchers. When research findings are published,
the data must not be published elsewhere, and any
presentation of these data at a conference should be
acknowledged.
622 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
Verbal Presentations
Researchers communicate their findings through
verbal presentations at local, national, and interna­
tional conferences. Presenting findings at a conference
requires receiving acceptance of your abstract as a
presenter, developing a research report, delivering the
report, and responding to questions.
Receiving Acceptance as a Presenter
Most research conferences require researchers to
submit an abstract, and acceptance as a presenter is
based on the quality of the abstract. The abstract
should be based on the theme of the conference and
the organizers’ criteria for reviewing the abstract.
As noted earlier, an abstract is a clear, concise
summary of a study that is usually limited to 100
to 250 words. The abstract submitted for a verbal
presentation is usually based on results from a
completed study that is not yet published. Findings
from a pilot study or preliminary findings would
be acceptable for a poster presentation (Pyrczak
& Bruce, 2007). The sponsors of a research confer­
ence circulate a call for abstracts 9 months to 1 year
before the conference. Many research journals and
newsletters publish these requests for abstracts, and
they are available electronically. In addition, confer­
ence sponsors email requests for abstracts to univer­
sities, major healthcare agencies, and known nurse
researchers.
The call for abstracts stipulates the format for the
abstract. Frequently, abstracts are limited to 1 page,
single-spaced, and include the content outlined in Box
27-1 (for an abstract of a quantitative study) and Box
27-2 (for an abstract of a qualitative study). When
abstracts are submitted online, you may be limited to
a specific number of words or characters. For elec­
tronic submissions, write and revise the abstract in a
separate document. Depending on the instructions,
you may copy and paste the text in a box on the
webpage or attach the file.
The title of your abstract must create interest, and
the body of your abstract “sells” the study to the
reviewers. An example of an abstract is presented in
Box 27-3. Names and affiliations would be removed
for review. Writing an abstract requires practice; fre­
quently, a researcher rewrites an abstract many times
until it meets all the criteria, including the word limit,
outlined by the conference sponsors. Careful atten­
tion to the criteria of the sponsoring agency should
assist you in developing and refining your abstract
and increase your chances of having the abstract
accepted for either a podium or poster presentation.
The Western Institute of Nursing (WIN) has an
Box 27-1 Outline for Quantitative
Study Abstract
I. Title of the Study
II. Introduction
Statement of the problem and purpose
Identification of the framework
III. Methodology
Design
Sample size
Identification of data analysis methods
IV. Results
Major findings
Conclusions
Implications for nursing
Recommendations for further research
Note: The title and authors with affiliations, a
conflict of interest statement, a brief reference list
of one or two key citations, and the
acknowledgment of funding source are not usually
considered in the word limitations for the abstract.
Box 27-2 Outline for a Qualitative
Study Abstract
I. Title of the Study
II. Introduction
Identification of phenomenon to be
studied
Statement of the aim of the study
Identification of the qualitative method
used to conduct the study
Evolution of the study (rationale for
conducting the study, historical context,
and experiential context)
III. Methodology
Discussion of the background, philosophy,
and assumptions for the method used to
conduct the study
Brief description of the sample, setting,
and data collection process
IV. Results
Brief description of data analysis
procedures
Major findings
Conclusions
Implications for nursing
Recommendations for further research
Note: The title and authors with affiliations, a
conflict of interest statement, a brief reference list
of one or two key citations, and the
acknowledgment of funding source are not usually
considered in the word limitations for the abstract.
 CHAPTER 27  Disseminating Research Findings 623

Box 27-3 Example of an Abstract: “Sleep Adequacy and Environmental Demands

among Racially Diverse Women”
Dawn E. Dailey, RN, MSN, Harvey C. Davis, RN, PhD Prevention’s (CDC) 2002 Behavioral Risk Factor
APRN, BC Assistant Professor Surveillance System (BRFSS) database. BRFSS is
Doctoral Candidate Department of Health and an ongoing data collection program designed to
Department of Family Human Services measure behavioral risk factors in random
Health Care Nursing San Francisco State University samples of adults 18 years and older. Of the
University of California San Francisco, CA 32,503 adults surveyed, 29,465 were Caucasian
San Francisco and 3,040 were African American women
San Francisco, CA ranging in age from 18 to 65 years (mean = 42
years). Participants were asked to report the
Bruce A. Cooper, PhD Kathryn A. Lee, RN, PhD, FAAN number of days they did not get enough sleep
Senior Statistician Professor within the last month. Days of inadequate sleep
Office of Research Department of Family Health in the past month were re-coded as adequate
University of California Care Nursing (0-11 days), moderate (12-23 days), or severe
San Francisco University of California San (24-30 days). Comparative analyses (t-test,
San Francisco, CA Francisco chi-square) were used to describe differences in
San Francisco, CA health-related characteristics and sleep
Aim: To describe environmental demands adequacy between the two groups of women.
associated with sleep inadequacy among Multiple regression analysis was used to identify
Caucasian and African American women. key factors related to insufficient sleep.
Background: Sleep deprivation is purported Results: While more Caucasian women (32%)
to affect an individual’s health through reported overall sleep inadequacy (moderate
physiological, cognitive, emotional, and social and severe) than African American women
mechanisms. Regardless of age, women are (30%), significantly more African American
more likely to suffer from poor sleep. However, women (22%) compared to Caucasians (16%)
little is known about sleep adequacy in reported severe problems with sleep (X2 = 61.02;
subgroups of American women. p = .0001). Preliminary findings indicate that
Conceptual basis: This study is based on an sleep inadequacy is associated with specific
adaptation of a framework developed by Lee health behaviors, chronic medical conditions,
and colleagues (1994) that describes internal and psychological symptoms.
and external environmental demands that Implications: Preliminary results indicate that
result in sleep disturbance and daytime perception of inadequate sleep permeates the
fatigue. Internal environmental demands are lives of women. Findings suggest that African
characterized as physical illness, depression, age, American women suffer from more severe sleep
smoking patterns, alcohol use, physical activity, problems than their Caucasian counterparts.
mental health, stress, and energy levels. External Inadequate sleep has a profound impact on
environmental demands are characterized as health, functioning, and well-being. These
marital status, income, employment status, findings build upon existing knowledge of sleep
educational level, number of children, role disturbances and can lead to the development
demands, and life events. of effective interventions to enhance sleep
Methods: A secondary analysis was conducted quality in diverse groups of women.
using the Centers for Disease Control and
Source: Dailey, D. E., Davis, H. C., Cooper, B.A., & Lee, K. A. (2005). Sleep inadequacy and environmental demands
among racially diverse women. Abstract for presentation at Western Institute of Nursing Conference, San
Francisco, CA.

excellent tutorial called, “Writing a WINning abstract”
by Lentz (2011b) from University of Washington
School of Nursing. This tutorial can be accessed
at the WIN website (http://www.ohsu.edu/son/win/
pcorner.shtml).
Some sponsors ask that you specify whether you
want to be considered for an oral podium presentation
or a poster presentation, whereas other sponsors decide
on a poster versus oral podium presentation based
on their own criteria or scoring system. Generally,
624 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
abstracts that describe smaller sample sizes and read
more like a small unfinished pilot work are less likely
to be accepted for an oral podium presentation. Some
sponsoring agencies require that you submit two
versions of your abstract: one with names and affilia­
tions that would be in their program or published
abstracts and another that removes all names and
affiliations so that the abstract is anonymous and
reviewers are blinded. Read the sponsor’s instructions
carefully because they usually require that content
of the abstract has not been published or presented
elsewhere. Instructions also indicate whether or not
accepted abstracts are published, usually as a supple­
mental issue of the sponsor’s affiliated professional
journal.
Developing an Oral Research Presentation
The presentation developed depends on the focus of
the conference, the audience, and the time designated
for each presentation. Some conferences focus on
certain sections of the research report, such as tool
development, data collection and analysis, findings, or
implications of the findings for nursing practice. It is
important to address the major sections of a research
report (introduction, methods, results, and discussion)
in your presentation. The content of a presentation
varies depending on whether the audience consists of
mainly researchers or clinical nurses. You would be
best prepared for this type of presentation if you
attended a previous meeting and if you practice your
presentation on a local audience of peers. If you are
unsure who your conference audience represents (cli­
nicians or researchers, students or experienced
researchers) or how large the audience will be (a con­
current session with an audience that specifically
selects your presentation based on the title and their
interest or a general session with the entire audience
of conference participants), it is important to ask the
contact person for the conference. In planning your
allotted time for the presentation, it also is helpful to
know if questions from the audience will be allowed
during your presentation, allowed at the end of your
presentation, or held until the end of your entire
session when participants direct their questions to any
one of the presenters in your session.
Time is probably the most important factor in
developing a presentation because many presenters
are limited to 10 or 15 minutes, with an additional 5
minutes for questions. As a guideline, you want to aim
for one slide per minute. Your title slide, acknowledg­
ment slide, and final slide of references or slide calling
for questions from the audience should be included in
the formula. These slides are included in the timing
because someone will be introducing you after settling
the audience, the previous speaker may be running
late, and you do not want to be encroaching on the
next speaker’s time. The WIN website also has a tuto­
rial on how to do a 10-minute oral research presenta­
tion developed by Lentz (2011b) (http://www.ohsu.edu/
son/win/pcorner.shtml). Your audience is there to hear
what is new in your area of research, not to hear all
the background review of literature that brought you
to this current research. For guidance, you should
spend 20% (2 minutes or 2 slides) of your time on the
title and introduction, 20% on the methodology, 40%
on the results, and 20% on the discussion and implica­
tions for practice and research.
Your title slide should provide the audience with
the gap in knowledge that you are addressing in your
presentation. Your introduction should acknowledge
funding sources and collaborators if applicable as well
as any conflict of interest. A very brief review of key
background literature and a simple diagram of the
conceptual framework should lead directly into the
research questions or hypotheses that address the gap
in knowledge. The methodology content includes a
brief identification of the design, sampling method,
measurement techniques, and analysis plan. The
content covered in the results section should start with
a table of your sample characteristics followed by a
slide of results for each question or hypothesis. The
presentation should conclude with a brief discussion
of findings, implications of your findings for clinical
practice, and recommendations for future research.
Most presenters find that the shorter the presentation
time, the greater the preparation time needed. If you
are limited to 10 minutes, you must be very selective
about which one or two research questions or hypoth­
eses you will focus on. If you have 15 or 20 minutes,
you may still limit your presentation to three research
questions or hypothesis but allow more time to discuss
the details regarding the strengths and limitations of
your research.
The longer the presentation, the more important
figures and pictures, or some animation, become. The
information presented on each slide should be limited
to eight lines or fewer, with six or fewer words per
line. A single slide should contain information that can
be easily read and examined in 30 seconds to 1 minute.
Only major points are presented on visuals, so use
single words, short phrases, or bulleted points to
convey ideas, not complete sentences. Figures such as
bar graphs and line graphs usually convey ideas more
clearly than tables. Pictures of the research setting,
equipment, and photographs of the research team help
the audience to visualize the research project better. A
 CHAPTER 27  Disseminating Research Findings 625
laser pointer may be useful to guide the audience to
your key point on the slide, but the deliberate and
careful use of color is more appealing to the audience,
can increase the clarity of the information presented,
and can call attention to a particular important statisti­
cal test and p value without need for a laser pointer.
However, avoid using particular shades of red color
for bulleted points or highlighted wording, particularly
if you have a dark background; red may appear fine
on your computer as you develop the slide, but becomes
difficult to see when projected to a large audience.
Preparing the script and visuals for a presentation
is difficult, so enlist the assistance of an experienced
researcher and audiovisual expert. Rehearse your pre­
sentation with experienced researchers, and use their
comments to refine your script, visuals, and presenta­
tion style. If your presentation is too long, synthesize
your script and provide handouts for important content.
PowerPoint slides provide an excellent format for
presenting an oral research report; they include
easy-to-read fonts, color, creative backgrounds, visuals
or pictures to clarify points, and animation options.
However, consulting an audiovisual expert will ensure
that your materials are clear and properly constructed,
with the print large enough and dark enough for the
audience to read. When the PowerPoint slides have
been developed, view them from the same vantage
point as the audience to ensure that each slide is clear
and can be visualized without totally darkening the
room.
Delivering a Research Report
and Responding to Questions
A novice researcher might attend conferences and
examine the presentation style of other researchers
before presenting an oral report. Even though each
researcher needs to develop his or her own presenta­
tion style, observing others can promote an effective
style. An effective presentation requires practice. You
need to rehearse your presentation several times, with
the script, until you are comfortable with the timing,
the content, and your presentation style. When practic­
ing, use the visuals so that you are comfortable with
the equipment. It is always advantageous to check out
the room where you will be presenting to see how
chairs are arranged and how the podium and screen
are situated.
Most conferences organize their oral presentations
by topic into a session moderated by an expert in the
field. The session usually includes a presentation by
the researcher, a comment by the session’s moderator,
and a question period before moving to the next
speaker. If your presentation is too long for the allotted
time, there will be no opportunity for questions from
the audience. When preparing for a presentation, try
to anticipate the questions that members of the audi­
ence might ask and rehearse your answers. Practice
your presentation with colleagues and ask them to
raise questions. If you practice making clear, concise
responses to specific questions, you will be less
anxious during your presentation. When giving a pre­
sentation, have someone make notes of the audience’s
questions, suggestions, or comments because they are
often useful when preparing a manuscript for publica­
tion or developing the next study.
Poster Sessions
Your research abstract may be accepted at a confer­
ence as a poster session rather than podium presenta­
tion. A poster session is a visual presentation of your
study, all on one surface, with either three or four
columns. Before developing a poster, follow the con­
ference sponsor’s directions for (1) the size limitations
or format restrictions for the poster, (2) the size of the
poster display area, and (3) the background and poten­
tial number of conference participants. Your institu­
tion may have a template with the logo that you are
required to use for the audience to identify your affili­
ation more easily. A poster usually includes the fol­
lowing content: the title of the study; investigator
and institution names; purpose; research objectives,
questions, or hypotheses (if applicable); framework;
design; sample; instruments; essential data collection
procedures; results; conclusions; implications for
nursing; recommendations for further research; a few
key references; and acknowledgments.
For clarity and conciseness, a poster often includes
pictures, tables, or figures to communicate the study.
Figure 27-4 presents the award-winning WIN poster
developed by Dr. Dawn Dailey when she was a doc­
toral student at University of California, San Fran­
cisco. This poster was presented at the WIN meeting
in 2005, based on acceptance of the abstract pre­
sented in Box 27-3. It has a polished, professional
look and presents the key aspects of the study using
a balance of text, figures, and color in a 3-column
format. Conference sponsors often provide boards
for displaying posters, so the poster can be rolled
to prevent creases and easily transported to the con­
ference in a protective tube. Online services (e.g.,
Makesigns.com) will ship the poster to your hotel
destination for an extra fee; this avoids any problem
with airport security or risk of misplacing your poster
while traveling.
A quality poster completely presents a study, yet
can be comprehended in 5 minutes or less. Bold
 626

UNIVERSITY OF CALIFORNIA SAN FRANCISCO

UCSF Sleep Adequacy and Environmental Demands among
Racially Diverse Women
University of California
DE Dailey, HC Davis, BA Cooper, & KA Lee UCSF School of Nursing
San Francisco

Introduction Results Summary

Sleep disturbance is purported to affect an individual’s health through 1. In this sample of women, African Americans were younger and fewer
Table 1: Selected Sample Characteristics Figure 2: Differences in Sleep Severity between
physiological, cognitive, emotional, and social mechanisms. were married. African American women also had lower incomes,
Caucasian and African American women
Regardless of age, women are more likely to suffer from poor sleep. more children in the home, and less education.
White Black
However, little is known about sleep inadequacy in subgroups of
n=29465 n=3040 2. Whites and Blacks experienced the same percentage of sleep
American women. 25
inadequacy. However, a higher percentage of Blacks reported severe
Internal sleep inadequacy.
Purpose
• Describe environmental demands associated with sleep inadequacy Age (years) 43 40
among Caucasian and African American women. 20 3. Internal variables accounted for 15% of the variance in sleep
Anxiety (days/month) 6.6 6.1
inadequacy. Very little (<4%) was due to external role demands.
Depression (days/month) 3.8 4.4
This study is based on an adaptation of a framework developed by Lee
and colleagues (1994) that describes internal and external Perceived health (days/month) 2.3 2.6 4. In this sample, anxiety, depression, age and perceived health were
15
environmental demands resulting in sleep disturbances and fatigue White the strongest predictors of sleep inadequacy. Ethnicity was not a
Weight (pounds) 156 172
(Figure 1 below). significant predictor of sleep inadequacy. These 3 variable accounted
Black for 15% of the variance in sleep inadequacy.
Smoker 26% 23%
10

Percentage
Internal External BMI (normal limits) 18% 29% 5. The strongest predictors of sleep inadequacy in the Caucasian
Environmental Environmental
Physical illness 49% 57% sample were anxiety, perceived health, age, depression,
Demands Demands
5 and number of children in the home.
Alcohol consumption (times/month) 2.6 1.5
Physical illness* External 6. The strongest predictors of sleep inadequacy in the African American
Material status*
Menstrual cycle <High school 7% 11% sample were anxiety, depression, age, physical illness and
Income* 0
phase employment.
Occupation <25,000 per year 24% 45%
Depression* Moderately Severely
Employment
Anxiety* Children in home 41% 56%
status*
Self-esteem Sleep Sleep Inadequacy
Level of education* Married 59% 27% Conclusions/Further Study
Age* Adequacy Number of Employed 68% 67%
Smoking*
children*
Sleeping patterns
Age of children • Sleep inadequacy permeates the lives of women.
Diet patterns*
Role demands
Weight*
Social support • Findings suggest that African American women suffer from more severe
Caffeine/alcohol Table 2: Multiple Regression Procedure, with Sleep Inadequacy Score as the Dependent Variable and 14
Positive life events sleep inadequacy than Caucasian women.
intake* Internal and External Demands as Independent Variables
Negative life events
Exercise patterns*
• Inadequate sleep is associated with health and functioning,
Explanatory Variables Cummulative R2 beta p
* Variables in CDC BRFSS 2002 database R2 Change • These finding support existing knowledge of sleep disturbances and its
Race/ethnicity .001 association with anxiety and depression.
Internal Variables .147 .146
Methods Physical illness .080 .0001 • Further research is needed to understand variables that affect sleep in
subgroups of women to promote the development of
Depression .013 .0001 effective interventions to enhance sleep quality.
• Secondary analysis using the Center for Disease Control and Anxiety .019 .0001
Prevention’s (CDC) 2002 Behavioral Risk Factor Surveillance System
Age –.007 .0001
(BRFSS) database. Reference:
• BRFSS is an ongoing data collection program designed to measure Smoker –.083 .0001
Lee, KA et al (1994). Fatigue as a response to environmental demands in
behavioral risk factors in random samples of adults 18 years and older Diet patterns –.001 .840
women’s lives. Image:Journal of Nursing Scholarship, 26(2),149-154
• Of the 62,341 individuals surveyed, 53% were women with ages Weight .001 .011
ranging from 18 to 65 years. Alcohol consumption –.054 .005 For additional information contact:
Exercise –.062 .0001 Dawn E. Dailey, RN, MSN, PhD(c)
Department of Family Health Care Nursing
External Variables .184 .038
Sleep Assessment University of California San Francisco
Marital status .013 .189
Sleep adequacy was assessed by asking: Income .001 .686
During the past 30 days, for about how many days have your felt Employment status .033 .001
UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research

you did not get enough rest or sleep?

Educational level .008 .118
Recoded categories: Children in home .053 .0001
• Adequate= 0-11 days
• Moderately inadequate = 12-23 days
• Severely inadequate = 24-30 days

Figure 27-4  Poster of “Sleep Adequacy and Environmental Demands Among Racial Diverse Women.” (From University of California San Francisco.)
 CHAPTER 27  Disseminating Research Findings 627
headings are used for the different parts of the research
report, followed by concise narratives or bulleted
phrases. Print size on a poster needs to be large enough
to be read at 3 feet (approximately 20 font size), but
the title or banner should be readable at 20 feet
(Shelledy, 2004). There may be an extra cost for
glossy finish, but the crisp appearance and readability
may be worth the price to you. Poster sessions usually
last 1 to 2 hours; you should remain by your poster
during this time and offer to answer any questions
when a viewer is present. Most researchers provide
conference participants with a copy of the 1-page
accepted abstract or a 1-page color handout of the
poster tacked to the poster board with contact informa­
tion, particularly if you cannot stand by your poster
for the entire allotted time. Websites are available to
assist you with research posters. You can also search
the Internet for “research poster” to view current web­
sites or a commercially available product (http://
www.makesigns.com), or visit the University of
Buffalo website at http://ublib.buffalo.edu/libraries/
asl/guides/bio/posters.html.
Conferences include both quantitative and qualita­
tive research posters. Narrative content does not lend
itself to the crisp presentation required on a poster. The
number of words on a poster, even for a qualitative
study, should be kept to a minimum, so you must
select the most critical information to include. It is
recommended that sections of the poster be color-
coordinated; research has been conducted on use
of color to attract people to a poster (Keegan &
Bannister, 2003).
Summary and implications sections are frequently
where a viewer looks first, given the limited time for
viewing many posters during a session. Because rich
narrative text is so meaningful in qualitative studies,
authors are advised to bold and enlarge the font for a
few particularly meaningful quotes, and use artwork
or photos that conceptualize the quote in a visual way.
Posters usually take 10 to 20 hours to develop based
on the complexity of the study and the experience of
the researcher. Novice researchers usually need more
than 20 hours to develop a poster. Important points in
poster development include planning ahead, seeking
the assistance of others, and limiting the information
on the poster (Shelledy, 2004). Your school or work­
place should have a template and color scheme for you
to use with the appropriate logos and colors.
One major advantage of a poster session is the
opportunity for one-to-one interaction between the
researcher and the viewer. At the end of the poster
session, individuals interested in a study frequently
stay to speak to the researcher. Have a notepad on
hand to record comments and contact information for
individuals conducting similar research. Exchanging
business cards and writing key information on the
back of the card is a useful practice. Poster sessions
provide an excellent opportunity to begin networking
with other researchers involved in the same area of
research. Conference participants occasionally request
your study instruments or other items, so it is essential
that you keep a record of their contact information and
specific requests.
Publishing Research Findings
Oral and poster presentations are a valuable means of
rapidly communicating findings, but their impact is
limited and the contents should not be previously pub­
lished. Even if the accepted abstract is published in a
supplemental volume of a journal associated with the
conference sponsors, you should be planning publica­
tion of the full findings for a research journal as you
prepare for the oral podium or poster presentation.
Published research findings are permanently recorded
in a journal or book and usually reach a larger audi­
ence than presentations do. There should be an
acknowledgment in the published report that the con­
tents of your paper were presented at a particular
research conference. The presentation and comments
from the audience can provide a basis for finalizing
your article for publication. Many journal editors are
conference attendees and may request your paper for
an article when they hear your oral presentation or see
your poster. Many researchers present their findings at
a conference and never submit the paper for publica­
tion; this could create the impression on a vita that a
researcher’s efforts are more focused on traveling than
on research dissemination. Studies with negative find­
ings (no significant difference or relationship) are fre­
quently not submitted for publication. These negative
findings can be as important to the development of
knowledge as positive findings because they direct
future research, and they should find a venue for pub­
lication and broad dissemination. Many authors strat­
egize placing these nonsignificant findings within a
publication that does have a key positive finding.
Publishing research findings is a rewarding experi­
ence, but the process demands a great deal of time
and energy. The manuscript rejections or requests for
major revisions that most authors receive can be dis­
couraging. However, you can take certain steps to
increase the probability of having a manuscript
accepted for publication. While you are developing
your study and writing the proposal, outline your plans
for publishing the findings. At this time, you and other
628 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
members of your research team should discuss and
determine authorship credit. Authorship can become a
complex issue when the research is a collaborative
project among individuals from different disciplines.
Some researchers develop the entire manuscript and
then face the decision of who will be first, second,
third, or last author. There are many ways to determine
authorship credit, but the decision should be accept­
able to all investigators involved. Authorship credit
should be given only to the individuals who made
substantial contributions to developing and imple­
menting a study and to writing the manuscript (Baggs,
2008). Some journals require that all authors sign a
written statement and checklist of their contributions.
A novice author may want to consider seeking help
from a more experienced author who can either be a
coauthor on the manuscript or receive an acknowledg­
ment for their contribution in the acknowledgment
section. In many cases, the last author is usually the
senior mentor or leader of the team who secured the
resources (funding, space, and equipment) for con­
ducting the study.
Publishing a Journal Article
Developing a manuscript for publication includes the
following steps: (1) selecting a journal, (2) developing
a query letter, (3) preparing a manuscript, (4) submit­
ting the manuscript for review, and (5) revising the
manuscript.
Selecting a Journal
Selecting a journal for publication of your study
requires knowledge of the basic requirements of the
journal, the journal’s review process, and recent arti­
cles published in the journal. A refereed journal
is peer-reviewed and uses referees or expert reviewers
to determine whether a manuscript is acceptable for
publication. In nonrefereed journals, the editor makes
the decisions to accept or reject manuscripts, but these
decisions are usually made after consultation with a
nursing expert. Most refereed journals require manu­
scripts to be reviewed anonymously, or blinded, by
two or three reviewers. In some cases, there are two
reviewers for the scientific content and one reviewer
for particular attention to the statistical content (Henly,
Bennett, & Dougherty, 2010). The reviewers are asked
to determine the strengths and weaknesses of a manu­
script, and their comments are anonymously sent
from the journal editor to the contact author. Most
academic institutions support the refereed system
and may recognize only publications that appear in
peer-reviewed journals for faculty seeking tenure and
promotion.
Having a manuscript accepted for publication
depends not only on the quality of the manuscript
but also on how closely the manuscript matches the
goals of the journal and its subscribers or audience
(Dougherty, Freda, Kearney, Baggs & Broome, 2011).
It is a good idea to review articles recently published
in the journal to which you plan to submit a manu­
script. This detailed review lets you know whether a
research topic has recently been addressed and whether
the research findings would be of interest to that jour­
nal’s readers. This process enables you to identify and
prioritize a few journals that would be appropriate for
publishing your findings. Checking the references in
your own manuscript may reveal certain journals that
publish research on your topic. Reviewing the journal’s
impact factor, the time line for their review process,
and the waiting period from acceptance to publication
date can also influence your decision on where to
submit your manuscript for review and publication.
Journal Impact Factor
Journal Citation Report provides quantitative mea­
sures to evaluate scientific journals, including data
on journal impact factors. Journal Citation Report
is produced by Thomson Institute for Scientific Infor­
mation. The impact factor is a measure of the fre­
quency with which the “average article” in a journal
has been cited in a given period of time (Garfield,
2006). A list of selected impact factors can also be
found by professional field (e.g., nursing, law, educa­
tion, economics) at http://www.sciencegateway.org/
rank/index.html.
The impact factor for a journal is calculated based
on a 3-year period and can be considered to be the
average number of times published papers are cited up
to 2 years after publication. The impact factor cannot
be calculated until all of the year’s publications are
included; the most current impact factor may be 1 to
2 years old. The impact factor for a journal can usually
be found at the journal’s website. Here is an example
of information found at one journal website in 2011:
2009 IMPACT FACTOR: 4.357
2009 IMPACT FACTOR RANKING: 2nd out of
70 reproductive medicine journals
TIMING: 84% of manuscripts published within 6
months
Generally, specialty journals in nursing have lower
impact factors than broad-based journals such as
Journal of the American Medical Association or New
England Journal of Medicine. Your work is important
to publish, and there is a journal that will match your
needs and the needs of its readership, but it would
be important to aim for the higher impact nursing
 CHAPTER 27  Disseminating Research Findings 629
journals that command a higher respect for the quality
of their reviewers and their review process and hence
the quality of their publications.
Developing a Query Letter
A query letter determines an editor’s interest in
reviewing your manuscript. This letter should be no
more than 1 page and usually includes the abstract and
the researcher’s qualifications for writing the article.
The length of the manuscript and the numbers of
tables or figures may be useful, and the editor may be
interested to know when, if ever, something on this
topic was last published in their journal. Some editors
appreciate a list of potential reviewers that you might
suggest. Address your query letter in an email to the
current editor of a journal. Indicate in the letter the
title of the manuscript you would like to submit, why
publishing the manuscript is important, and why the
readers of the journal would be interested in reading
the manuscript. Even if a letter is not required by a
journal, some researchers send a query letter because
the response (positive or negative) enables them to
make the final selection of a journal for submitting
their manuscript. Often an editor responds that the
journal is planning a special issue on a particular topic
and provides the due dates so that you can prepare
well in advance. Other journals, such as Advances in
Nursing Science, accept manuscripts submitted only
for their special topic issues, and their website will
indicate due dates by topic.
Preparing a Manuscript
A manuscript is written according to the format out­
lined by the journal. Guidelines for developing a
manuscript are usually published in the issue of the
journal or on the journal’s website. Follow these
guidelines explicitly to increase the probability of
your manuscript being accepted for publication. Infor­
mation provided for authors includes (1) directions for
manuscript preparation, (2) discussion of copyright
and conflict of interest, and (3) guidelines for submis­
sion of the manuscript.
Writing research reports for publication requires
skills in technical writing that are not used in other
types of publications. Technical writing condenses
information and is stylistic. The Publication Manual
of the American Psychological Association (APA,
2010); A Manual for Writers of Research Papers,
Theses, Dissertations (Turabian, Booth, Colomb, &
Williams, 2007); and the Chicago Manual of Style
(University of Chicago Press Staff, 2010) are consid­
ered useful sources for quality technical writing. Most
journals stipulate the format style required for their
journal. In a review of 65 nursing journals, Northam,
Yarbough, Haas, and Duke (2010) noted that 36 (55%)
require APA format.
A quality research report has no errors in punc­
tuation, spelling, or sentence structure. It is also
important to avoid confusing words, clichés, jargon,
and excessive wordiness and abbreviations. Word pro­
cessing programs have “tools” commands that have
the capacity to proofread manuscripts for errors.
However, as the author, you still need to respond and
correct sentences that the software program has identi­
fied as problematic. These program tools also do a
word count to ensure that your manuscript adheres to
the limitations specified in the journal guidelines.
Knowledge about the author guidelines provided
by the journal and a background in technical writing
will help you to develop an outline for a proposed
manuscript. The brief outlines presented in Tables
27-1 and 27-8 must be expanded in detail to guide
your writing of a manuscript. You can use the outline
to develop a rough draft of your article, which you will
revise numerous times. Present the content of your
article logically and concisely under clear headings,
and select a title that creates interest and reflects the
content. The APA manual (APA, 2010) provides
detailed directions regarding appropriate terms to use
in describing study results and manuscript prepara­
tion. Consider using an article from the journal as a
guide or template; this can help you see the general
length of an introduction and discussion section, how
tables are presented, how references are cited, and
how acknowledgments are worded.
Developing a well-written manuscript is difficult.
Often, universities and other agencies offer writing
seminars to assist students and faculty in preparing a
publication. Some faculty members who chair thesis
and dissertation committees also assist their students
in developing an article for publication. In this situa­
tion, the faculty member is almost always the second
author for the article. The APA manual (APA, 2010)
has a section on how to reduce the content of a thesis
or dissertation for publication.
When you are satisfied with your manuscript, ask
one or two colleagues to review it for organization,
completeness of content, and writing style. Colleagues’
comments can guide you in making any final revisions.
You may also want to ask a friend or family member
to read the paper. Although your friend may not under­
stand the topic or statistical results, he or she should
be able to read the paper and understand the message
you are trying to communicate. The manuscript
should be expertly typed according to the journal’s
specifications. If the journal has an international focus,
630 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
it would be important to specify that your sample is
from a particular geographic area such as the United
States. For example, if the journal is British, appropri­
ate spelling is important (e.g., “labor” would be spelled
“labour”); software spell check tools have options for
American English, British English, and other lan­
guages. The reference list for the manuscript needs to
be in a complete and correct format. Computer pro­
grams are available with bibliography systems that
enable you to compile a consistent reference list for­
matted in any commonly accepted journal style. With
these programs, you can maintain a permanent file of
reference citations. When you need a reference list for
a manuscript, you can select the appropriate references
from the collection and use the program to format for
the requirements of a particular journal.
Submitting a Manuscript for Review
Guidelines in each journal indicate the name of the
editor and the address for manuscript submission.
Submit your manuscript to only one journal at a time;
only when you confirm that your manuscript is not
accepted should you submit to a different journal.
Most journals now accept only manuscripts submitted
electronically, and the editor provides a Portable Doc­
ument Format (PDF) version to reviewers when they
accept the offer to review the manuscript. When sub­
mitting the manuscript, include your complete mailing
address, phone number, fax number, and email address.
The corresponding author who submits the manuscript
usually receives notification of receipt of the manu­
script within 1 to 2 days if sent by email, and in many
cases the notification is sent to all authors listed on the
title page of the manuscript.
Peer Review
Most journals use some form of peer review process
to evaluate the quality of manuscripts submitted for
publication. The manuscript is usually sent to two or
three reviewers with guidelines from the editor for
performing the review. In most cases, the review is
“blinded,” which means that the reviewers do not
know who the author is, and the author does not know
who is reviewing the paper. For reviewers to be
blinded, journal instructions will indicate that any
materials in the manuscript that identify the authors or
institutions should be omitted and replaced with
brackets to indicate that something was intentionally
removed from the text—“[removed for blind review].”
For research papers, reviewers are asked to evalu­
ate the validity of the study. Reviewers consider
whether the methodology was adequate to address the
research question or hypotheses and whether the
findings are trustworthy and correctly interpreted. For
example, if results were not statistically significant,
was a power analysis performed? Reviewers also
evaluate whether the discussion was appropriate given
the findings and whether the author adequately dis­
cussed the clinical implications of the findings without
going beyond the actual data. Reviewers are also
asked to comment on the relevancy of the reference
citations, the usefulness of any tables or figures, and
the consistency between the title, abstract, and text of
the manuscript. Reviewers also look for the strengths
and limitations of the study, which the authors should
convey in their discussion. Every study has its limita­
tions, and a limitation is not a reason for rejecting the
manuscript. However, reviewers want to see that the
authors have accurately identified and addressed limi­
tations for the readers.
Responding to Requests to Revise a Manuscript
After reviewing your manuscript, the journal editor
will reach one of four possible decisions: (1) accept
the manuscript as submitted, (2) accept the manuscript
pending minor revisions, (3) tentative acceptance of
the manuscript pending major revisions, or (4) rejec­
tion of the manuscript. Accepting a manuscript as sub­
mitted is extremely rare. The editor will send you a
letter that indicates acceptance and the likely date of
publication.
Most manuscripts are returned to the author for
minor or major revisions before they are published.
Many of these returned manuscripts are never revised.
If a reviewer identifies a major or minor limitation of
the study, that limitation should be considered and can
be added to your other limitations in the discussion
section. An author may also incorrectly interpret the
request for revision as a rejection and assume that
a revised manuscript would also be rejected. This
assumption is not usually true because revising a man­
uscript based on reviewers’ comments improves the
quality of the manuscript. When editors return a man­
uscript for revision, they include the reviewers’ actual
comments or a summary of the comments to direct the
revision. These reviewers and the editor have devoted
time to reviewing your manuscript, and you should
make the necessary revisions or respond with your
rationale for not making a specific change requested
by a reviewer and return the revised manuscript to the
same journal for reconsideration.
It is important to review each comment from a
reviewer carefully and make the revisions that improve
the quality of the research report without making inac­
curate statements about the study. Often an editor will
provide guidance about how to respond to a reviewer

if you ask for help. When you have revised your man­
uscript based on the reviewers’ comments, it should
be resubmitted with a cover letter explaining exactly
how you responded to each review comment or rec­
ommendation. It is helpful to number each comment
and recommendation, followed by your response and
description of your modification as well as page
number in your manuscript for that revision. You can
also provide a table with a column for the requested
changes and a second column with the changes made
or the rationale for why a change could not be made.
The second column should also include the number of
the page where a change was made. Often an editor
requests that the changes in the manuscript be identi­
fied with “track changes” or highlighting. In some
cases, you may disagree with a reviewer’s recommen­
dation. If so, provide a rationale for your dis­agreement,
but do not ignore any comment or recommendation.
Sometimes the revised manuscript and your cover
letter are returned to the reviewers, and still further
modification is requested in the paper before it is pub­
lished. Although these experiences are frustrating,
they provide the opportunity to improve your writing
skills and logical development of ideas. Frequently,
editors request that the final manuscript be submitted
electronically as a clean copy with no track changes
or highlights.
An author who receives a rejection can feel devas­
tated, but he or she is not alone and should never take
it personally. All authors, even experienced ones, have
had manuscripts rejected. Manuscripts are rejected for
various reasons. When a group of nursing journal
editors surveyed manuscript reviewers and asked them
to identify the most important indicators for a manu­
script’s contribution to nursing, there were five key
factors (Dougherty et a., 2011). Table 27-10 identifies
TABLE 27-10  Important Indicators of a
Manuscript’s Contribution   for research reports because there is more rapid access
to Nursing
Factor
Knowledge or research evidence
Timeliness/topic of current interest
Newly emerging idea
Generalizability across population
or international boundaries
Contribution to theory
From Dougherty, M. C., Freda, M. C., Kearney, M. H., Baggs, J. G., &
Broome, M. (2011). Online survey of nursing journal peer reviewers:
Indicators of quality in manuscripts. Western Journal of Nursing Research,
33(4), 506–521.
CHAPTER 27  Disseminating Research Findings 631
these issues and their mean frequencies. “Timeliness
of the topic” was more important than “contribution
to theory” and reflected the predominance of review­
ers who review for nursing research journals. When
you receive a rejection notice, give yourself a cooling-
off period and then ask colleagues to help you deter­
mine what the reviewers were saying about why the
manuscript was rejected. The editor’s introductory
comments to the author in a rejection letter can often
help to guide you in your responses to reviewers’
concerns. Most manuscripts, especially those that are
poorly written, can be revised and submitted to a dif­
ferent journal that was your second choice.
Publishing Research Findings
in Online Journals
Many print journals are moving to online formats.
These journals continue to provide their traditional
print version but also have a website with access to
some or all of the articles in the printed journal. The
number of online nursing journals is also growing.
These online-only journals have some distinct advan­
tages for authors. Email links from author to editor
and editor to reviewer allow papers to be submitted
and sent electronically and allow reviewer comments
to be sent back to the editor more quickly. Reviewer
comments are compiled by the editor more quickly
and sent to the author by email. When the author
revises the article and resubmits it by email, the editor
can ask the same reviewers to evaluate the article
again. Once the article is judged to be satisfactory, it
is accepted for publication. This process is particularly
important for international scientific communications.
Online journals use “continuous publication,” which
means that there is no wait for approved articles to
be published because the editor does not have to
wait until the next issue is scheduled for publication.
The notion of an “issue” is becoming antiquated as a
result of electronic publishing. Approved articles are
placed online almost immediately, which is important
to recent research findings by other researchers and
Number (%) clinicians interested in facilitating evidence-based
1404 (83.8%)
practice. There may be added fees for publishing in
1153 (68.8%) an online format, but the possibilities of dialogue
1141 (68.1%) with readers, including other researchers in the same
698 (41.7%) field of study, are great. However, additional work
needs to be done to ensure that online publications
534 (31.9%) are secured and that each publication is permanently
available with a permanent identifier for citation and
linking. Journal publishers have begun to assign
a Digital Object Identifier (DOI) to each article. The
DOI system is managed by the International DOI
632 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
Foundation, and identifiers are assigned to all types of
digital work. The DOI will never change, even if the
location for that work does change. The use of DOIs
is expected to increase and become accepted as the
permanent identifier for scientific and scholarly pub­
lications (DOI Handbook, 2006).
Not all online journals are refereed or provide peer
review. You may wish to check for information on the
peer review process at the online journal website. Peer
review is essential to scholars in the university tenure
track system. Electronic publishing may result in a
more open peer review process. Baggs (2011) dis­
cussed some issues to be aware of when considering
Internet postings, particularly related to multiple ver­
sions and intellectual property issues.
Publishing in an online journal has potential advan­
tages. The constraint on length of the manuscript,
imposed because of the cost of print publishing,
usually does not exist. Multiple tables, figures, graph­
ics, and even streaming audio and video are possibili­
ties with online journals. Animations can be created
to assist the reader in visualizing ideas. Links may be
established with full-text versions of citations from
other online sources. It is possible to track the number
of times the article has been accessed to assess its
impact on the scientific community. Electronic list­
servs and chat rooms may be available to discuss the
paper. All of these capabilities are not currently avail­
able with every online journal. The technology to
provide them exists, but online journals with some
of these advanced technologies cover their costs
by charging subscription fees or obtaining financing
through advertisements. Online journals are rapidly
emerging, and websites are available (www.medscape.
com/pages/public/publications) that provide a listing
and electronic access to the online health journals.
Publishing Research Findings in Books
Some qualitative studies and large, complex quantita­
tive studies are published as chapters within books,
monographs, or books. Publishing a book requires
extensive commitment on the part of the researcher.
In addition, the researcher must select a publisher and
convince the publisher to support the book project. A
prospectus must be developed that identifies the pro­
posed content of the book, describes the market
readership for the book, and includes a rationale for
publishing the book. The publisher and researcher
must negotiate a contract that is mutually acceptable
regarding (1) the content and length of the book, (2)
the time required to complete the book, (3) the per­
centage of royalties that the author will receive, (4)
any financial coverage to be offered in advance, and
(5) how it will be marketed. The researcher must fulfill
the obligations of the contract by producing the pro­
posed book within the agreed time frame. Publishing
a book is a significant accomplishment and an effec­
tive, but sometimes slow, means of communicating
research findings.
Duplicate Publications and Self-Plagiarism
Duplicate publication is the practice of publishing
the same article or major portions of the article in two
or more print or electronic media without notifying the
editors or referencing the other publication in the ref­
erence list (Baggs, 2008). Duplicate publication is a
form of plagiarism (self-plagiarism) and is not only
unethical but also poor practice because it limits an
opportunity for others to publish new knowledge, arti­
ficially inflates the importance of a study topic, clut­
ters the literature with redundant information, rewards
researchers for publishing the same content twice,
and may violate copyright law. Many journal editors
screen a manuscript for plagiarism using software pro­
grams. If portions of the material have been presented
at a scientific meeting in the form of an oral podium
or poster presentation, this should be acknowledged
along with funding sources and any potential conflict
of interest.
Journals require the submission of an original man­
uscript, not previously published, so submitting a
manuscript that has been previously published without
referencing the duplicate work or notifying the editor
of the previous publication is unethical and a form of
scientific misconduct (see Chapter 9). Previous publi­
cations related to the study must be cited in the text
of the manuscript and the reference list. Editors have
the responsibility of developing a policy on duplicate
publications and informing all authors, reviewers, and
readers of this policy. In addition, editors must ensure
that readers are informed of duplicate materials by
adequate citation of the materials in the article’s text
and reference list. A duplicate publication can result
in retractions and refusal to accept other manuscripts
for review from the author (Baggs, 2008; International
Committee of Medical Journal Editors, 2011).
KEY POINTS
• Communicating research findings, the final step
in the research process, involves developing a
research report and disseminating it through pre­
sentations and publications to audiences of nurses,
healthcare professionals, policy makers, and health­
care consumers.
 CHAPTER 27  Disseminating Research Findings 633
• Both quantitative and qualitative research reports
include four basic sections: (1) introduction, (2)
methods, (3) results, and (4) discussion.
• In a quantitative research report, the introduction
briefly identifies the problem that was studied and
presents an empirical and theoretical basis for the
study. The methods section of the report describes
how the study was conducted; the results section
reveals what was found; and the discussion section
includes the limitations, conclusions that support or
refute other published work, implications for
nursing practice, and recommendations for further
research.
• In qualitative research, the introduction section
identifies the phenomenon under investigation and
describes the type of qualitative study that was
conducted. The methods section includes both the
general and the applied methods for the study con­
ducted. The results section of the research report
includes the data analysis procedures and presenta­
tion of the findings, and the discussion section
includes conclusions, limitations, implications for
nursing, and recommendations for further research.
• The greatest impact on nursing practice can be
achieved by communicating nursing research find­
ings to nurses, other health professionals, policy
makers, and healthcare consumers.
• Research findings are presented at conferences and
meetings through oral podium and poster presenta­
tions of selected portions of your research study;
the report developed depends on the focus of the
conference, the audience, and the time designated
for each presentation.
• A poster session is a visual presentation of a study.
Conference sponsors provide information concern­
ing (1) size limitations or format restrictions for the
poster and (2) the size of the poster display area.
The home institution should provide (1) the institu­
tion’s logo to place with your title and affiliations
and (2) any requirements for the color scheme.
• Developing a manuscript for publication includes
the following steps: (1) selecting a journal, (2)
writing a query letter, (3) preparing an original
manuscript, (4) submitting the manuscript for
review, and (5) responding to requests for revision
of the manuscript.
• Selecting a journal for publication of a study
requires knowledge of the basic requirements of the
journal, the journal’s refereed status, its impact
factor, and recent articles published in the journal.
• An increasing area of concern in publishing is the
practice of duplicate publications where the same
article or major portions of the article are published
in two or more print or electronic media without
notifying the editors or referencing the other pub­
lication in the reference list.
REFERENCES
Agency for Healthcare Research and Quality (AHRQ). (2011).
Evidence-based practice centers: Synthesizing scientific evidence
to improve quality and effectiveness in health care. Retrieved from
http://www.ahrq.gov/clinic/epc/.
American Psychological Association (APA). (2010). Publication
manual of the American Psychological Association (6th ed.).
Washington, DC: Author.
Baggs, J. G. (2008). Issues and rules for authors concerning author­
ship versus acknowledgements, dual publication, self plagiarism,
and salami publishing [Editorial]. Research in Nursing & Health,
31(4): 295–297.
Baggs, J. G. (2011). The dissertation manuscript option, internet
posting, and publication [Editorial]. Research in Nursing &
Health, 34(2), 89–90.
Benner, P. E. (1994). Interpretative phenomenology: Embodiment,
caring, and ethics in health and illness. Thousand Oaks, CA: Sage
Publications.
Bennett, J. A. (2006). The consolidated standards of reporting trials
(CONSORT): Guidelines for reporting randomized trials. Nursing
Research, 54(2), 128–132.
Bernstein, N. (2011). Desperately seeking sleep. Parenting
Magazine. Retrieved from http://www.parenting.com/article/
desperately-seeking-sleep-21354392?page=0,2.
Borge, C. R., Wahl, A. K., & Moum, T. (2010). Association of
breathlessness with multiple symptoms in chronic obstructive
pulmonary disease. Journal of Advanced Nursing, 66(12),
2688–2700.
Chapman, S. A., Wides, C. D., & Spetz, J. (2010). Payment regula­
tions for advanced practice nurses: implications for primary care.
Policy, Politics, & Nursing Practice, 11(2), 89–98.
Chun, K. M. (2006). Conceptual and measurement issues in family
acculturation research. In M. H. Bornstein & L. R. Cote (Eds.),
Acculturation and parent-child relationships: Measurement and
development (pp. 63e78). Mahwah, N.J: Lawrence Erlbaum
Associates.
Chun, K. M., Chesla, C. A., & Kwan, C. M. L. (2011). “So we adapt
step by step”: Acculturation experiences affecting diabetes man­
agement and perceived health for Chinese American immigrants.
Social Science & Medicine, 72(2), 256–264.
Cohen, M. Z., Kahn, D. L., & Steeves, R. H. (2000). Hermeneutic
phenomenological research: A practical guide for nurse practi-
tioners. Thousand Oaks, CA: Sage Publications.
Conn, V. S., & Rantz, M. J. (2003). Research methods: Managing
primary study quality in meta-analyses. Research in Nursing &
Health, 26(4), 322–333.
Craig, J. V., & Smyth, R. L. (2012). The evidence-base practice
manual for nurses (3rd ed.). Edinburgh, UK: Churchill
Livingstone.
Dailey, D. E., Davis, H. C., Cooper, B. A., & Lee, K. A. (2005).
Sleep inadequacy and environmental demands among racially
diverse women. Abstract for presentation at the Western Institute
of Nursing Conference, San Francisco, CA. Available from http://
hdl.handle.net/10755/158158.
634 UNIT FOUR  Analyzing Data, Determining Outcomes, and Disseminating Research
Denzin, N. K., & Lincoln, Y. S. (2005). Introduction: The discipline
and practice of qualitative research. In N. K. Denzin & Y. S.
Lincoln (Eds.) The Sage handbook of qualitative research (3rd
ed.) (pp. 1–32). Thousand Oaks, CA: Sage.
Digital Object Identifier (DOI) Handbook. (2006). Retrieved from
doi:10.1000/186.
Donaldson, N., Rutledge, D., & Geiser, K. (2008). Role of the
external coach in advancing research translation in hospital-based
performance improvement. In K. Henriksen, J. B. Battles, M. A.
Keyes, & M. L. Grady (Eds.), Advances in patient safety:
New directions and alternative approaches (Vol. 2: Culture and
redesign). Rockville, MD: Agency for Healthcare Research &
Quality.
Dougherty, M. C., Freda, M. C., Kearney, M. H., Baggs, J. G., &
Broome, M. (2011). Online survey of nursing journal peer review­
ers: Indicators of quality in manuscripts. Western Journal of
Nursing Research, 33(4), 506–521.
Floyd, J. A., Galvin, E. A., Roop, J. C., Oermann, M. H., &
Nordstrom, C. K. (2010). Graphics for dissemination of meta-
analyses to staff nurses. Nursing Research, 18(1), 72–86.
Garfield, E. (2006). The history and meaning of the journal impact
factor. Journal of the American Medical Association, 295(1),
90–93.
Goyal, D., Gay, C., & Lee, K. A. (2010). How much does low
socioeconomic status increase the risk of prenatal and postpartum
depressive symptoms in first-time mothers? Women’s Health
Issues, 20(2), 96–104.
Henly, S. J., Bennett, J. A., & Dougherty, M. C. (2010). Scientific
and statistical reviews of manuscripts submitted to Nursing
Research: Comparison of completeness, quality, and usefulness.
Nursing Outlook, 58(4), 188–199.
Heo, S., Moser, D. K., Lennie, T. A., Zambroski, C. H., & Chung,
M. L. (2007). A comparison of health-related quality of life
between older adults with heart failure and healthy older adults.
Heart & Lung, 36(1), 16–24.
International Committee of Medical Journal Editors [ICMJE].
(2011). Uniform requirements for manuscripts submitted to bio-
medical journals: Writing and editing for biomedical publication.
Retrieved from http://www.icmje.org/index.html.
Keegan, D. A., & Bannister, S. L. (2003). Effect of colour coordina­
tion of attire with poster presentation on poster popularity. Cana-
dian Medical Association Journal, 169(12), 1291–1292.
Lee, K. A., Lentz, M. J., Taylor, D. L., Mitchell, E. S., & Woods,
N. F. (1994). Fatigue as a response to environmental demands in
women’s lives. Image: Journal of Nursing Scholarship, 26(2),
149-154.
Lee, K. A., & Gay, C. L. (2011). Can modifications to the
bedroom environment improve the sleep of new parents? Two
randomized controlled trials. Research in Nursing & Health,
34(1), 7–19.
Lentz, M. (2011a). Writing a WINning abstract. Retrieved from
http://www.ohsu.edu/son/win/pcorner.shtml.
Lentz, M. (2011b). Yes it is possible: A research report in 10 minutes.
Retrieved from http://www.ohsu.edu/son/win/pcorner.shtml .
Marshall, C., & Rossman, G. B. (2010). Designing qualitative
research (5th ed.). Thousand Oaks, CA: Sage.
Melnyk, B. M., & Fineout-Overholt, E. (2010). Evidence-based
practice in nursing & healthcare: A guide to best practice (2nd
ed.). Philadelphia: Lippincott Williams & Wilkins.
Munhall, P. (2012). Nursing research: A qualitative perspective (5th
ed.). Sudbury, MA: Jones & Bartlett Learning.
Northam, S., Yarbough, S., Haas, B., & Duke, G. (2010). Journal
editor survey: Information to help authors publish. Nurse Educa-
tor, 35(1), 29–36.
Pallant, J. (2007). SPSS survival manual. New York, NY: Open
University Press, McGraw-Hill.
Perez-Escamilla, R., & Putnik, P. (2007). The role of acculturation
in nutrition, lifestyle, and incidence of type 2 diabetes among
Latinos. The Journal of Nutrition, 137, 860-870.
Polit, D. F., & Northam, S. (2011). Impact factors in nursing jour­
nals. Nursing Outlook, 59(1), 18–28.
Pyrczak, F., & Bruce, R. R. (2007). Writing empirical research
reports: A basic guide for students of the social and behavioral
sciences (6th ed.). Los Angeles, CA: Pyrczak.
Salant, T., & Lauderdale, D. S. (2003). Measuring culture: a critical
review of acculturation and health in Asian immigrant popula­
tions. Social Science & Medicine, 57, 71-90.
Sandelowski, M. (2008). Reading, writing and systematic review.
Journal of Advanced Nursing, 64(1), 104–110.
Sandelowski, M. J. (2010). Getting it right [Editorial]. Research in
Nursing & Health, 33(1), 1–3.
Santacroce, S. J., Maccarelli, L. M., & Grey, M. (2004). Methods:
Intervention fidelity. Nursing Research, 53(1), 63–66.
Saver, C. (2006). Tables and figures: Adding vitality to your article.
AORN Journal, 84(6), 945–950.
Shelledy, D. C. (2004). How to make an effective poster. Respira-
tory Care, 49(10), 1213–1216.
Stein, K. F., Sargent, J. T., & Rafaels, N. (2007). Intervention
research: Establishing fidelity of the independent variable in
nursing clinical trials. Nursing Research, 56(1), 54–62.
Turabian, K. L., Booth, W. C., Colomb, G. G., & Williams, J. M.
(2007). A manual for writers of research papers, theses, disserta-
tions, seventh edition: Chicago style for students and research-
ers. Chicago, IL: University of Chicago Press.
University of Chicago Press Staff. (2010). The Chicago manual of
style (16th ed.). Chicago, IL: University of Chicago Press.
Wilson, I. B., & Cleary, P. D. (1995). Linking clinical variables with
health-related quality of life: A conceptual model of patient out­
comes. Journal of the American Medical Association, 273(1),
59–65.
UNIT FIVE
Proposing and Seeking Funding for Research
Writing Research Proposals
  http://evolve.elsevier.com/Grove/practice/
W
ith a background in the quantitative, qualita-
tive, outcomes, and intervention research
methodologies, you are ready to propose a
study. A research proposal is a written plan that iden-
tifies the major elements of a study, such as the research
problem, purpose, and framework, and outlines the
methods and procedures to conduct the proposed study.
A proposal is a formal way to communicate ideas about
a study to seek approval to conduct the study and
obtain funding. Researchers who are seeking approval
to conduct a study submit the proposal to a select group
for review and, in many situations, verbally defend the
proposal. Receiving approval to conduct research has
become more complicated because of the increasing
complexity of nursing studies, the difficulty involved
in recruiting study participants, and increasing con-
cerns over legal and ethical issues. In many large
hospitals and healthcare corporations, both the lawyer
and the institutional review board (IRB) evaluate the
research proposals. The expanded number of health-
care studies being conducted has led to conflict among
investigators over who has the right to recruit potential
research participants. The increased number of pro-
posed studies has resulted in greater difficulty in
obtaining funding. Researchers need to develop a
quality study proposal to facilitate university and clini-
cal agency IRB approval, obtain funding, and conduct
the study successfully. This chapter focuses on writing
a research proposal and seeking approval to conduct a
study. Chapter 29 presents the process of seeking
funding for research.
Writing a Research Proposal
A well-written proposal communicates a signifi­
cant, carefully planned research project; shows the
28
CHAPTER
qualifications of the researchers; and generates support
for the project. Conducting research requires precision
and rigorous attention to detail. Reviewers often judge
a researcher’s ability to conduct a study by the quality
of the proposal. A quality study proposal is clear,
concise, and complete. Writing a quality proposal
involves (1) developing ideas logically, (2) determin-
ing the depth or detail of the content of the proposal,
(3) identifying critical points in the proposal, and
(4) developing an esthetically appealing copy (Martin
& Fleming, 2010; Merrill, 2011; Offredy & Vickers,
2010).
Developing Ideas Logically
The ideas in a research proposal must logically build
on each other to justify or defend a study, just as a
lawyer would logically organize information in the
defense of a client. The researcher builds a case to
justify why a problem should be studied and proposes
the appropriate methodology for conducting the
study. Each step in the research proposal builds on the
problem statement to give a clear picture of the study
and its merit (Merrill, 2011). Universities, medical
centers, federal funding agencies, and grant writing
consultants have developed websites to help research-
ers write successful proposals for quantitative, qualita-
tive, outcomes, and intervention research. For example,
the University of Michigan provides an online guide
for proposal development (http://www.drda.umich.
edu/proposals/PWG/pwgcomplete.html). The National
Institute of Nursing Research (NINR, 2012) provides
online training for developing nurse scientists at http://
www.ninr.nih.gov/Training/OnlineDevelopingNurse
Scientists/. You can use a search engine of your choice,
such as Google, and search for research proposal
development training; proposal writing tips; courses
635
636 UNIT FIVE  Proposing and Seeking Funding for Research
on proposal development; and proposal guidelines
for different universities, medical centers, and govern-
ment agencies. In addition, various publications have
been developed to help individuals improve their sci-
entific writing skills (American Psychological Asso-
ciation [APA], 2010; Offredy & Vickers, 2010;
Turabian, Booth, Colomb, & Williams, 2007; Univer-
sity of Chicago Press Staff, 2010).
Determining the Depth of a Proposal
The depth or detail of the content of a proposal is
determined by guidelines developed by colleges or
schools of nursing, funding agencies, and institutions
where research is conducted. Guidelines provide spe-
cific directions for the development of a proposal and
should be followed explicitly. Omission or misinter-
pretation of a guideline is frequently the basis for
rejection or requiring revision. In addition to follow-
ing the guidelines, you need to determine the amount
of information necessary to describe each step of your
study clearly. Often the reviewers of your proposal
have varied expertise in the area of your study. The
content in a proposal needs to be detailed enough to
inform different types of readers yet concise enough to
be interesting and easily reviewed (Martin & Fleming,
2010). The guidelines often stipulate a page limit,
which determines the depth of the proposal. The rele-
vant content of a research proposal is discussed later in
this chapter and varies based on the purpose of the
proposal.
Identifying Critical Points
The key or critical points in a proposal must be evident,
even to a hasty reader. You might highlight your criti-
cal points with bold or italicized type. Sometimes
researchers create headings to emphasize critical
content, or they may organize the content into tables
or graphs. It is critical in a proposal to detail the back-
ground and significance of the research problem and
purpose, study methodology, and research production
plans (data collection and analysis plan, personnel,
schedule, and budget) (APA, 2010; Offredy & Vickers,
2010; Turabian et al., 2007).
Developing an Esthetically Appealing Copy
An esthetically appealing copy is typed without spell-
ing, punctuation, or grammatical errors. A proposal
with excellent content that is poorly typed or formatted
is not likely to receive the full attention or respect of
the reviewers. The format used in typing the proposal
should follow the guidelines developed by the review-
ers or organization. If no particular format is requested,
researchers commonly follow APA (2010) format. An
appealing copy is legible (the print is dark enough to
be read) with appropriate tables and figures to com-
municate essential information. You need to submit
the proposal by the means requested as a mailed hard
copy, an email attachment, or uploaded file.
Content of a Research Proposal
The content of a proposal is written with the interest
and expertise of the reviewers in mind. Proposals
are typically reviewed by faculty, clinical agency
IRB members, and representatives of funding insti-
tutions. The content of a proposal varies with the
reviewers, the guidelines developed for the review,
and the type of study (quantitative or qualitative)
proposed. This section addresses the content of
(1) a student proposal for both quantitative and
qualitative studies, (2) condensed research proposals,
and (3) preproposals.
Content of a Student Proposal
Student researchers develop proposals to communi-
cate their research projects to the faculty and members
of university and agency IRBs (see Chapter 9 for
details on IRB membership and the approval process).
Student proposals are written to satisfy requirements
for a degree and are usually developed according to
guidelines outlined by the faculty. The faculty member
who will be assisting with the research project (the
chair of the student’s thesis or dissertation committee)
generally reviews these guidelines with the student.
Each faculty member has a unique way of interpreting
and emphasizing aspects of the guidelines. In addition,
a student needs to evaluate the faculty member’s back-
ground regarding a research topic of interest and deter-
mine whether a productive working relationship can
be developed. Faculty members who are actively
involved in their own research have extensive knowl-
edge and expertise that can be helpful to a novice
researcher. Both the student and the faculty member
benefit when a student becomes involved in an aspect
of the faculty member’s research. This collaborative
relationship can lead to the development of essential
knowledge for providing evidenced-based nursing
practice (Brown, 2009; Craig & Smyth, 2012; Melnyk
& Fineout-Overholt, 2011).
The content of a student proposal usually requires
greater detail than a proposal developed for an agency
or funding organization. The proposal is often the first
three or four chapters of the student’s thesis or dis-
sertation, and the proposed study is discussed in the
future tense—that is, what the student will do in
conducting the research. A student research proposal
 CHAPTER 28  Writing Research Proposals 637

TABLE 28-1  Quantitative Research Proposal Guidelines for Students

Chapter I Introduction
A. Background and significance of the problem
B. Statement of the problem
C. Statement of the purpose
Chapter II Review of Relevant Literature
A. Review of theoretical literature
B. Review of relevant research
C. Summary
Chapter III Framework
A. Development of a framework
(Develop a map of the study framework, define concepts in the map, describe relationships or propositions in
the map, indicate the focus of the study, and link concepts to study variables)
B. Formulation of objectives, questions, or hypotheses
C. Definitions (conceptual and operational) of study variables
D. Definition of relevant terms
Chapter IV Methods and Procedures
A. Description of the research design
(Model of the design, strengths and weaknesses of the design validity)
B. Identification of the population and sample
(Sample size, use of power analysis, sample criteria, and sampling method including strengths and weaknesses)
C. Selection of a setting
(Strengths and weaknesses of the setting)
D. Presentation of ethical considerations
(Protection of subjects’ rights and university and healthcare agency review processes)
E. Description of the intervention if appropriate for the type of study
(Provide a protocol for implementing the intervention, detail who will implement the intervention, and describe
how intervention fidelity is ensured)
F. Selection of measurement methods
(Reliability, validity, scoring, and level of measurement of the instruments as well as plans for examining
reliability and validity of the instruments in the present study; precision and accuracy of physiological
measures)
G. Plan for data collection
(Data collection process, training of data collectors if appropriate, schedule, data collection forms, and
management of data)
H. Plan for data analysis
(Analysis of demographic data; analyses for research objectives, questions, or hypotheses; level of significance
if appropriate; and other analysis techniques)
I. Identification of limitations
(Methodological and theoretical limitations)
J. Discussion of communication of findings
References Include references cited in the proposal and follow APA (2010) format
Appendices Presentation of a study budget and timetable

usually includes a title page with the title of the pro-
posal, the name and credentials of the investigator,
university name, and the date. You need to devote time
to developing the title so that it accurately reflects the
scope and content of the proposed study (Martin &
Fleming, 2010).
Content of a Quantitative Research Proposal
A quantitative research proposal usually includes a
table of contents that reflects the following chapters
or sections: (1) introduction, (2) review of relevant
literature, (3) framework, and (4) methods and pro­
cedures. Some graduate schools require in-depth
development of these sections, whereas others require
a condensed version of the same content. Another
approach is that proposals for theses and dissertations
be written in a format that can be transformed into a
publication. Table 28-1 outlines the content often
covered in the chapters of a student quantitative
research proposal.
638 UNIT FIVE  Proposing and Seeking Funding for Research
Introduction
The introductory chapter identifies the research topic
and problem and discusses their significance and
background. The significance of the problem addresses
its importance in nursing practice and the expected
generalizability of the findings. The magnitude of a
problem is partly determined by the interest of nurses;
other healthcare professionals; policy makers; and
healthcare consumers at the local, state, national, or
international level. You can document this interest
with sources from the literature. The background
describes how the problem was identified and histori-
cally links the problem to nursing practice. Your back-
ground information might also include one or two
major studies conducted to resolve the problem, some
key theoretical ideas related to the problem, and pos-
sible solutions to the problem. The background and
significance form the basis for your problem state-
ment, which identifies what is not known and the need
for further research. Follow your problem statement
with a succinct statement of the research purpose or
the goal of the study (see Chapter 5) (Martin &
Fleming, 2010; Merrill, 2011).
Review of Relevant Literature
The review of relevant literature provides an over-
view of the essential information that will guide you
as you develop your study and includes relevant the-
oretical and empirical literature (see Table 28-1).
Theoretical literature provides a background for
defining and interrelating relevant study concepts,
whereas empirical literature includes a summary and
critical appraisal of previous studies. Here you will
discuss the recommendations made by other research-
ers, such as changing or expanding a study, in rela-
tion to the proposed study. The depth of the literature
review varies; it might include only recent studies
and theorists’ works, or it might be extensive and
include a description and critical appraisal of many
past and current studies and an in-depth discussion of
theorists’ works. The literature review might be pre-
sented in a narrative format or in a pinch table that
summarizes relevant studies (see Chapter 6) (Pinch,
1995). The literature review shows that you have a
command of the current empirical and theoretical
knowledge regarding the proposed problem (Merrill,
2011; Offredy & Vickers, 2010).
This chapter concludes with a summary. The
summary includes a synthesis of the theoretical litera-
ture and findings from previous research that describe
the current knowledge of a problem (Merrill, 2011).
Gaps in the knowledge base are also identified, with
a description of how the proposed study is expected
to contribute to the nursing knowledge needed for
evidence-based practice.
Framework
A framework provides the basis for generating and
refining the research problem and purpose and linking
them to the relevant theoretical knowledge in nursing
or related fields. The framework includes concepts and
relationships among concepts or propositions, which
are sometimes represented in a model or a map (see
Chapter 7). Middle-range theories from nursing and
other disciplines are frequently used as frameworks
for quantitative studies, and the proposition or propo-
sitions to be tested from the theory need to be identi-
fied (Smith & Liehr, 2008). The framework needs to
include the concepts to be examined in the study, their
definitions, and their link to the study variables (see
Table 28-1). If you use another theorist’s or research-
er’s model from a journal article or book, letters docu-
menting permission to use this model from the
publisher and the theorist or researcher need to be
included in your proposal appendices.
In some studies, research objectives, questions, or
hypotheses are developed to direct the study (see
Chapter 8). The objectives, questions, or hypotheses
evolve from the research purpose and study frame-
work, in particular the proposition to be tested, and
identify the study variables. The variables are concep-
tually defined to show the link to the framework, and
they are operationally defined to describe the proce-
dures for manipulating or measuring the study vari-
ables. You also will need to define any relevant terms
and to identify assumptions that provide a basis for
your study.
Methods and Procedures
The researcher describes the design or general strategy
for conducting the study, sometimes including a
diagram of the design (see Chapter 11). Designs for
descriptive and correlational studies are flexible and
can be unique to the study being conducted (Kerlinger
& Lee, 2000). Because of this uniqueness, the descrip-
tions need to include the design’s strengths and weak-
nesses. Presenting designs for quasi-experimental and
experimental studies involves (1) describing how the
research situation will be structured; (2) detailing the
treatment to be implemented (Chlan, Guttormson, &
Savik, 2011); (3) explaining how the effect of the
treatment will be measured; (4) specifying the vari-
ables to be controlled and the methods for controlling
them; (5) identifying uncontrolled extraneous vari-
ables and determining their impact on the findings; (6)
describing the methods for assigning subjects to

the treatment group, comparison or control group, or
placebo group; and (7) exploring the strengths and
weaknesses of the design (Shadish, Cook, & Camp-
bell, 2002). The design needs to account for all the
objectives, questions, or hypotheses identified in the
proposal. If a pilot study is planned, the design should
include the procedure for conducting the pilot and for
incorporating the results into the proposed study (see
Table 28-1).
Your proposal should identify the target popula-
tion to which your study findings will be generalized
and the accessible population from which the sample
will be selected. You need to outline the inclusion and
exclusion criteria you will use to select a study par-
ticipant or subject and present the rationale for these
sample criteria. For example, a participant might be
selected according to the following sample criteria:
female, age 18 to 60 years, hospitalized, and 1 day
status post abdominal surgery. The rationale for these
criteria might be that the researcher wants to examine
the effects of a selected pain management interven-
tion on women who have recently undergone hospi-
talization and abdominal surgery. The sampling
method and the approximate sample size are dis-
cussed in terms of their adequacy and limitations in
investigating the research purpose (Thompson, 2002).
A power analysis usually is conducted to determine
an adequate sample size to identify significant rela-
tionships and differences in studies (see Chapter 15)
(Aberson, 2010).
A proposal includes a description of the proposed
study setting, which frequently includes the name of
the agency and the structure of the units or sites where
the study is to be conducted. The specific setting is
often identified in the proposal but not in the final
research report. The agency you select should have the
potential to generate the type and size of sample
required for the study. Your proposal might include the
number of individuals who meet the sample criteria
and are cared for by the agency in a given time period.
In addition, the structure and activities in the agency
need to be able to accommodate the proposed design
of the study. If you are not affiliated with this agency,
it would be helpful if you had a letter of support for
your study from the agency.
Ethical considerations in a proposal include the
rights of the subjects and the rights of the agency
where the study is to be conducted. Describe how you
plan to protect subjects’ rights as well as the risks and
potential benefits of your study. Also, address the steps
you will take to reduce any risks that the study might
present. Many agencies require a written consent
form, and that form is often included in the appendices
CHAPTER 28  Writing Research Proposals 639
of the proposal. With the implementation of the Health
Insurance Portability and Accountability Act (HIPAA),
healthcare agencies and providers must have a signed
authorization form from patients to release their health
information for research. You must also address the
risks and potential benefits of the study for the institu-
tion (Martin & Fleming, 2010; Offredy & Vickers,
2010). If your study places the agency at risk, outline
the steps you will take to reduce or eliminate these
risks. It is also necessary for you to state that the pro-
posal will be reviewed by the thesis or dissertation
committee, university IRB, and agency IRB.
Some quantitative studies are focused on testing the
effectiveness of an intervention, such as quasi-
experimental studies or randomized controlled trials.
In these types of studies, the elements of the interven-
tion and the process for implementing the intervention
must be detailed (Bulecheck, Butcher, & Dochterman,
2008). You need to develop a protocol that details the
elements of the intervention and the process for imple-
menting them (see Chapter 14 and the example quasi-
experimental study proposal at the end of this chapter).
Intervention fidelity needs to be ensured during a
study so that the intervention is consistently imple-
mented to designated study participants (Chlan et al.,
2011; Santacroce, Maccarelli, & Grey, 2004).
Describe the methods you will use to measure
study variables, including each instrument’s reliabil-
ity, validity, methods of scoring, and level of measure-
ment (see Chapter 16). A plan for examining the
reliability and validity of the instruments in the present
study needs to be addressed. If an instrument has no
reported reliability and validity, you may need to
conduct a pilot study to examine these qualities. If the
intent of the proposed study is to develop an instru-
ment, describe the process of instrument development
(Waltz, Strickland, & Lenz, 2010). If physiological
measures are used, address the accuracy, precision,
sensitivity, selectivity, and error rate of the instrument
(Ryan-Wenger, 2010). A copy of the interview ques-
tions, questionnaires, scales, physiological measures,
or other tools to be used in the study is usually included
in the proposal appendices (see Chapter 17). You must
obtain permission from the authors to use copyrighted
instruments, and letters documenting that permission
has been obtained must be included in the proposal
appendices.
The data collection plan clarifies what data are to
be collected and the process for collecting the data. In
this plan you will identify the data collectors, describe
the data collection procedures, and present a schedule
for data collection activities. If more than one person
will be involved in data collection, it is important to
640 UNIT FIVE  Proposing and Seeking Funding for Research
describe methods used to train your data collectors to
ensure consistency. The method of recording data is
often described, and sample data recording sheets are
placed in the proposal appendices. Also, discuss any
special equipment you will use or develop to collect
data for the study, and address data security, including
the methods of data storage (see Chapter 20).
The plan for data analysis identifies the analysis
techniques that will be used to summarize the demo-
graphic data and answer the research objectives,
questions, or hypotheses. The analysis section is best
organized by the study objectives, questions, or
hypotheses. The analysis techniques identified need to
be appropriate for the type of data collected (Grove,
2007). For example, if an associative hypothesis
is developed, correlational analysis is planned. If a
researcher plans to determine differences among
groups, the analysis techniques might include a t-test
or analysis of variance (ANOVA) (Munro, 2005). A
level of significance (α = 0.05, 0.01, or 0.001) is also
identified (see Chapters 21 through 25). Often, a
researcher projects the type of results that will be gen-
erated from data analysis. Dummy tables, graphs, and
charts can be developed to present these results and are
included in the proposal appendices if required by the
guidelines. The researcher might project possible find-
ings for a study and indicate what support or nonsup-
port of a proposed hypothesis would mean in light of
the study framework and previous research findings.
The methods and procedures chapter of a pro-
posal usually concludes with a discussion of the
study’s limitations and a plan for communication of
the findings. Both methodological and theoretical
limitations are addressed. Methodological limitations
might include areas of weakness in the design, sam-
pling method, sample size, measurement tools, data
collection procedures, or data analysis techniques;
theoretical limitations set boundaries for the general-
ization of study findings. The accuracy with which
the conceptual definitions and relational statements
in a theory reflect reality has a direct impact on the
generalization of study findings. Theory that has
withstood frequent testing through research provides
a stronger framework for the interpretation and gen-
eralization of findings. A plan is included for com-
municating the research through presentations to
audiences of nurses, other health professionals,
policy makers, and healthcare consumers and publi-
cation (see Chapter 27).
A budget and timetable are frequently included in
the proposal appendices. The budget projects the
expenses for the study, which might include the cost
for data collection tools and procedures; special
equipment; consultants for data analysis; computer
time; travel related to data collection and analysis;
typing; copying; and developing, presenting, and pub-
lishing the final report. Study budgets requesting
external funding for researchers’ time include investi-
gators’ salaries and secretarial costs. You need a time-
table to direct the steps of your research project and
increase the chance that you will complete the project
on schedule. A timetable identifies the tasks to be
done, who will accomplish these tasks, and when these
tasks will be completed. An example proposal for a
quasi-experimental study is presented at the end of this
chapter to guide you in developing your study
proposal.
Content of a Qualitative Research Proposal
Qualitative research proposal guidelines are unique
for the development of knowledge and theories using
various qualitative research methods. A qualitative
proposal usually includes the following content areas:
(1) introduction; (2) research philosophy and general
method; (3) applied method of inquiry; and (4) current
knowledge, limitations, and plans for communication
of the study findings (Marshall & Rossman, 2011;
Munhall, 2012; Patton, 2002; Sandelowski, Davis, &
Harris, 1989). Guidelines are presented in Table 28-2
to assist you in developing a qualitative research
proposal.
Introduction
The introduction usually provides a general back-
ground for the proposed study by identifying the phe-
nomenon, clinical problem, issue, or situation to be
investigated and linking it to nursing knowledge. The
general aim or purpose of the study is identified and
provides the focus for the qualitative study to be con-
ducted. The study purpose might be followed by
research questions that direct the investigation
(Munhall, 2012; Offredy & Vickers, 2010). For
example, a possible aim or purpose for an ethno-
graphic study might be to “describe the coping pro-
cesses of Mexican American adults with type 2
diabetes receiving care in a federally funded clinic.”
The research questions might focus on the influences
of real-world problems, cultural elements, and the
clinic environment on the coping processes of these
adults. Thus, the study questions might include any
of the following: How do Mexican American adults
respond to a new diagnosis of type 2 diabetes? What
is the impact of type 2 diabetes on Mexican American
adults and their families over time? What community,
clinic, and family types of support exist for Mexican
American adults with type 2 diabetes? What does it
 CHAPTER 28  Writing Research Proposals 641

TABLE 28-2  Qualitative Research Proposal Guidelines for Students

Chapter I Introduction
A. Identify the phenomenon to be studied
B. Identify the study purpose or aim and its significance
C. State the study questions or objectives
D. Describe the evolution of the study
  1. Provide a rationale for conducting study
  2. Place the study in context historically
  3. Discuss the researcher’s experience with phenomenon
  4. Discuss the relevance of the study to nursing
Chapter II Philosophical and Conceptual Foundation and General Method for the Proposed Study
A. Identify the type of qualitative research (phenomenological research, grounded theory research, ethnographic
research, exploratory-descriptive qualitative research, and historical research) to be conducted
B. Describe the philosophical and theoretical basis for the research method
C. Explain the research assumptions
D. Discuss the general steps, procedures, and outcomes for this method
E. Translation of concepts or terms
Chapter III Method of Inquiry
A. Demonstrate the researcher’s credentials for conducting this qualitative study
B. Select a site and population
C. Describe the plan for the researcher’s role in the following
  1. Entry into the site and approval to collect data
  2. Selection of study participants
  3. Ethical considerations
D. Describe the plan for data collection
  1. Data to be collected
  2. Procedures for data collection
  3. Procedures for recording data during data collection
E. Describe the plan for data analysis conducted with data collection
  1. Steps for coding information
  2. Use of specific data analysis procedures advanced in the specific research method (phenomenology research,
grounded theory research, ethnography research, exploratory-descriptive qualitative research, and historical
research)
  3. Steps to be taken to verify the information
Chapter IV Current Knowledge, Limitations, and Plans for Communication of the Study
A. Summarize and reference relevant literature as appropriate for the type of qualitative study
B. Disclose anticipated findings, hypotheses, and hunches
C. Discuss procedures to remain open to unexpected information
D. Discuss limitations of the study
E. Identify plans for communication of findings (Marshall & Rossman, 2011; Munhall, 2012)
References Include references cited in the proposal and follow APA (2010) format
Appendices Present the study budget and timetable

mean to Mexican American adults to have their dia-
betes under control?
The introduction also includes the evolution of the
study and its significance to nursing practice, patients,
the healthcare system, and health policy. The discus-
sion of the evolution of the study often includes how
the problem developed (historical context), who or
what is affected by the problem, and the researcher’s
experience with the problem (experiential context).
Whenever possible, the significance and evolution of
the study purpose needs to be documented from the
literature (Munhall, 2012). The significance of a
study may include the number of people affected,
how this phenomenon affects health and quality of
life, and the consequences of not understanding this
phenomenon. Marshall and Rossman (2011) identi-
fied the following questions to assess the significance
of a study: (1) Who has an interest in this domain of
inquiry? (2) What do we already know about the
topic? (3) What has not been answered adequately in
642 UNIT FIVE  Proposing and Seeking Funding for Research
previous research and practice? (4) How will this
research add to knowledge, practice, and policy in
this area? The introduction section concludes with an
overview of the remaining sections that are covered
in the proposal.
Philosophical and Conceptual Foundation and
General Methods for the Proposed Study
This section introduces the philosophical and concep-
tual foundation for the qualitative research method
(phenomenological research, ethnographic research,
grounded theory research, exploratory-descriptive
qualitative research, or historical research) selected
for the proposed study. The researcher provides a
rationale for the qualitative method selected and dis-
cusses its ability to generate the knowledge needed in
nursing (see Table 28-1). The investigator introduces
the philosophy, essential elements of the philosophy,
and the assumptions for the specific type of qualitative
research to be conducted.
The philosophy varies for the different types of
qualitative research and guides the conduct of the
study. For example, a proposal for a phenomenologi-
cal study might indicate the purpose of the study is to
understand the experience of young and middle-aged
women receiving news about a family BRCA 1/2
genetic mutation. “The specific study aims are to (a)
describe the experiences of women learning about a
family BRCA 1/2 mutation, (b) describe the meaning
of genetic risk to female biologic relatives of BRCA
1/2 mutation carriers, and (3) gain an understanding
of practical knowledge used in living with risk”
(Crotser & Dickerson, 2010, p. 367). Genetic testing
has determined that 5% to 10% of breast cancers are
caused by inherited gene mutations such as BRCA 1
or BRCA 2. “Heideggerian hermeneutic phenomenol-
ogy was selected to guide this study.… By listening
to the stories of women who lived the experience,
HCPs [healthcare providers] will understand the
meaning of living with risk through the language used
to express their life view (Heidegger, 1975)” (Crotser
& Dickerson, 2010, p. 358). Assumptions about the
nature of the knowledge and the reality that underlie
the type of qualitative research to be conducted are
also identified. The assumptions and philosophy
provide a theoretical perspective for the study that
influences the focus of the study, data collection and
analysis, and articulation of the findings.
Method of Inquiry
Developing and implementing the methodology of
qualitative research require an expertise that some
believe can be obtained only through a mentorship
relationship with an experienced qualitative researcher.
The role of the researcher and the intricate techniques
of data collection and analysis are thought to be best
communicated through a one-to-one relationship.
Thus, planning the methods of a qualitative study
requires knowledge of relevant sources that describe
the different qualitative research techniques and proce-
dures (Marshall & Rossman, 2011; Miles & Huberman,
1994; Munhall, 2012; Patton, 2002), in addition to
requiring interaction with a qualitative researcher.
The proposal needs to reflect the researcher’s
credentials for conducting the particular type of quali-
tative study proposed (see Chapter 12 for details on
qualitative research methods).
Identifying the methods for conducting a qualita-
tive study is a difficult task because sometimes the
specifics of the study design emerge during the study.
In contrast to quantitative research, in which the
design is a fixed blueprint for a study, the design in
qualitative research emerges or evolves as the study is
conducted. You must document the logic and appro-
priateness of the qualitative method and develop a
tentative plan for conducting your study. Because this
plan is tentative, researchers reserve the right to
modify or change the plan as needed during the
conduct of the study (Sandelowski et al., 1989). How­
ever, the design or plan must be (1) consistent with
the philosophical approach, study purpose, and spe-
cific research aims or questions; (2) be well conceived;
and (3) address prior criticism, as appropriate (Fawcett
& Garity, 2009). The tentative plan describes the
process for selecting a site and population and the
initial steps taken to gain access to the site. Having
access to the site includes establishing relationships
that facilitate recruitment of the participants necessary
to address the research purpose and answer the
research questions. For the research question, “How
do Mexican American adults cope with a new diagno-
sis of type 2 diabetes while receiving care in federally
funded clinics?” the participants might be identified in
a specific clinic or by contacting particular healthcare
providers. Although initial contact might be made
through a clinic, the interviews and observations
might occur in the community, at family gatherings,
or in the participants’ homes.
The researcher must gain entry into the setting,
develop a rapport with the participants that will facili-
tate the detailed data collection process, and protect
the rights of the participants (Marshall & Rossman,
2011; Sandelowski et al., 1989). You need to address
the following questions in describing the researcher’s
role: (1) What is the best setting for the study? (2)
How will I ease my entry into the research site? (3)

How will I gain access to the participants? (4) What
actions will I take to encourage the participants to
cooperate? (5) What precautions will I take to protect
the rights of the participants and to prevent the setting
and the participants from being harmed? You need to
describe the process you will follow to obtain informed
consent and the actions you will take to decrease study
risks. The sensitive nature of some qualitative studies
increases the risk for participants, which makes ethical
concerns and decisions a major focus of the proposal
(Munhall, 2012; Patton, 2002).
The primary data collection techniques used
in qualitative research are observation and in-depth
interviewing. Observations can range from highly
detailed, structured notations of behaviors to ambigu-
ous descriptions of behaviors or events. The interview
can range from structured, closed-ended questions
to unstructured, open-ended questions (Marshall &
Rossman, 2011; Munhall, 2012). You need to address
the following questions when describing the proposed
data collection process: (1) What data will be col-
lected? For example, will the data be field notes from
memory, audio recordings of interviews, transcripts of
conversations, DVDs of events, or examination of
existing documents? (2) What techniques or proce-
dures will the research team use to collect the data?
For example, if interviews are to be conducted, will a
list of the proposed questions be included in the
appendix? (3) Who will collect data and provide any
training required for the data collectors? (4) Where
will sources of data be located? In historical research,
data are collected through an exhaustive review of
published and unpublished literature. (5) How will the
data be recorded and stored?
The methods section also needs to address how you
will document the research process. For example, you
might keep a research journal or diary during the
course of the study. These notes can document the
day-to-day activities, methodological events, decision-
making procedures, and personal notes about the
informants. This information becomes part of the audit
trail that you can provide to ensure the quality of the
study (Miles & Huberman, 1994; Munhall, 2012;
Patton, 2002).
The methods section of the proposal also includes
the analysis techniques and the steps for conducting
these techniques. In qualitative research, data collec-
tion and analysis often occur simultaneously. The data
are usually in the form of notes, digital files, audio
recordings, DVDs, and other material obtained from
observation, interviews, and completing question-
naires. Through qualitative analysis techniques, these
data are organized to promote understanding and
CHAPTER 28  Writing Research Proposals 643
determine meaning (see Chapter 12) (Patton, 2002).
Researchers also need to identify software programs
they plan to use for data analysis.
Current Knowledge Base, Limitations, and
Plans for Communication of the Study
This section of the proposal summarizes and docu-
ments all relevant literature that was reviewed for the
study. Similar to quantitative research, qualitative
studies require a literature review to provide a basis
for the study purpose and to clarify how this study will
expand nursing knowledge (Marshall & Rossman,
2011; Munhall, 2012). This initial literature review is
often conducted to establish the significance of the
study and to develop research questions to guide the
study. In phenomenological and grounded theory
research, an additional literature review is usually con-
ducted toward the end of the research project. The
findings from a phenomenological study are compared
and combined with findings from the literature to con-
tribute to the current knowledge of the phenomenon.
In grounded theory research, the literature is used to
explain, support, and extend the theory generated in
the study (Glaser & Strauss, 1965). In all types of
qualitative studies, the findings obtained are examined
in light of the existing literature (see Chapter 4).
You need to describe how the literature reviewed
has influenced your proposed research methods.
Biases and previous experience with the research
problem need to be addressed, as does their potential
impact on the proposed study. Often, anticipated find-
ings, hypotheses, and hunches are identified before the
study is conducted, followed by a discussion of the
procedures that might be used to remain open to new
information. You will also need to address the limita-
tions of your proposed study in the context of limita-
tions of similar studies.
Conclude your proposal by describing how you
plan to communicate your findings to various audi-
ences through presentations and publications. Often, a
realistic budget and timetable are provided in the
appendix. A qualitative study budget is similar to a
quantitative study budget and includes costs for data
collection tools, software, and recording devices; con-
sultants for data analysis; travel related to data collec-
tion and analysis; transcription of recordings; copying
related to data collection and analysis; and developing,
presenting, and publishing the final report. However,
one of the greatest expenditures in qualitative research
is the researcher’s time. Develop a timetable to project
how long the study will take; often a period of 2 years
or more is designated for data collection and analysis
(Marshall & Rossman, 2011; Munhall, 2012; Patton,
644 UNIT FIVE  Proposing and Seeking Funding for Research
2002). You can use your budget and timetable to make
decisions regarding the need for funding.
Excellent websites have been developed to assist
novice researchers in identifying an idea for quali­
tative study and developing a qualitative research
proposal and reports (see www.nova.edu/ssss/QR/
qualres.html). The Office of Behavior and Social Sci-
ences Research within the National Institutes of Health
has a website to assist researchers in developing quali-
tative and quantitative research proposals for funding
(http://grants.nih.gov/grants/writing_application
.htm). You can use these websites and other publica-
tions to promote the quality of your qualitative research
proposal. The quality of a proposal is based on the
potential scientific contribution of the research to
nursing knowledge; the research philosophy guiding
the study; the research methods; and the knowledge,
skills, and resources available to the investigators
(Marshall & Rossman, 2011; Munhall, 2012; Patton,
2002).
Content of a Condensed Proposal
The content of proposals developed for review by
clinical agencies and funding institutions is usually a
condensed version of the student proposal. However,
even though these proposals are condensed, the logical
links between components of the study need to be
clearly shown. A condensed proposal often includes
a statement of the problem and purpose; previous
research that has been conducted in the area (usually
limited to three to five studies); the framework, vari-
ables, design, sample, ethical considerations, and
plans for data collection and analysis; and plans for
dissemination of findings.
A proposal submitted to a clinical agency needs to
identify the specific setting clearly, such as the emer-
gency department or intensive care unit, and the pro-
jected time span for the study. Members of clinical
agencies are particularly interested in the data collec-
tion process and involvement of institutional person-
nel in the study. The researcher needs to identify any
expected disruptions in institutional functioning, with
plans for preventing these disruptions when possible.
The researcher must recognize that anything that
slows down or disrupts employee functioning costs
the agency money and can interfere with the quality
of patient care. By showing that you are aware of these
concerns and have proposed ways to minimize their
effects, you increase the probability of obtaining
approval to conduct your study.
Various companies, corporations, and organiza-
tions provide funding for research projects. A proposal
developed for these types of funding institutions
frequently includes a brief description of the study, the
significance of the study to the institution, a timetable,
and a budget. Most of these proposals are brief and
might contain a 1-page summary sheet or abstract at
the beginning of the proposal that summarizes the
steps of the study. The salient points of the study are
included on this page in easy-to-read, nontechnical
terminology. Some proposal reviewers for funding
institutions are laypersons with no background in
research or nursing. An inability to understand the
terminology might put the reviewer on the defensive
or create a negative reaction, which could lead to
disapproval of the study. When a funding institution
is examining multiple studies, the summary sheet is
often the basis for final decisions. The summary
should be concise, informative, and designed to sell
the study.
In proposals for both clinical and funding agencies,
the investigator needs to document his or her research
background and supply a curriculum vitae if requested.
The research review committee for approval of funding
will be interested in previous research, research pub-
lications, and clinical expertise, especially if a clinical
study is proposed. If you are a graduate student, the
committee may request the names of the university
committee members and verification that your pro-
posal has been approved by the student’s thesis or
dissertation committee and the university IRB. Chapter
29 provides details on proposals submitted for funding.
Content of a Preproposal
Sometimes a researcher sends a preproposal or query
letter rather than a proposal to a funding institution. A
preproposal is a short document of 4 to 5 pages plus
appendices that explores the funding possibilities for
a research project. The parts of the preproposal are
logically ordered as follows: “(1) letter of transmittal,
(2) proposal for research, (3) personnel, (4) facilities,
and (5) budget” (Malasanos, 1976, p. 223). The pro-
posal provides a brief overview of the proposed
project, including the research problem, purpose, and
methodology (brief description), and, most important,
a statement of the significance of the work to knowl-
edge in general and the funding institution in particu-
lar. By developing a preproposal, researchers are able
to determine the agencies interested in funding their
study and limit submission of their proposals to only
institutions that indicate an interest.
Seeking Approval for a Study
Seeking approval to conduct a study is an action that
should be based on knowledge and guided by purpose.

Obtaining approval for a study from a research review
committee or IRB requires understanding the approval
process, writing a research proposal for review, and,
in many cases, verbally defending the proposal. Little
has been written to guide the researcher who is going
through the labyrinth of approval mechanisms. This
section provides a background for obtaining approval
to conduct a study.
Clinical agencies and healthcare corporations
review studies for the following reasons: (1) to evalu-
ate the quality of the study, (2) to ensure that adequate
measures are being taken to protect human subjects,
and (3) to evaluate the impact of the study on the
reviewing institution (Merrill, 2011; Offredy &
Vickers, 2010). What does the researcher hope to
result from this institutional review? Most researchers
hope to receive approval to collect data at the review-
ing institution and to obtain support for the proposed
study. IRB reviews sometime identify potential risks
or problems related to studies that can be resolved
before the studies are implemented.
Approval Process
An initial step in seeking approval is to determine
exactly what committees in which agencies must grant
approval before the study can be conducted. You need
to take the initiative to determine the formal approval
process rather than assume that you will be told if a
formal review system exists. Information on the
formal research review system might be obtained from
administrative personnel, an online website, special
projects or grant officers, chairs of IRBs in clinical
agencies, clinicians who have previously conducted
research, university IRB chairs, and university faculty
who are involved in research.
Graduate students usually require approval from
their thesis or dissertation committee, the university
IRB, and the agency IRB where the data are to be
collected. University faculty members conducting
research seek approval for the studies from the univer-
sity IRB and the agency IRB. Nurses conducting
research in an agency where they are employed must
seek approval only at that agency. If the researcher
seeks outside funding, additional review committees
are involved. Not all studies require full review by the
IRB (see Chapter 9 for the types of studies that qualify
for exempt or expedited review). However, the IRB,
not the researcher, determines the type of review that
the study requires for conduct in that agency.
When multiple committees must review a study,
sometimes the respective committees agree that the
review for the protection of human subjects will be
done by only one of the committees, with the findings
CHAPTER 28  Writing Research Proposals 645
of that committee generally being accepted by the
other committees. For example, if the university IRB
examined and approved a proposal for the protection
of human subjects, funding agencies usually recognize
that review as sufficient. Reviews in other committees
focus on approval to conduct the study within the
institution or decisions to provide study funding.
As part of the approval process, the researcher must
determine the agency’s policy regarding (1) the use of
the name of the clinical facility in reporting findings,
(2) the presentation and publication of the study, and
(3) the authorship of publications. The facility’s name
is used when presenting or publishing a study only
with prior written administrative approval. The
researcher may feel freer to report findings that could
be interpreted negatively in terms of the institution if
the agency is not identified. Some institutions have
rules that limit what is presented or published in a
study, where it is presented or published, and who is
the presenter or author. Before conducting a study,
researchers, especially employees of healthcare agen-
cies, must clarify the rules and regulations of the
agency regarding authorship, presentations, and pub-
lications. In some cases, recognition of these rules
must be included in the proposal if it is to be approved.
Preparing Proposals for Review Committees
The initial proposals for theses and dissertations are
often developed as part of a formal class. The faculty
members teaching the class provide students with spe-
cific proposal guidelines approved by the graduate
faculty and assist them in developing their initial pro-
posals. If students elect to conduct a thesis, they ask
an appropriate faculty member to serve as chair. With
the assistance of the chair, the student identifies com-
mittee members with expertise in the focus of the
proposed study or in conducting research who can
work effectively together to refine the final proposal.
The number of committee members varies for theses
(usually a chair and two members) and dissertations
(often a chair and four members) and with the univer-
sity requirements. This proposal requires approval
by the thesis or dissertation committee and the univer-
sity IRB.
Conducting research in a clinical agency requires
approval by the agency IRB. This committee has the
responsibility to (1) provide researchers with copies
of institutional policies and requirements, (2) screen
proposals for conducting research in the agency, and
(3) assist the researcher with the IRB process. The
approval process policy and proposal guidelines are
usually available from the chair of the IRB; and the
guidelines should be followed carefully, particularly
646 UNIT FIVE  Proposing and Seeking Funding for Research
page limitations. Some committees refuse to review
proposals that exceed these limitations. Reviewers on
these committees are usually evaluating proposals in
addition to other full-time responsibilities, and their
time is limited.
Investigators also need to be familiar with the
IRB’s process for screening proposals. Most agency
IRBs screen proposals for (1) scientific merit, (2) pro-
tection of human rights, (3) congruence of the study
with the agency’s research agenda, and (4) impact of
the study on patient care (Merrill, 2011). Researchers
need to develop their proposal with these ideas in
mind. They also need to determine whether the com-
mittee requires specific forms to be completed and
submitted with the research proposal. Other important
information can be gathered by addressing the follow-
ing questions: (1) How often does the committee
meet? (2) How long before the next meeting? (3) What
materials should be submitted before the meeting?
(4) When should these materials be submitted?
(5) How many copies of the proposal are required?
(6) What period of time is usually involved in com-
mittee review?
Social and Political Factors
Social and political factors play an important role in
obtaining approval to conduct a study. You need to
treat the review process with as much care as devel-
opment of the study. The dynamics of the relation-
ships among committee members is important to
assess. This detail is especially important in the
selection of a thesis or dissertation committee to
ensure that the members are willing to work together
productively. Thorough assessment of the social and
political situation in which the study will be reviewed
and implemented may be crucial to success of
the study.
Clinical agency IRBs may include nurse clinicians
who have never conducted research, nurse research-
ers, and researchers in other disciplines. The reactions
of each of these groups to a study could be very dif-
ferent. Sometimes IRB committees are made up pri-
marily of physicians, which is frequently the case in
health science centers. Physicians are often not ori-
ented to nursing research methods and might need
additional explanations related to the research meth-
odology. However, most physicians are strong sup-
porters of nursing research, helpful in suggesting
changes in design to strengthen the study, and eager
to facilitate access to subjects.
The researcher needs to anticipate potential
responses of committee members and prepare the
proposal to elicit a favorable response. It is wise to
meet with the chair of the agency IRB early in the
development of a proposal. This meeting could facili-
tate proposal development, rapport between the
researcher and agency personnel, and approval of the
research proposal.
In addition to the formal committee approval mech-
anisms, you will need the tacit approval of the admin-
istrative personnel and staff who are affected by the
conduct of your study. Obtaining informal approval
and support often depends on the way in which a
person is approached. Demonstrate interest in the
institution and the personnel as well as interest in the
research project. The relationships formed with agency
personnel should be equal, sharing ones because these
people can often provide ideas and strategies for con-
ducting the study that you may not have considered.
The support of agency personnel during data collec-
tion can also make the difference between a successful
and an unsuccessful study (Merrill, 2011).
Conducting nursing research can benefit the institu-
tion as well as the researcher. Clinicians have an
opportunity to see nursing research in action, which
can influence their thinking and clinical practice if the
relationship with the researcher is positive. These cli-
nicians may be having their first close contact with a
researcher, and interpretation of the researcher’s role
and the aspects of the study may be necessary. In addi-
tion, clinicians tend to be more oriented in the present
than researchers are, and they need to see the immedi-
ate impact that the study findings can have on nursing
practice in their institution. Interactions with research-
ers might help clinicians see the importance of research
in providing evidence-based practice and encourage
them to become involved in study activities in the
future (Offredy & Vickers, 2010). Conducting research
and providing evidence-based practice are essential if
a hospital is to achieve and maintain Magnet status.
The award of Magnet status from the American Nurses
Credentialing Center (ANCC, 2012) is prestigious to
an institution and validates the excellence in evidence-
based nursing care provided by the facility.
Verbal Presentation of a Proposal
Graduate students writing theses or dissertations are
frequently required to present their proposals verbally
to university committee members, which are called
thesis or dissertation proposal defenses. Most clinical
agencies require researchers to meet with the IRB to
discuss their proposals. In a verbal presentation of a
proposal, reviewers can evaluate the researcher as a
person, the researcher’s knowledge and understanding

of the content of the proposal, and his or her ability to
reason and provide logical explanations related to the
study. These face-to-face meetings give the researcher
the opportunity to encourage committee members to
approve his or her study.
Appearance is important in a personal presentation
because it can give an impression of competence or
incompetence. These presentations are business-like,
with logical and rational interactions, so one should
dress in a business-like manner. The committee might
perceive individuals who are casually dressed as not
valuing the review process.
Nonverbal behaviors are important during the
meeting as well; appearing calm, in control, and con-
fident projects a positive image. Plan and rehearse
your presentation to reduce anxiety. Obtain informa-
tion on the personalities of committee members, their
relationships with each other, the vested interests of
each member, and their areas of expertise because this
can increase your confidence and provide a sense of
control. It is important to arrive at the meeting early
to assess the environment for the meeting and select
a seat carefully. Because you are the presenter, all
members of the committee need to be able to see you.
However, selecting a seat on one side of a table with
all the committee members on the other side could
make you feel uncomfortable and simulate an inter-
rogation rather than a scholarly interaction. Sitting at
the side of a table rather than at the head might be a
strategic move to elicit support.
The verbal presentation of the proposal usually
begins with a brief overview of the study. Your pre-
sentation needs to be carefully planned, timed, and
rehearsed. Salient points should be highlighted, which
you can accomplish with the use of audiovisuals. After
the presentation, the reviewers will ask questions, so
be prepared to defend or justify the methods and pro-
cedures used in your study. Sometimes it is beneficial
to practice responding to questions related to the study
with a friend; this rehearsal will help you to determine
the best way to defend your ideas without appearing
defensive. When the meeting ends, you need to thank
the members of the committee for their time. If the
committee did not make a decision regarding the study
during the meeting, ask when the decision will be
made.
Revising a Proposal
Reviewers sometimes suggest changes in a proposal
that improve the study methodology; however, some
of the changes requested may benefit the institution
but not the study. Remain receptive to the suggestions,
CHAPTER 28  Writing Research Proposals 647
explore with the committee the impact of the changes
on the proposed study, and try to resolve any conflicts.
Usually reviewers make valuable suggestions that
might improve the quality of a study or facilitate the
data collection process. Revision of the proposal is
often based on these suggestions before the study is
implemented.
Sometimes a study requires revision while it is
being conducted because of problems with data col-
lection tools or subjects’ participation. However, if
clinical agency personnel or representatives of funding
institutions have approved a proposal, the researcher
needs to examine the situation seriously before making
major changes in the study. Before revising a proposal,
address three questions: (1) What needs to be changed?
(2) Why is the change necessary? (3) How will the
change affect implementation of the study and the
study findings? Students need to seek advice from
the faculty before revising their studies. Sometimes it
is beneficial for seasoned researchers to discuss their
proposed study changes with other researchers or
agency personnel for suggestions and additional
viewpoints.
If a revision is necessary, revise your proposal
and discuss the change with members of the IRB in
the agency where the study is being conducted. The
IRB members might indicate that the investigators
can proceed with the study or that the revised pro-
posal might need additional review. If a study is
funded, the study changes must be discussed with the
representatives of the funding agency. The funding
agency has the power to approve or disapprove the
changes. However, realistic changes that are clearly
described and backed with a rationale will probably
be approved.
Example of a Quantitative
Research Proposal
An example proposal of a quasi-experimental study
is included to guide you in developing a research
proposal for a thesis, dissertation, or research project
in your clinical agency. The content of this proposal
is brief and does not include the detail normally pre-
sented in a thesis or dissertation proposal. However,
the example provides you ideas regarding the content
areas that would be covered in developing a proposal
for a quasi-experimental study. This proposal was
developed by Dr. Kathy Daniel (2011), faculty at
The University of Texas at Arlington College of
Nursing.
648 UNIT FIVE  Proposing and Seeking Funding for Research
“THE EFFECT OF NURSE PRACTITIONER
DIRECTED TRANSITIONAL CARE ON
MEDICATION ADHERENCE AND
READMISSION OUTCOMES OF ELDERLY
CONGESTIVE HEART FAILURE PATIENTS
Kathryn Daniel, PhD, RN, ANP-BC, GNP-BC
Chapter 1
Introduction
Hospitalized patients with chronic health diagnoses such
as congestive heart failure (CHF), pneumonia, and
stroke are often readmitted to acute care hospitals
within a 30 day interval for potentially preventable eti-
ologies. These unnecessary readmissions carry a signifi-
cant cost to Medicare and have been targeted for
non-reimbursement. Hospitals and healthcare systems
are eager to implement programs that can safely and
effectively reduce unnecessary readmissions. Their
interests are also tempered by the realization that either
way, whether by administrative non-reimbursement
policy or actual prevention of unnecessary readmissions,
such admissions will no longer be the source of revenue,
but rather a cost to the organization. Even though some
readmissions will not be preventable, the burden will
likely be on the hospital organization to justify payment
(Stauffer et al., 2011).
Estimates of the prevalence of heart failure vary.
However, older adults, defined as those 65 years of age
and older, have documented higher rates of CHF,
6-10%. The trends over the past decade are an older
age at first hospital admission for adults with CHF and
an older age at death. This is probably secondary to
technological advances and evidence-based guidelines
for the care of individuals with heart failure. Despite
these trends, the cost for management of CHF in the
United States (U.S.) accounts for nearly 2% of the total
cost of health care in the country (Mosterd & Hoes,
2007; Solomon et al., 2005).
CHF patients have one of the highest readmission
rates to the hospital within 30 days of any diagnosis.
Nationally 25% of patients discharged from the hospi-
tal after an acute care stay for heart failure, are read-
mitted to the hospital within 30 days (Jencks, Williams,
& Coleman, 2009). Reports are as high as 50% of
those readmitted from the community had no
follow-up with their primary care provider prior to
readmission. When patients are readmitted to the hos-
pital within the 30 day period, hospitals may not be
reimbursed for subsequent hospitalizations. In 2004,
premature CHF readmissions cost the Medicare
system an estimated 17.4 billion dollars (Jencks et al.,
2009).
Prognosis remains poor once CHF is diagnosed. From
the date of index hospitalization, the 30-day mortality
rate is between 10 and 20%. Mortality at one year,
and five years is estimated between 30 to 40% and
60 to 70% respectively. Most individuals will die with
progressively worsening symptoms while others will
succumb to fatal arrhythmias (Mosterd & Hoes, 2007;
Solomon et al., 2005). With these high morbidity
and mortality rates, individuals with CHF need addi-
tional healthcare in the community to manage their
disease and decrease their rates of premature hospital
readmission.
Chapter 2
Review of Relevant Literature
Care for this population is fragmented and uncoordi-
nated. Systems of care today often are connected to
sites of care, so when patients are discharged from
acute care settings to home or to other settings and
back again, there are many opportunities for gaps in
care. Vulnerable complex frail patients with new prob-
lems or questions about management of existing prob-
lems have few knowledgeable resources to help them
navigate the new landscape of their health. More and
more hospital care is rendered by hospitalist providers
who do not follow patients after discharge from the
acute care setting, but refer patients back to their out-
patient providers for care after discharge. Communica-
tion between inpatient and outpatient silos of care may
be absent and is frequently delayed.
Medically complex patients who have multiple chronic
diseases and few socioeconomic resources are the most
vulnerable within this group and most likely
to be readmitted. Silverstein, Qin, Mercer, Fong, and
Haydar (2008) found that male African American
patients over age 75 with multiple medical comorbidi-
ties, admitted for a medicine service admission (not
surgical) and who had Medicare only as a payer source
have the highest risk of readmission. CHF was the
highest single predictor of readmission, but other
co-morbidities such as cancer, COPD, or chronic renal
failure were also contributing factors. The period of
greatest vulnerability for readmission is within the first
month after hospitalization and before patients have
been seen in the office by their primary care provider
(PCP) after hospital discharge.

Adverse drug events are a leading cause of readmis-
sion (Morrissey, McElnay, Scott, & McConnell, 2003).
Medication reconciliation and adherence are important
in the post discharge situation. Patients and families do
the best they can to relay their drug information to
inpatient providers, but they may forget things or
assume the provider knows what they are taking.
Because patients have had an acute change in their
health, their medication regimens are often modified
during their hospital stay. In addition, inpatient medica-
tion choices are influenced by hospital formularies. Even
when diligent providers discharge patients with pre-
scriptions for their new or modified medications, these
choices may not be available on the patients’ drug for-
mulary plan. So when they present these prescriptions
to their local pharmacy after discharge from the hospi-
tal, the new medication may not be available to them
or is too costly for them to afford. Inpatient providers
may also be unaware of all the medications the patient
already has at home and duplicate drugs or drug classes
that the patient has on hand (Corbett, Setter, Daratha,
Neumiller, & Wood, 2010).
Nurse practitioners (NPs) are educated to manage
chronic diseases and understand systems of care. Thus,
NPs are in a unique position within the healthcare
system to have significant positive effects on patient
outcomes, thereby decreasing readmissions, improving
patient physical and mental health outcomes, and
decreasing the costs of care (Naylor et al., 2004). Trials
using the transitional care model have been very favor-
able, both in controlled research settings and in real
world settings. Patients followed by a transitional care
NP have had substantial reduction in 30 day readmis-
sions (Naylor et al., 2004; Neff, Madigan, & Narsavage,
2003; Stauffer et al., 2011; Zhao & Wong, 2009). Yet in
spite of success in prevention of unnecessary readmis-
sions, balancing the cost of such programs must be
weighed against decreasing revenue streams before
hospitals will support them (Stauffer et al., 2011).
“Transitional care programs utilizing advanced prac-
tice nurses have consistently reduced readmissions of
vulnerable patients. Medication management is an
important part of the transitional care NP role. What is
not known is the effect of a transitional care NP program
focused on medication management on readmission
rates and medication adherence of elderly individuals
with CHF. Thus, the purpose of this study is to examine
the effects of an NP directed transitional care program
on the hospital readmission rate and medication adher-
ence of elderly CHF patients.
CHAPTER 28  Writing Research Proposals 649
Chapter 3
Framework
The Transitional Care Model provides comprehensive
in-hospital planning and home follow-up for chronically
ill, high-risk older adults hospitalized for common
medical and surgical conditions [see Figure 28-1]. This
model was initially developed by Dorothy Brooten in
the 1980s with a population of high risk pregnant
women and low birth weight infants (Brooten et al.,
1987; Brooten et al., 1994). Later Naylor and colleagues
(2004) developed it further in high risk elderly popula-
tions focusing on patients with CHF (Brooten et al.,
2002; Naylor et al., 2004). Multiple randomized con-
trolled trials (RCTs) support its effectiveness in reducing
unnecessary readmissions (Naylor et al., 2004; Neff
et al., 2003; Ornstein, Smith, Foer, Lopez-Cantor, &
Soriano, 2011; Williams, Akroyd, & Burke, 2010; Zhao
& Wong, 2009).
The goals of care provided by the transitional care
model focus on empowering the patient and family
through coordination of care and medical management
of disease and co-morbidities as needed with the ability
to make changes immediately based on set protocols,
health literacy, self-care management, and collaboration
with other providers and families to prevent unneces-
sary hospital readmissions. Figure 28-1 illustrates the
inter-relationship of concepts of this model (Transitional
care model—when you or a loved one requires care).
Patients who are more vulnerable, either socially or
physically, or complex, would utilize more aspects of the
transitional care model, whereas patients with more
resources (social and physical) need less support during
transitions of care. According to this model’s conceptual
relationships, when advanced practice nurses educate
patients about self-management skills, they are more
adherent to the overall plan of care. Thus, these chroni-
cally ill individuals have fewer unnecessary readmissions
and greater medication adherence (Brooten, Youngblut,
Kutcher, & Bobo, 2004).
The purpose of this study is to determine the effect
of an NP directed transitional care program on medica-
tion adherence and hospital readmission rate of dis-
charged elderly adults with CHF. The independent
variable (IV) is the NP directed transitional care program
and the dependent variables (DVs) are medication
adherence and hospital readmission rates. This study
will compare the medication adherence and read­mission
rate of medically complex elderly CHF patients who
receive NP directed transitional care with medication
650 UNIT FIVE  Proposing and Seeking Funding for Research

Screening

Maintaining
relationship Engaging
elder/caregiver

Coordination of care Managing symptoms

Patient

Educating/promoting
Assuring continuity
self-management

Collaborating

Figure 28-1  Transitional Care Model. (From Transitional Care Model. Retrieved from http://www.transitionalcare.info/)

management, and those who receive standard home Conceptual Operational

health nursing services. The following table summarizes Variables Definitions Definitions
the conceptual and operational definitions for the inde-
pendent (IV) and dependent variables (DV) in this study. DV: hospital Outcome which reflects Any unplanned
readmission inadequate training and readmission to an
Conceptual Operational rate preparation of patients/ acute care hospital
Variables Definitions Definitions family to manage new/ reported to study
IV: nurse Time-limited services Enrollment and chronic health conditions or investigators within 30
practitioner delivered by specially trained participation in a breakdown in communication days of hospital
directed nurse practitioners to at risk nurse practitioner between patient/family and discharge. Number of
transitional populations designed to directed transitional provider (Coleman & Boult, days from hospital
care program ensure continuity and avoid care program 2003). discharge to
preventable poor outcomes as including medication readmission will be
they move across sites of care management after an measured.
and among multiple providers acute care hospital
(Brooten et al., 1987; stay for CHF (see
Coleman & Boult, 2003). protocol in Appendix
A).

Conceptual Operational
Variables Definitions Definitions
DV: medication Adherence to the medical Score on the Morisky
adherence plan of care which reflects Medication Adherence
shared values, goals, and Scale measured on
decision making between intake and at 30 days
patients, families, and from index
providers (Rich, Gray, hospitalization
Beckham, Wittenberg, & discharge (Morisky,
Luther, 1996). Ang, Krousel-Wood, &
Ward, 2008).
Hypotheses:
1. CHF patients receiving an NP directed transitional
care program with medication management will
have greater medication adherence than CHF
patients who receive standard home health nursing
services after discharge from an acute care
hospitalization for CHF.
2. CHF patients receiving an NP directed transitional
care program with medication management will
have fewer readmissions within 30 days of
discharge from index hospitalization and number of
days to readmission will be greater than CHF
patients who receive standard home health nursing
services after discharge from an acute care
hospitalization for CHF.
Chapter 4
Methods and Procedures
Design
The design for this study will be a quasi-experimental
pretest posttest design comparing readmission out-
comes of patients who received NP led transitional care
with similar patients who did not receive transitional
care at 30 days after index hospitalization discharge
(Burns & Grove, 2009). Figure 28-2 provides a model
of the study design identifying the implementation of the
IV (see Appendix A) and the measurement of the DVs.
The study will also compare pretest posttest medication
adherence scores between the experimental and stan-
dard care groups at 30 days. The protocol for conduct-
ing the study is presented in Appendix B. The proposal
will be submitted to the Institutional Review Boards
(IRBs) of The University of Texas at Arlington (UTA) and
a selected healthcare system for approval. After approv-
als are obtained, patients admitted to one of the partici-
pating hospitals in the system who have an admitting
diagnosis of CHF will be screened for eligibility. Eligible
patients will be approached by study personnel who will
CHAPTER 28  Writing Research Proposals 651
explain the opportunity to participate in the study after
discharge from the hospital. Patients who consent to
participate will be randomized into either the experi-
mental (intervention) group or the comparison (stan-
dard care) group. Demographic information, medical
status, and pretest medication adherence will be col-
lected from all patients who consent to be in the study
before discharge from the hospital. Outcome measures
(hospital readmission rate and posttest medication
adherence) will be recorded at 30 days after discharge
using the data collection form in Appendix C. The
pretest and posttest design with a comparison group
has uncontrolled threats to validity due to selection,
maturation, instrumentation, and the possible interac-
tion between selection and history (Burns & Grove,
2009; Shadish, Cook, & Campbell, 2002). Randomiza-
tion of subjects to the treatment, controlled implemen-
tation of the study treatment, and quality measurement
methods strengthen the study design.
Ethical Considerations
University and Clinical Agency IRB approvals will be
obtained. All study personnel who have access to the
data or to participants will complete human subject
protection training before beginning to participate in
study delivery. All participants will have the study
explained to them in detail and have all their questions
answered before signing consent forms to participate in
the study. The consent form for this study is presented
in Appendix D. The participants will receive a copy of
their signed consent form.
Time frame: This entire study is projected to take
one year. Subject recruitment will begin after IRB
approval and informational in-services are presented to
the nursing and social work staff in the participating
hospitals. Data collection and analysis of readmission
outcomes and mortality will begin with the recruitment
of participants and will end 30 days after the last partici-
pant is recruited (see the Study Protocol in Appendix B).
Intervention and Procedures
Patients who consent to participate in the study will be
visited by the transitional care NP who will be following
them after discharge for an intake visit before they are
discharged from the hospital. The same NP will visit the
patient in their home within 24 hours of discharge from
the hospital to monitor the patient’s condition, review
the goals of care and plans for care, provide patient
education as needed, and manage any new issues as they
emerge. The NP will also manage all aspects of the
patients’ medications. The NP will make at least weekly
652 UNIT FIVE  Proposing and Seeking Funding for Research
Measurement of
dependent variable(DV)
Randomized Pretest
experimental group
of discharged CHF Medication adherence
patients
Randomized control
Pretest
group of discharged
CHF patients Medication adherence
Treatment: Under control of researcher
Approach to analysis:
Comparison of pretest and posttest scores
Comparison of comparison and experimental groups
Comparison of pretest/posttest differences between samples
Uncontrolled threats to validity:
Testing
Instrumentation
Mortality
Restricted generalizability as control increases
Figure 28-2  Classic experimental design.
home visits for the entire study period, carefully inquir-
ing about any interval emergency department visits or
hospital admissions. Patients who are readmitted to the
hospital may be retained in the study for the full study
period (30 days) even though they have already reached
the end-point of readmission so that medication adher-
ence can be measured. At the end of the 30 days, all
patients in the study will be contacted and/or visited at
home by study staff to capture outcome measures. The
intervention and study protocols were developed to
ensure intervention fidelity (see Appendices A and B)
(Dumas, Lynch, Laughlin, Smith, & Prinz, 2001; Erlen &
Sereika, 2006; Moncher & Prinz, 1991).
NPs who will be delivering transitional care to study
patients will receive study related training that explicitly
reviews the 2009 Focused Update incorporated into the
American College of Cardiology Foundation/American
Heart Associated (ACCF/AHA) 2005 Guidelines for the
Diagnosis and Management of Heart Failure in Adults
(Jessup et al., 2009) as well as training in study protocols
Manipulation of Measurement of
independent variable(IV) dependent variable(DV)
Treatment Posttest
NP directed transitional Medication adherence
care program Hospital readmission rate
Posttest
Medication adherence
Hospital readmission rate
(weekly visits) and study related measures. Since only
40 patients will be in the intervention group, one NP is
expected to be able to manage 40 patients over a one
year period. To ensure study continuity and coverage
for holidays and scheduled absences, a second NP
employed in the agency will also be trained. Study
recruitment and outcome measures will be accom-
plished via a study registered nurse who will be trained
on study information and procedures (see Appendix B)
and the patient consent process.
Subjects and Setting
Sample criteria: An electronic search of the inpatient
database each night at midnight will reveal all patients in
the participating hospitals with qualifying diagnosis of
CHF who are age 75 or older. Other inclusion criteria
are the patient must be at least 75 years old, have a
minimum of three chronic disease states, male gender,
and have Medicare, Medicaid, and or charity status as a
payer source. These criteria are selected based on

information from Silverstein and colleagues (2008),
which revealed these characteristics specifically increased
risk of readmission in a similar population. Study person-
nel will eliminate any patients who have already been
offered participation. Patients who are on ventilator
support or vaso-active drips will be deferred until they
are stable enough to begin discharge planning. Patients
who are being discharged on hospice or who have
already participated are not eligible to participate.
Patients who are on dialysis will be excluded due to their
unique needs and resources.
A power analysis was conducted to determine the
desired sample size. Since this intervention is known to
be effective in preventing readmission with a moderate
effect size, the effect size of 0.45 was chosen with α =
0.05 and power of 0.80, indicating a sample size of 70 was
required for the study with 35 participants in both the
intervention and comparison groups (Aberson, 2010).
Ten percent will be added to each group to accommo-
date for attrition. This leaves a final required sample size
of 40 for each group. When the required sample size of
80 has been secured, recruitment will stop. Due to the
large population of elderly CHF patients in these hospi-
tals, the sample is hoped to be obtained in 4-6 months.
Demographic variables of interest will be collected to
describe the study sample and compare the sample with
the population for representativeness. Race, gender,
age, chronic illnesses, marital status, educational level,
and healthcare insurance will be collected using the data
collection form in Appendix C. Socioeconomic status
and literacy are known predictors of health status and
utilization (Silverstein et al., 2008). Describing relation-
ships between these factors and patient outcomes may
be important in explaining study outcomes. The study
participants’ addresses will be obtained also for contact
by NPs following hospital discharge.
Instruments
The Morisky Medication Adherence Scale will be admin-
istered to all subjects who agree to participate in the
study during intake and at 30 days post initial hospital
discharge (Morisky et al., 2008). This tool has estab-
lished sensitivity of 93% and specificity of 53% when
utilized with a similar population of older adults taking
anti-hypertensive medications. It consists of eight
questions, seven asking for yes/no answers about the
patient’s self-reported adherence over the preceding
two weeks and a final question with a five point Likert
style question. High adherence is associated with a
score greater than 6 on the scale (see Appendix E).
Low/medium adherence was significantly associated
CHAPTER 28  Writing Research Proposals 653
with poor blood pressure control, while high adhering
patients (80.3%) were more likely to have blood pres-
sure controlled (Morisky et al., 2008). Test-retest pro-
cedures were utilized to produce consistency of
performance measures from one group of subjects on
two separate occasions which were then correlated
with the norm reference of actual blood pressure mea-
surements (Waltz, Strickland, & Lenz, 2010). Item-total
correlations were >0.30 for each of the eight items in
the scale with Cronbach’s alpha of 0.83. Confirmatory
factor analysis revealed a unidimensional scale with all
items loading to a single factor.
The Morisky Medication Adherence Scale is appropri-
ate for the proposed study because it was validated on
a similar population of older outpatients who were
mostly minority (76.5% black). The questions specifi-
cally ask about ‘blood pressure medicines’ which are the
primary medications utilized in CHF management. This
eight question instrument is derived from a previously
validated four question version (Morisky, Green, &
Levine, 1986).
Procedure
Eligible participants will have the study explained to
them by the study recruiter who will obtain consent
from those who are willing to participate. The recruiter,
a registered nurse who is part of the study team, will
also capture demographic and medical data, and admin-
ister the Morisky Medication Adherence Scale to all
participants (see Appendix E). Patients assigned to the
transitional care intervention will be visited by a transi-
tional care NP before being discharged home (see
Appendix B for Study Protocol).
On the day after discharge, the transitional care NP
will visit the patient in their home to evaluate their
home situation and resources as well as review the plan
of care. For the next 30 days, the transitional care NP
will visit the patient on at least a weekly basis. The visit
will conform to the transitional care visit guideline in
Appendix A so that intervention fidelity will be main-
tained. At all times a transitional care NP will be
available by telephone. Outcome measures (hospital
readmissions and medication adherence) will be mea-
sured at 30 days after discharge using the data collection
form in Appendix C and the Morisky Medication Adher-
ence Scale in Appendix E. The study recruiter will also
do these measures to decrease potential for bias.
Plan for Data Management and Analysis
“Demographic data will be analyzed and NP actions
and their frequency of use will be examined using
654 UNIT FIVE  Proposing and Seeking Funding for Research
descriptive statistics. All encounter content with patients
will be recorded in the electronic health record, which
all transitional care staff will have access to at all times.
The documentation of weekly scheduled visits from the
transitional care NP will follow a template so that all
areas are consistently addressed with all study partici-
pants and intervention fidelity is assured (Erlen &
Sereika, 2006). Differences in the interval level data
produced by the Morisky Medication Adherence Scale
will be examined with t-test at pre-test between the
intervention and comparison groups to ensure the
groups were similar at the start of the study. Differences
will also be examined between pretest and posttest, and
at posttest between the intervention and comparison
groups. Differences in readmission rates will be exam-
ined at 30 days between the intervention and compari-
son groups. IBM Statistical Package for Social Sciences
Statistics 19 will be used to analyze the data. Alpha will
be set at 0.05 to conclude statistical difference. The
statistical tests will be an independent t-test between
two groups and a dependent t-test comparing pre- and
posttests. Bonferroni correction for multiple t-tests will
be done to reduce the risk of a type I error (Maxwell,
& Delaney, 2004).
References
Aberson, C. L. (2010). Applied power analysis for the
behavioral sciences. New York, NY: Routledge Taylor
& Francis Group.
Brooten, D., Kumar, S., Brown, L. P., Butts, P., Finkler,
S. A., Bakewell-Sachs, S., et al. (1987). A random-
ized clinical trial of early hospital discharge and
home follow-up of very-low-birth-weight infants. In
L. T. Rinke (Ed.), Outcome measures in home care:
Research (pp. 95–106). New York, NY: National
League for Nursing.
Brooten, D., Naylor, M. D., York, R., Brown, L. P.,
Munro, B. H., Hollingsworth, A. O., et al. (2002).
Lessons learned from testing the quality cost model
of advanced practice nursing (APN) transitional care.
Journal of Nursing Scholarship, 34(4), 369–375.
Brooten, D., Roncoli, M., Finkler, S., Arnold, L.,
Cohen, A., & Mennuti, M. (1994). A randomized
trial of early hospital discharge and home follow-up
of women having cesarean birth. Obstetrics and
Gynecology, 84(5), 832–838.
Brooten, D., Youngblut, J. M., Kutcher, J., & Bobo, C.
(2004). Quality and the nursing workforce: APNs,
patient outcomes and health care costs. Nursing
Outlook, 52(1), 45–52.
Burns, N., & Grove, S. K. (2009). The practice
of nursing research: Appraisal, synthesis, and
generation of evidence (6th ed.). St. Louis, MO:
Saunders.
Coleman, E. A., & Boult, C. (2003). Improving the
quality of transitional care for persons with complex
care needs. Journal of the American Geriatrics Society,
51(4), 556–557.
Corbett, C. F., Setter, S. M., Daratha, K. B., Neumiller,
J. J., & Wood, L. D. (2010). Nurse identified hospital
to home medication discrepancies: Implications for
improving transitional care. Geriatric Nursing, 31(3),
188–196.
Dumas, J. E., Lynch, A. M., Laughlin, J. E., Smith, E. P.,
& Prinz, R. J. (2001). Promoting intervention fidelity.
conceptual issues, methods, and preliminary results
from the EARLY ALLIANCE prevention trial.
American Journal of Preventive Medicine, 20(1),
38–47.
Erlen, J. A., & Sereika, S. M. (2006). Fidelity to a
12-week structured medication adherence interven-
tion in patients with HIV. Nursing Research, 55(2),
S17-S22.
Jencks, S. F., Williams, M. V., & Coleman, E. A. (2009).
Rehospitalizations among patients in the Medicare
fee-for-service program. New England Journal of
Medicine, 360(14), 1418–1428. Retrieved from
http://dx.doi.org/10.1056/NEJMsa0803563.
Jessup, M., Abraham, W. T., Casey, D. E., Feldman, A.
M., Francis, G. S., Ganiats, T. G. et al. (2009). 2009
focused update: ACCF/AHA guidelines for the
diagnosis and management of heart failure in adults:
A report of the American College of Cardiology
Foundation/American Heart Association task force
on practice guidelines: Developed in collaboration
with the international society for heart and lung
transplantation. Circulation, 119(14), 1977–2016.
Maxwell, S. E., & Delaney, H. D. (2004). Designing
experiments and analyzing data: A model comparison
perspective (2nd ed.). Mahwah, NJ: Lawrence
Erlbaum Associates.
Moncher, F. J., & Prinz, R. J. (1991). Treatment fidelity
in outcome studies. Clinical Psychology Review, 11(3),
247–266.
Morisky, D. E., Ang, A., Krousel-Wood, M., & Ward,
H. J. (2008). Predictive validity of a medication
adherence measure in an outpatient setting. Journal
of Clinical Hypertension, 10(5), 348–354.
Morisky, D. E., Green, L. W., & Levine, D. M. (1986).
Concurrent and predictive validity of a self-reported
measure of medication adherence. Medical Care,
24(1), 67–74.
Morrissey, E. F. R., McElnay, J. C., Scott, M., &
McConnell, B. J. (2003). Influence of drugs,

demographics and medical history on hospital
readmission of elderly patients: A predictive model.
Clinical Drug Investigation, 23(2), 119–128.
Mosterd, A., & Hoes, A. W. (2007). Clinical
epidemiology of heart failure. Heart, 93(9),
1137–1146.
Naylor, M. D., Brooten, D. A., Campbell, R. L.,
Maislin, G., McCauley, K. M., & Schwartz, J. S.
(2004). Transitional care of older adults hospitalized
with heart failure: A randomized, controlled trial
[corrected] [published erratum appears in J Am
Geriatr Soc 2004 Jul;52(7):1228]. Journal of the
American Geriatrics Society, 52(5), 675–684.
Neff, D. F., Madigan, E., & Narsavage, G. (2003).
APN-directed transitional home care model:
Achieving positive outcomes for patients with
COPD. Home Healthcare Nurse, 21(8), 543–550.
Ornstein, K., Smith, K. L., Foer, D. H., Lopez-Cantor,
M. T., & Soriano, T. (2011). To the hospital and back
home again: A nurse practitioner-based transitional
care program for hospitalized homebound people.
Journal of the American Geriatrics Society, 59(3),
544–551.
Rich, M. W., Gray, D. B., Beckham, V., Wittenberg, C.,
& Luther, P. (1996). Effect of a multidisciplinary
intervention on medication compliance in elderly
patients with congestive heart failure. The American
Journal of Medicine, 101(3), 270–276.
Shadish, W. R., Cook, T. D., & Campbell, D. T. (2002).
Experimental and quasi-experimental designs for
generalized causal inference. Boston, MA: Houghton,
Mifflin and Company.
APPENDIX A: Intervention Protocol for Transitional Care Nurse Practitioner (TCNP)
Visit Protocol
1. Patients are visited for their initial visit within 24-48
hours of discharge from the hospital.
2. Only NPs who have been trained on CHF protocols
and transitional care protocols and are included on
the study IRB protocol may visit/interact with study
patients.
3. On the first visit the TCNP will review the hospital
discharge plan of care with the patient. A family
caregiver is identified on the hospital visit or first
home visit. This person should be present and
included in all visits and supervise the patient’s
needs in the home. On every visit the following will
be addressed by the TCNP.
a. Review the plan of care given to the patient on
discharge from the hospital.
CHAPTER 28  Writing Research Proposals 655
Silverstein, M. D., Qin, H., Mercer, S. Q., Fong, J., &
Haydar, Z. (2008). Risk factors for 30-day hospital
readmission in patients over 65 years of age. Baylor
University Medical Center Proceedings, 21(4),
363–372.
Solomon, S. D., Zelenkofske, S., McMurray, J. J. V.,
Finn, P. V., Velazquez, E., Ertl, G., et al. (2005).
Sudden death in patients with myocardial infarction
and left ventricular dysfunction, heart failure, or
both. New England Journal of Medicine, 352(25),
2581–2588.
Stauffer, B. D., Fullerton, C., Fleming, N., Ogola, G.,
Herrin, J., Stafford, P. M., et al. (2011). Effectiveness
and cost of a transitional care program for heart
failure: A prospective study with concurrent
controls. Archives of Internal Medicine, 14(14),
1238–1243.
Transitional care model—when you or a loved one
requires care. Retrieved 8/8/2011, 2011, from http://
www.transitionalcare.info/Abou-1783.html.
Waltz, C. F., Strickland, O. L., & Lenz, E. R. (2010).
Measurement in nursing and health research (4th ed.).
New York, NY: Springer Publishing Company.
Williams, G., Akroyd, K., & Burke, L. (2010). Evalua-
tion of the transitional care model in chronic heart
failure. British Journal of Nursing (BJN), 19(22),
1402–1407.
Zhao, Y., & Wong, F. K. Y. (2009). Effects of a
postdischarge transitional care programme for
patients with coronary heart disease in China: A
randomised controlled trial. Journal of Clinical
Nursing, 18(17), 2444–2455.
b. On all visits after the initial visit, inquire about any
unplanned visits to any hospital.
c. Ask about any new problems, issues, or
symptoms that have arisen since hospital
discharge.
d. Conduct a brief review of systems, looking
specifically for any changes since discharge from
the hospital.
e. Review log of daily weights/teach if needed to do
daily weights before breakfast and after voiding
each morning.
f. Conduct a focused physical exam with careful
attention to cardiovascular and respiratory exam
on every visit; other systems as indicated by any
patient complaints.
656 UNIT FIVE  Proposing and Seeking Funding for Research
g. Review all recommended medications with
the patient and caregiver by physically viewing
the supply. On the first visit to the home, if the
patient does not have a ‘medminder,’ the
TCNP will provide one to the patient/family at
no cost and set up the medications for the first
week. The available quantities and dosages on
hand will be monitored on all medications, not
just CHF medications. (Anticipate unexpected
problems to arise here with possible
duplication of drug classes, unavailable
meds, etc.)
h. Review indication, rationale, schedule, and
possible side effects of every medication.
i. Provide patient/family education as needed on
dietary choices, exercise, as needed medications,
etc.
j. When possible and needed the TCNP will adjust
medications as required to accommodate
individual patient plan formulary.
k. Adjust/titrate meds as indicated to achieve goals
of care.
l. Order lab tests necessary to monitor patient
response to medication changes.
m. Order any other medications/tests/referrals
indicated by patient exam and complaints.
n. Consult immediately with primary care provider
(PCP)/cardiologist for any unexpected
deterioration in patient condition.
APPENDIX B: Study Protocol
Recruiting/Intake—Study Registered
Nurse (RN)
1. Generate CHF list from hospital IT.
2. Compare list to track daily discharges of patients
already recruited.
3. Screening for eligibility: Inclusion sample criteria
a. Service area is 30 miles from the hospital:
Use Internet directions program if you are
unsure how far the patient lives from the
facility.
b. Must have heart failure diagnosis
c. 75 and + in age
d. African American
e. Male gender
f. Medicare, non-funded or Medicaid
g. Patient resides in a private residence, assisted
living facility, or residential care home.
o. Communicate any changes in medication regimen
in writing for patient/caregiver.
p. Record visit in electronic health record (EHR);
forward copy to patient’s PCP for review. Visit
template in EHR will include fields to capture the
above items ‘c-p.’
q. On final home visit at the end of 4th week,
collect Morisky Medication Adherence Scale for
study.
r. After final visit at the end of the 4th week,
compose discharge summary and send to PCP.
Study RN Protocol for Comparison Group
a. The study RN will recruit, consent, and randomize
patients. After consent is obtained, she will also
obtain demographic information and the pre-test
Morisky Medication Adherence Scale on all
participants.
b. The study RN will contact all usual care patients by
telephone at the end of each week during the study
period of four weeks to inquire about any interval
hospital admissions.
c. On the final telephone call to the usual care
participant at the end of week four, the study RN
will also collect the post-test Morisky Medication
Adherence Scale.
d. The study RN will also contact all transitional care
participants at the end of week four to collect
post-test Morisky Medication Adherence Scale.
4. Exclusion sample criteria:
a. Patients discharged home on hospice.
b. Patients on dialysis.
c. Patients on ventilators or vaso-active drips
should not be approached until they are in the
discharge planning stage.
5. If patient meets all above inclusion and exclusion
sample criteria: they will be approached for study
participation
6. Introduce yourself to the patient and family.
7. Explain the opportunity to participate in the study
after discharge from the hospital and what is
involved. If patients agree to participate, give them
consent to read or read to them if desired.
8. a Ask them to sign consent if they wish to
participate.
9. If they decline to participate, thank them for giving
you their time. Make a note in the chart that they

were offered study participation and have refused,
that they are not in the study.
10. For those patients who consent to participate in
the study:
a. Collect patients’ demographic, medical, and
educational information.
b. Administer the Morisky medication adherence
scale.
c. Confirm their address and phone number.
d. Give them your card and phone number.
e. Randomize participant to either the intervention
or comparison group. Let them know which
group they will be in and when to expect
contact again.
f. Intervention group will be visited by NP in
hospital and within 24 hours of discharge from
hospital in their residence, then weekly
throughout study period. Place a transitional
care ‘sticker’ on the chart to alert inpatient staff
that we are following the patient who was
assigned to the intervention group.
g. Usual care group will receive weekly phone
call from study RN to determine any hospital
readmissions, plus one end of study data
collection of Morisky medication adherence
scale.
Intervention Group:
A transitional care nurse practitioner is preferably certi-
fied as an Adult/Gerontology Primary Care Nurse Prac-
titioner, although other nurse practitioners with
significant geriatric expertise will be considered. Other
types of advanced practice nurses will not be included
in this trial although they were included in much of the
original studies by Brooten et al. (2004) and Naylor
et al. (2004). All transitional care nurse practitioners will
complete a standardized orientation and training
oasis-ebl|Rsalles|1476726504
program focusing on a review of national heart failure
guidelines as well as principles of geriatric care, patient
and caregiver goal setting, and educational and behav-
ioral strategies focused on patient and caregiver needs.
Scripting for Transitional Care Program
Introduction during Inpatient Visit
1. Introduction of yourself to the patient/family
2. You were randomly chosen to be in the
Transitional Care group. The goal of the program
is to help people (and their families) with heart
problems learn how to best manage their illness
at home.
CHAPTER 28  Writing Research Proposals 657
a. Heart failure has more hospital readmissions than
any other problem in the United States
b. 20% of all people discharged with this problem
return to the hospital within 30 days
c. Patients followed in transitional care programs
have had lower readmission rates.
3. This is how the program works:
a. I meet you here in the hospital (probably only
one time).
b. I come to see you very soon after you go home; I
will be there within 24-48 hours.
c. I will see you every week for one month at a
minimum; we can add more visits to this if
needed for you and your family.
4. I work with your doctors and keep them informed
of how things are going at home. I am a nurse
practitioner; I am not a home healthcare nurse,
although I will work with your home healthcare
nurse as needed. Go into more explanation re:
differences etc. as needed, give them brochure on
‘what is a nurse practitioner.’
a. Why NPs can do more.
b. NP can prescribe and make medication
changes if necessary and keep your physician
informed.
c. NP can address new problems which might come
up.
d. Your Medicare benefit and supplemental insur-
ance will pay for my visits; you will not
be billed for any uncovered co-pays.
5. The goal of the program is not to slow you down,
we do not want to interfere with your other
activities, and we want you to continue to be able
to do as much as you can do.
a. We will review your medications at every visit.
b. I will ask you each week about any
readmissions to any hospital since the previous
visit.
c. Each week we will review your plan of care, how
you are doing, and about any new problems or
issues that arise.
d. The study RN who recruited you to the study
will contact you at the end of the study and ask
you the same questions that she asked after you
initially consented to participate (Morisky
Medication Adherence Scale).
6. There will be different levels of coordination
involved with each patient.
a. I may discuss your case with your hospital nurse
and hospitalist/cardiologist if needed.
658 UNIT FIVE  Proposing and Seeking Funding for Research

b. I may discuss your case with your primary care update the database on any hospitalizations that
provider if needed during intervention period; have occurred since the last interval data collection
he/she will receive a copy of the record for every (specifically how many days since discharge to
visit. readmission). On the final call, the Morisky
c. I will provide a comprehensive discharge summary Medication Adherence Scale will also be collected.
to your PCP when discharged from transitional
care service after one month.
Study RN—Data Collection on Comparison
Group Patients
1. Call all usual care patients at 7, 14, 21, and 30 days
after discharge. On each occasion he/she will

APPENDIX C: Data Collection Form

Days since Discharge

Data Collection Form without Readmission
All Other
Pretest Heart Diagnoses Post-test
Morisky Failure (One End End End End Morisky
Medication Diagnosis Line/ of of of of Medication
Study Years of Adherence (ICD-9 ICD-9 Week Week Week Week Adherence
ID Age Gender Race Education Score Code) Code) 1 2 3 4 Score

APPENDIX D: Informed Consent
Principal Investigator Name:
Kathryn Daniel, PhD, RN, ANP-BC, GNP-BC
Title of Project:
The Effect of Nurse Practitioner Directed Transitional
Care on Medication Adherence and Read­mission Out-
comes of Elderly Congestive Heart Failure Patients
Introduction
You are being asked to participate in a research study.
Your participation is voluntary. Please ask questions if
there is anything you do not understand.
Purpose
This study is designed to examine the effects of nurse
practitioner directed transitional care program on medi-
cation adherence and hospital readmission of elderly
patients who have congestive heart failure. Nationally,
20% or more of patients who are hospitalized with
congestive heart failure are readmitted to the hospital
within 30 days, often for reasons that are preventable.
Transitional care using nurse practitioners has been
shown to have positive benefits for many people like
you after they are discharged from the hospital. This
study is designed to determine whether medication
adherence is also related to decreased hospital
readmissions.
Duration
This study will last for 4 weeks after you are discharged
from the hospital.
Procedures
After you have read this form and agreed to participate,
the intake nurse will gather some basic information from
you. Then you will be randomly assigned to receive
usual care or transitional care after you are discharged
from the hospital.
If you are assigned to the usual care group, you will
be given the care your physician orders for you to
receive upon discharge from the hospital. In addition,
you will be telephoned at your home once per week for
4 weeks by a study nurse who will ask you whether or
not you have been back to the hospital. On the 4th and
final week’s call, she will also ask you some additional
questions about how you take your medications.
“If you are assigned to the transitional care nurse
practitioner group, your assigned transitional care
nurse practitioner will come to your room and
CHAPTER 28  Writing Research Proposals 659
introduce herself to you before you are discharged from
the hospital. You will also receive the care ordered by
your doctor after you are discharged from the hospital
including at least weekly visits and telephone support
from the transitional care nurse practitioner. The tran-
sitional care nurse practitioner will work with you and
your doctors to bridge the gap between hospital dis-
charge and your return to your usual primary healthcare
provider as you learn to manage the changes in your
health.
Possible Benefits
There are no direct benefits to you for participating in
this research; however your participation will help us to
determine whether or not nurse practitioner led tran-
sitional care can decrease unnecessary hospital readmis-
sions and improve medication adherence. It is possible
that having direct access to the transitional care nurse
practitioner may provide you with more timely evalua-
tion and management of problems that occur during the
4 weeks after discharge from the hospital.
Compensation
You will not receive any compensation for your partici-
pation in this study.
Possible Risks/Discomforts
You may return to your usual state of health and activi-
ties rapidly and thus not feel the need for a visit from
the nurse practitioner or a phone call from the study
nurse every week for 4 weeks.
Alternative Procedures/Treatments
There are no alternatives to participation, except not
participating. You will always receive the care ordered
by your physician.
Withdrawal from the Study
You may discontinue your participation in this study at
any time without any penalty or loss of benefits.
Number of Participants
We expect 80 participants to enroll in this study.
Confidentiality
If in the unlikely event it becomes necessary for the
Institutional Review Board to review your research
records, then The University of Texas at Arlington
will protect the confidentiality of those records to the
660 UNIT FIVE  Proposing and Seeking Funding for Research
extent permitted by law. Your research records will not
be released without your consent unless required by
law or a court order. The data resulting from your par-
ticipation may be made available to other researchers
in the future for research purposes not detailed within
this consent form. In these cases, the data will contain
no identifying information that could associate you with
it, or with your participation in any study.
If the results of this research are published or pre-
sented at scientific meetings, your identity will not be
disclosed.
Contact for Questions
Questions about this research or your rights as a research
subject may be directed to Kathryn Daniel at (xxx)xxx-
xxxx. You may contact the chairperson of the UT Arling-
ton Institutional Review Board at (xxx)-xxx-xxxx in the
event of a research-related injury to the subject.
Consent Signatures
As a representative of this study, I have explained the
purpose, the procedures, the benefits, and the risks that
are involved in this research study:
____________________________________________
______________________________
APPENDIX E: Morisky Medication Adherence Scale
Please complete the following scale by circling the
best response that fits you.
1. Do you sometimes forget to take your medications?
Yes/no
2. Over the past 2 weeks, were there any days when
you did not take your medication? Yes/no
3. Have you ever cut back or stopped taking your
medication without telling your doctor because you
felt worse when you took it? Yes/no
4. When you travel or leave home, do you sometimes
forget to bring along your medications? Yes/no
5. Did you take your medicine yesterday? Yes/no
6. When you feel like your blood pressure is under
control, do you sometimes stop taking your
medication? Yes/no
7. Taking medication everyday is a real inconvenience
for some people. Do you ever feel hassled about
KEY POINTS
• This chapter focuses on writing a research proposal
and seeking approval to conduct a study.
Signature and printed name of principal investigator
or person obtaining consent/Date
By signing below, you confirm that you have read or
had this document read to you.
You have been informed about this study’s purpose,
procedures, possible benefits and risks, and you have
received a copy of this form. You have been given the
opportunity to ask questions before you sign, and you
have been told that you can ask other questions at any
time.
You voluntarily agree to participate in this study. By
signing this form, you are not waiving any of your legal
rights. Refusal to participate will involve no penalty or
loss of benefits to which you are otherwise entitled, and
you may discontinue participation at any time without
penalty or loss of benefits, to which you are otherwise
entitled.
____________________________________________
_________________________
Signature of volunteer/Date
sticking to your blood pressure treatment plan?
Yes/no
8. How often do you have difficulty remembering to
take all your medications? (Select one)
Never
Occasionally, but less than half the time
About half the time
More than half the time
Almost all the time
Morisky, D. E., Ang, A., Krousel-Wood, M., & Ward,
H. J. (2008). Predictive validity of a medication adher-
ence measure in an outpatient setting. Journal of Clinical
Hypertension, 10(5), 348-354. doi:10.1111/j.1751-7176.
2008.07572.x.
• A research proposal is a written plan that identifies
the major elements of a study, such as the problem,
purpose, and framework, and outlines the methods
and procedures to conduct a study.

• Writing a quality proposal involves (1) developing
the ideas logically, (2) determining the depth or
detail of the proposal content, (3) identifying the
critical points in the proposal, and (4) developing
an esthetically appealing copy.
• A quantitative research proposal usually has four
chapters or sections: (1) introduction, (2) review of
relevant literature, (3) framework, and (4) methods
and procedures.
• A qualitative research proposal generally includes
the following chapters or sections: (1) introduction;
(2) philosophical and conceptual foundation and
general method; (3) method of inquiry; and (4)
current knowledge, limitations, and plans to com-
municate the study.
• Most clinical agencies and funding institutions
require a condensed proposal, which usually
includes a problem, a purpose, previous research
conducted in the area, a framework, variables,
design, sample, ethical considerations, plan for data
collection and analysis, and plan for dissemination
of findings.
• Sometimes a researcher will send a preproposal or
query letter to a funding institution rather than a
proposal; and the parts of the preproposal are logi-
cally ordered as follows: (1) letter of transmittal,
(2) proposal for research, (3) personnel, (4) facili-
ties, and (5) budget.
• Seeking approval for the conduct or funding
of a study is a process that involves submission
of a proposal to a selected group for review
and, in many situations, verbally defending that
proposal.
• Research proposals are reviewed to (1) evaluate the
quality of the study, (2) ensure that adequate mea-
sures are being taken to protect human subjects,
and (3) evaluate the impact of conducting the study
on the reviewing institution.
• Proposals sometimes require revision before or
during the implementation of a study; if a change
is necessary, the researcher needs to discuss the
change with the members of the university and
clinical agency IRBs and the funding institution.
• An example of a brief quantitative research pro-
posal of a quasi-experimental study is provided.
REFERENCES
Aberson, C. L. (2010). Applied power analysis for the behavioral
sciences. New York, NY: Routledge Taylor & Francis Group.
American Nurses Credentialing Center (ANCC). (2012). Magnet
program overview. Retrieved from http://www.nursecredentialing
.org/Magnet/ProgramOverview.aspx.
CHAPTER 28  Writing Research Proposals 661
American Psychological Association (APA). (2010). Publication
manual of the American Psychological Association (6th ed.).
Washington, DC: Author.
Brown, S. J. (2009). Evidence-based nursing: The research-practice
connection. Sudbury, MA: Jones & Bartlett Publishers.
Bulecheck, G., Butcher, H., & Dochterman, J. (Eds.). (2008).
Nursing interventions classification (NIC) (5th ed.). St. Louis,
MO: Elsevier.
Chlan, L. L., Guttormson, J. L., & Savik, K. (2011). Methods:
Tailoring a treatment fidelity framework for an intensive care unit
clinical trial. Nursing Research, 60(5), 348–353.
Craig, J. V., & Smyth, R. L. (2012). The evidence-based practice
manual for nurses (3rd ed.). Edinburgh, UK: Churchill
Livingstone.
Crotser, C. B., & Dickerson, S. S. (2010). Women receiving news
of a family BRCA 1/2 mutation: Messages of fear and empower-
ment. Journal of Nursing Scholarship, 42(4), 367–378.
Daniel, K. (2011). The effect of nurse practitioner directed transi-
tional care on medication adherence and readmission outcomes of
elderly congestive heart failure patients. Unpublished proposal.
Fawcett, J., & Garity, J. (2009). Evaluating research for evidence-
based nursing practice. Philadelphia, PA: F. A. Davis Company.
Glaser, B., & Strauss, A. L. (1965). Discovery of substantive theory:
A basic strategy underlying qualitative research. American Behav-
ioral Scientist, 8(1), 5–12.
Grove, S. K. (2007). Statistics for health care research: A practical
workbook. St. Louis, MO: Saunders.
Heidegger, M. (1975). Poetry, language, thought (A. Hofstadter
Trans.). New York, NY: Harper & Row (original work published
1971).
Kerlinger, F. N., & Lee, H. B. (2000). Foundations of behavioral
research (4th ed.). Fort Worth, TX: Harcourt College.
Malasanos, L. J. (1976). What is the preproposal? What are its
component parts? Is it an effective instrument in assessing funding
potential of research ideas? Nursing Research, 25(3), 223–224.
Marshall, C., & Rossman, G. B. (2011). Designing qualitative
research (5th ed.). Los Angeles, CA: Sage.
Martin, C. J. H., & Fleming, V. (2010). A 15-step model for writing
a research proposal. British Journal of Midwifery, 18(12),
791–798.
Melnyk, B. M., & Fineout-Overholt, E. (2011). Evidence-based
practice in nursing & healthcare: A guide to best practice (2nd
ed.). Philadelphia, PA: Lippincott Williams & Wilkins.
Merrill, K. C. (2011). Developing an effective quantitative research
proposal. The Art & Science of Infusion Nursing, 34(3),
181–186.
Miles, M. B., & Huberman, A. M. (1994). Qualitative data analysis:
A source book of new methods (2nd ed.). Beverly Hills, CA: Sage.
Munhall, P. L. (2012). Nursing research: A qualitative perspective
(5th ed.). Sudbury, MA: Jones & Bartlett.
Munro, B. H. (2005). Statistical methods for health care research
(5th ed.). Philadelphia, PA: Lippincott.
National Institute of Nursing Research (NINR). (2012). Online:
Developing nurse scientists. Retrieved from http://www.ninr.nih
.gov/Training/OnlineDevelopingNurseScientists/.
Offredy, M., & Vickers, P. (2010). Developing a healthcare research
proposal: An interactive student guide. Oxford, United Kingdom:
Wiley-Blackwell.
662 UNIT FIVE  Proposing and Seeking Funding for Research
Patton, M. Q. (2002). Qualitative research and evaluation methods
(3rd ed.). Thousand Oaks, CA: Sage.
Pinch, W. J. (1995). Synthesis: Implementing a complex process.
Nurse Educator, 20(1), 34–40.
Ryan-Wenger, N. A. (2010). Evaluation of measurement precision,
accuracy, and error in biophysical data for clinical research and
practice. In C. F. Waltz, O. L. Strickland, & E. R. Lenz (Eds.),
Measurement in nursing and health research (4th ed.) (pp. 371–
383). New York, NY: Springer Publishing Company.
Sandelowski, M., Davis, D. H., & Harris, B. G. (1989). Artful
design: Writing the proposal for research in the naturalist para-
digm. Research in Nursing & Health, 12(2), 77–84.
Santacroce, S. J., Maccarelli, L. M., & Grey, M. (2004). Methods:
Intervention fidelity. Nursing Research, 53(1), 63–66.
Shadish, W. R., Cook, T. D., & Campbell, D. T. (2002). Experimen-
tal and quasi-experimental designs for generalized causal infer-
ence. Chicago, IL: Rand McNally.
Smith, M. J., & Liehr, P. R. (2008). Middle range theory for nursing
(2nd ed.). New York, NY: Springer Publishing Company.
Thompson, S. K. (2002). Sampling (2nd ed.). New York, NY: John
Wiley & Sons.
Turabian, K. L., Booth, W. C., Colomb, G. G., & Williams, J. M.
(2007). A manual for writers of research papers, theses, disserta-
tions, seventh edition: Chicago style for students and research-
ers. Chicago, IL: University of Chicago Press.
University of Chicago Press Staff. (2010). The Chicago manual of
style (16th ed.). Chicago, IL: University of Chicago Press.
Waltz, C. F., Strickland, O. L., & Lenz, E. R. (2010). Measurement
in nursing and health research (4th ed.). New York, NY: Springer
Publishing Company.
   http://evolve.elsevier.com/Grove/practice/
29
Seeking Funding for Research
S
eeking funding for research is important both for
the researcher and for the profession. Well-
designed studies can be expensive. As the rigor
and complexity of the study design increase, the cost
tends to increase. By obtaining funding, the researcher
can conduct a complex, well-designed study. Funding
indicates that others have reviewed the study and rec-
ognize its scientific and social merit. The scientific
credibility of the profession is related to the quality of
studies conducted by its researchers. Thus, scientific
credibility and funding for research are interrelated.
The nursing profession has invested a great deal
of energy in increasing the sources of funding and
amount of money available for nursing research. Each
award of funding enhances the status of the researcher
and increases the possibilities of greater funding for
later studies. In addition, funding provides practical
advantages. For example, funding may reimburse
part or all of the researcher’s salary and release the
researcher from other responsibilities, allowing the
researcher to devote time to conducting the study.
Funding provides you with the resources to hire
research assistants and study coordinators to facilitate
data collection and enhance your productivity. Thus,
skills in seeking funding for research are as important
as skills in the conduct of research.
Building a Program of Research
As a novice researcher, you may have had the dream
of writing a grant proposal to the federal government
or a national foundation for your first study and receiv-
ing a large grant that covers your salary, the salaries
of research assistants and secretarial support, equip-
ment, computers, and payments to subjects for their
time and effort. In reality, this scenario seldom occurs
for an inexperienced researcher. A new researcher is
usually caught in the difficult position of needing
experience to get funded and needing funding to get
CHAPTER
the release time to conduct research and gain the
needed experience. One way of resolving this dilemma
is to design initial studies that can realistically be done
without release time and with little or no funding.
This approach requires a commitment to put in extra
hours of work, which is often unrewarded monetarily
or socially. However, when well carried out and pub-
lished, these types of studies provide the credibility
one needs to begin the process toward major grant
funding. Guidelines for proposals for federal funding
usually include a section of the proposal in which
researchers are expected to describe their own research
studies that serves as evidence of their ability to con-
ceptualize, implement, and complete a study, includ-
ing disseminating the findings. Funders want assurance
that if they fund a proposal, their money will not be
wasted and that the findings of the study will be
published.
An aspiring career researcher needs to initiate a
program of research in a specific area of study and seek
funding in this area. A program of research consists of
the studies that a researcher conducts, starting with
small, simple studies and moving to larger, complex
studies over time. For example, if your research inter-
est is to promote health in rural areas, you need to
plan a series of studies that focus on promoting rural
health. Early studies may be small with each single
successive study building on the findings of the previ-
ous study. Finckeissen (2008) described this approach
as having a meta-model of research with alternative
solutions to a research problem. The findings of each
study suggest new solutions or provide evidence that
another solution is ineffective. Dr. Jean McSweeney,
PhD, RN, FAHA, FAAN, Professor at the College of
Nursing, University of Arkansas of Medical Sciences,
is an example of a nurse researcher who has built a
program of research. Dr. McSweeney’s area of clinical
practice was critical care, and she became very inter-
ested in cardiac patients. To complete her PhD, she
663
664 UNIT FIVE  Proposing and Seeking Funding for Research

TABLE 29-1  Publications Reflecting a Program of Research: Exemplar of McSweeney’s Program of

Research in Cardiovascular Health of Women
Citations from Oldest to Most Recent
McSweeney, J. C. (1993). Explanatory models of a myocardial event: Linkages between perceived causes and modifiable health
behaviors. Rehabilitation Nursing Research, 2(1), 39–49.
McSweeney, J. C. & Crane, P. B. (2001). An act of courage: Women’s decision-making processes regarding outpatient cardiac
rehabilitation attendance. Rehabilitation Nursing, 26(4), 132–140.
Crane, P. B. & McSweeney, J. C. (2003). Exploring older women’s lifestyle changes after myocardial infarction. Medsurg Nursing,
12(3), 170–076.
McSweeney, J. C., Cody, M., O’Sullivan, P., Elberson, D., Moser, D. K., & Gavin, B. J. (2003). Women’s early warning symptoms of
acute myocardial infarction. Circulation, 108(21), 2619–2623.
McSweeney, J. C., O’Sullivan, P., Cody, M., Crane, P. B. (2004). Development of the McSweeney Acute and Prodromal Myocardial
Infarction Symptom Survey. Journal of Cardiovascular Nursing, 19(1), 58–67.
McSweeney, J. C. & Coon, S. (2004). Women’s inhibitors and facilitators associated with making behavioral changes after myocardial
infarction. Medsurg Nursing, 13(1), 49–56.
McSweeney, J. C., Lefler, L. L., & Crowder, B. F. (2005). What’s wrong with me? Women’s coronary heart disease diagnostic
experiences. Progress in Cardiovascular Nursing, 20(2), 48–57.
McSweeney, J. C., Lefler, L. L., Fischer, E. P., Naylor, A. J., & Evans, L. K. (2007). Women’s prehospital delay associated with
myocardial infarction: Does race really matter? The Journal of Cardiovascular Nursing, 22(4), 279–285.
McSweeney, J. C., Pettey, C. M., Fischer, E. P., & Spellman (2009). Going the distance. Research in Gerontological Nursing, 2(4),
256–264.
McSweeney, J. C., Cleves, J. A., Zhao, W., Lefler, L. L., & Yang, S. (2010). Cluster analysis of women’s prodromal and acute
myocardial infarction by race and other characteristics. The Journal of Cardiovascular Nursing, 25(4), 104–110.
McSweeney, J. C., O’Sullivan, P., Cleves, M. A., Lefler, L. L., Cody, M., et al. (2010). Racial differences in women’s prodromal and
acute symptoms of myocardial infarction. American Journal of Critical Care, 19(1), 63–73.
Beck, C., McSweeney, J. C., Richards, K. C., Roberson, P. K., Tsai, P.-F., & Souder, E. (2010). Challenges in tailored intervention
research. Nursing Outlook, 58(2), 104–110.
McSweeney, J. C., Pettey, C. M., Souder, E., & Rhoads, S. (2011). Disparities in women’s cardiovascular health. Journal of Obstetric,
Gynecologic, and Neonatal Nursing, 40(3), 362–371.

conducted a qualitative study with patients and their

significant others to explore behavior changes after a
myocardial infarction. Her first post-dissertation study Contribution
was a qualitative study of women’s motivations to
change their behavior after a myocardial infarction.
She continued by conducting a series of quantitative
studies that built on the findings of the previous studies. Capacity Capital
Table 29-1 lists publications by Dr. McSweeney that
indicate the trajectory of her research program. Publi-
cation of the studies increased the credibility of the
Indicates the overlap of contribution,
researcher and provided the foundation for future
capacity, and capital which constitutes
funding. an ideal focus for a program of research
How do you decide on the focus of your program
of research? The ideal focus of a program of research Figure 29-1  Ideal focus for a program of research.
is the intersection of a potential contribution to science,
your capacity, and the capital that you can assemble.
Figure 29-1 shows the ideal study as overlapping
circles of contribution, capacity, and capital. reviewing the literature and finding a significant gap
in knowledge. Maybe this knowledge is needed to
Contribution solve a health problem, or it is the evidence needed
Contribution refers to the gap in knowledge that your to advance nursing science. The research focus is
research will address. This gap will be identified by broader than a single study. For example, a program

of research could be developed related to adherence
to antihypertensive medications. The first study might
be a qualitative study of the reasons patients give for
not taking their antihypertensive medications. Build-
ing on that knowledge, the next study might be a
descriptive quantitative study of differences in atti-
tudes of men and women toward antihypertensive
medications. A later study might measure the out-
comes of antihypertensive patients who received an
adherence intervention compared with a group of
patients who did not receive an intervention. Another
consideration related to the research focus is your
capacity to study the problem.
Capacity
Capacity consists of the internal resources you
possess, such as your intellect, emotional maturity,
knowledge, and skills. Because a program of research
develops over several years, perseverance and com-
mitment are necessary elements of your capacity.
Which areas of nursing and health stimulate your
curiosity and sustain your interest? Think about the
topics or areas of nursing practice in which you are
the most interested. Which patients or clinical areas
stimulate your curiosity? Maybe you have a personal
connection to a particular area, such as a nurse
researcher who is studying autism because she is the
mother of a son with autism. Maybe a family member
died at a young age from undiagnosed cardiac disease,
and now you are passionate to understand the
decision-making process related to diagnosis of
cardiac disease. Your research focus may evolve over
time, but, ideally, your passion for a specific topic or
group of patients would provide the basis for a long
research career.
Knowledge related to your research area is another
element of your capacity related to a specific research
topic. This knowledge may come from educational
programs, personal study, and clinical experience. If
you are interested in genomic research, what is your
knowledge of genes and interactions among them
and the environment? Have you had a course in
genetics or learned the laboratory skills to gather and
analyze cellular level data? If you are interested in
the effects of positioning on the hemodynamics of
unstable, acute patients, have you ever worked in a
critical care unit? One aspect of building a research
career is to continue to expand your capacity in your
focus area, but in the beginning, selecting an area in
which you have baseline knowledge is helpful. Your
capacity also includes grantsmanship, the knowl-
edge and skill you have related to securing and man-
aging grants.
CHAPTER 29  Seeking Funding for Research 665
Capital
People and funding are the capital you need to conduct
research over time. The primary purpose of this
chapter is to describe how to increase your monetary
capital. When a potential contribution to science, your
capacity, and available capital overlap, you have found
an ideal focus for your research career. This chapter
also provides suggestions on how to develop your
grantsmanship and increase your capacity in the area
in which you believe you can make a meaningful
contribution.
Getting Started
Level of Commitment
Developing your capacity as a researcher in a specific
area takes a great deal of time and effort. Writing
proposals for funding is hard work. Before you begin,
reflect on whether your motivation is external or inter-
nal. If your motivation is external, you are committed
to seeking funding because of the potential to receive
rewards from your employer, to earn the high regard
of your peers, or to be eligible for a promotion. If your
motivation is internal, you are convinced that more
knowledge is needed to benefit your patients. Both
external and internal motivations are valid reasons to
be committed to a program of research; however, an
internally motivated researcher will conduct studies
with limited funding and continue to seek additional
funding even in the absence of external rewards. As
an element of capacity, your level of commitment will
determine your ability to persevere and develop a
program of research.
Support of Other People
Even the most internally motivated person may
experience times of discouragement and need the
support of peers. Peers who share common values,
ways of thinking, and activities can be a reference
group for a novice researcher. This occurs when you
identify with the group, take on group values and
behavior, and evaluate your own values and behavior
in relation to those of the group. A new researcher
moving into grantsmanship may need to switch from
a reference group that views research and grant
writing to be either over their heads or not worth
their time to a group that values this activity. From
this group, you may receive support and feedback
necessary to develop grant-writing skills and enact a
program of research. In addition, you will have the
opportunity to provide similar support and feedback
to your peers.
666 UNIT FIVE  Proposing and Seeking Funding for Research
Networking is a process of developing channels of
communication among people with common interests
who may not work for the same employer or may be
geographically scattered. Contacts may be made
through social media, computer networks, mail, tele-
phone, or arrangements to meet at a conference (Adeg-
bola, 2011). Strong networks are based on reciprocal
relationships. A professional network can provide
opportunities for brainstorming, sharing ideas and
problems, and discussing grant-writing opportunities.
In some cases, networking may lead to the members
of a professional network writing a grant that will be
a multisite study with data collected in each member’s
home institution. When a proposal is being developed,
the network, which may become a reference group, can
provide feedback at various stages of development of
the proposal. Adegbola (2011) provides practical tips
on how to develop and maintain a professional network.
Through networking, nurses interested in a particu-
lar area of study can find peers, content experts,
and mentors. A content expert may be a clinician or
researcher who is known for his or her work in the
area in which you are also interested. Through your
review of the literature, you identify a researcher who
has developed an instrument to measure a variable that
you have decided must be included in your proposed
study. For example, you want to measure a biological
marker of stress and you have read several studies in
which an experienced researcher measured the vari-
able using a specific piece of equipment. You can also
search the Virginia Henderson International Nursing
Library to find funded and unfunded researchers
on different topics (http://www.nursinglibrary.org/vhl/
pages/aboutus.html). Contact the researcher through
email to make a telephone appointment to discuss the
strengths and weaknesses of this particular measure-
ment, or you may arrange to meet at an upcoming
conference.
A mentor is a person who is more experienced
professionally and willing to work with a less experi-
enced professional to achieve his or her goals. Because
funded nursing researchers are few, the need for men-
toring is greater than the number of available mentors
(Maas, Conn, Buckwalter, Herr, & Tripp-Reimer,
2009). Finding a mentor may take time and require
significant effort. Much of the information needed is
transmitted verbally, requires actual participation in
grant-writing activities, and is best learned in a mentor
relationship. This type of relationship requires a will-
ingness by both professionals to invest time and
energy. A mentor relationship has characteristics of
both a teacher-learner relationship and a close friend-
ship. Each individual must have an affinity for the
other, from which a close working relationship can be
developed. The relationship usually continues for a
long period.
Grantsmanship
Grantsmanship is not an innate skill; it must be learned.
Learning grant-related skills requires a commitment of
both time and energy. However, the rewards can be
great. Strategies used to learn grantsmanship are
described in the following sections and are listed in
order of increasing time commitment, involvement,
and level of expertise needed. These strategies are
attending grantsmanship courses, working with experi-
enced researchers, joining research organizations, and
participating on research committees or review panels.
Attending Courses and Workshops
Some universities offer elective courses on grantsman-
ship. Continuing education programs or professional
conferences sometimes offer topics related to grants-
manship. The content of these sessions may include the
process of grant writing, techniques for obtaining grant
funds, and sources of grant funds. In some cases, rep-
resentatives of funding agencies are invited to explain
funding procedures. This information is useful for
developing skill in writing proposals. Not all courses
or educational opportunities for learning grantsman-
ship require attendance at a conference because some
seminars are offered as webinars or online courses.
Apprenticeship
Volunteering to assist with the activities of another
researcher is an excellent way to learn research and
grantsmanship. Graduate students can gain this expe-
rience by becoming graduate research assistants.
Through an apprenticeship, you may gain experience
in writing grants and reading proposals that have been
funded. Examining proposals that have been rejected
can be useful if the comments of the review committee
are available. The criticisms of the review committee
point out the weaknesses of the study and clarify the
reasons why the proposal was rejected. Examining
these comments on the proposal can increase your
insight as a new grant writer and prepare you for
similar experiences. However, some researchers are
sensitive about these criticisms and may be reluctant
to share them. If an experienced researcher is willing,
it is enlightening to hear his or her perceptions and
opinions about the criticisms. Ideally, by working
closely with an experienced researcher, you have the
opportunity to demonstrate your commitment, and the
researcher invites you to become a permanent member
of a research team.

TABLE 29-2  Regional Nursing Research Organizations
Region Website
Eastern Nursing Research Society http://www.enrs-go.org
Southern Nursing Research Society http://www.snrs.org
Midwest Nursing Research Society http://www.mnrs.org
Western Institute of Nursing http://www.winursing.org/
Regional Nursing Research Organizations
In the United States, nurse researchers in each region
have formed regional research organizations. Table
29-2 lists these organizations and their websites. Each
of these regional organizations has an annual confer-
ence and provides opportunities for nursing students
to display a poster or present a small study of their
early research projects. These conferences are an
excellent opportunity to network and meet more expe-
rienced researchers (Adegbola, 2011).
Serving on Research Committees
Research committees and institutional review boards
exist in many healthcare and professional organiza-
tions. Through membership on these committees, con-
tacts with researchers can be made. Also, many
research committees are involved in reviewing pro-
posals for the funding of small grants or granting
approval to collect data in an institution. Often reading
proposals for approval for research involving human
subjects or for funding can give the novice researcher
insight into the importance of clarity and organization
in the research proposal. Reviewing proposals and
making decisions about funding help researchers
become better able to critique and revise their own
proposals before submitting them for review.
Identifying Funding Sources
Types of Grants
Two main types of grants are sought in nursing: project
grants and research grants. Project grant proposals
are written to obtain funding for the development of
new educational programs in nursing, such as a
program designed to teach nurses to provide a new
type of nursing care or as a project to support nursing
students seeking advanced degrees. These grants may
fund a project manager to achieve the goals of the
grant. Although these programs may involve evalua-
tion, they seldom involve research. For example, the
effectiveness of a new approach to patient care may
be evaluated, but the findings can seldom be general-
ized beyond the unit or institution in which the patient
CHAPTER 29  Seeking Funding for Research 667
Email
info@enrs-go.org
info@snrs.org
info@mnrs.org
win@ohsu.edu
care was provided. The emphasis is on implementing
the project, not on conducting research. Research
grants provide funding to conduct a study. Although
the two types of grant proposals have similarities, they
have important differences in writing techniques, flow
of ideas, and content. This chapter focuses on seeking
funding for research. Within research grants, propos-
als vary depending on the source of funding. Proposals
for federal funding are the most complex and include
a significant amount of information about your institu-
tion’s resources and capacity to support the study. The
section on Government Funding provides additional
information on types of federal proposals.
Nongovernment Funding
Private or Local Funding
The next step is to determine potential sources for
small amounts of research money. In some cases, man-
agement in the employing institution can supply
limited funding for research activities if a good case
is presented for the usefulness of the study to the
institution. In many universities, funds are available
for intramural grants, which you can obtain competi-
tively by submitting a brief proposal to a university
committee. Local chapters of nursing organizations
have money available for research activities. Sigma
Theta Tau International, the honor society for nurses,
provides small grants for nursing research that can
be obtained through submission to local, regional,
national, or international review committees. Another
source are organizations, such as the local chapters of
the American Cancer Society and the American Heart
Association. Although grants from the national offices
of these organizations require sophisticated research,
local or state levels of the organization may have small
amounts of funds available for studies in the organiza-
tion’s area of interest.
Private individuals who are locally active in phi-
lanthropy may be willing to provide financial assis-
tance for a small study in an area appealing to them.
You need to know who to approach and how and when
to make the approach to increase the probability of
successful funding. Sometimes this approach requires
668 UNIT FIVE  Proposing and Seeking Funding for Research

knowing someone who knows someone who might be TABLE 29-3  National Specialty Nursing
willing to provide financial support. Acquiring funds
Organizations That Fund Research
from private individuals requires more assertiveness
than other approaches to funding. Organization or
Requests for funding need not be limited to a single Association Website
source. If you need a larger amount of money than one Academy of Medical http://www.amsn.org/
source can supply, seek funds from one source for a Surgical Nurses
specific research need and from another source for American Association of http://www.aacn.org/
Critical Care Nurses
another research need. Also, one source may be able
to provide funds for a small segment of time; you can Association of Women’s http://www.awhonn.org./
Health, Obstetric, and awhonn/
then approach another source to seek funding for
Neonatal Nurses
another phase of the study. You can also combine these
Emergency Nurses http://www.ena.org/Pages/
two strategies.
Association default.aspx
Seeking funding from local sources is less demand-
Hospice and Palliative Care http://www.hpna.org/
ing in terms of formality and length of the proposal
Nurses
than is the case with other types of grants. Often, the
National Association of http://www.orthonurse.org/
process is informal and may require only a 2- or 3-page Orthopaedic Nurses
description of the study. Provide a clear, straightfor-
National Gerontological http://www.ngna.org/
ward description of the study and how the findings Nursing Association
will contribute to practice or future studies. The impor-
Oncology Nursing Society http://www.ons.org/
tant thing is to know what funds are available and
Society of Pediatric Nurses http://www.pedsnurses.org/
how to apply for them. Some of these funds go unused
index.html
each year because nurses are unaware of the existence
Wound Ostomy and http://www.wocn.org/
of the funds or think that they are unlikely to be suc-
Continence Nurses Society
cessful in obtaining the money. This unused money
leads granting agencies or potential donors to conclude
that nurses do not need more money for research.
Small grants do more than just provide the funds These grants are usually for less than $7500 each year,
necessary to conduct the research. They are the first are very competitive, and are awarded to new investi-
step in your being recognized as a credible researcher gators with promising ideas. Receiving funding from
and in being considered for more substantial grants for these organizations is held in high regard. Information
later studies. When you receive a grant, no matter how regarding these grants is available from the American
small, include this information on your curriculum Nurses Foundation (http://www.anfonline.org/) and
vitae. Also, list your participation in funded studies, Sigma Theta Tau International (http://www.nursing
even if you were not the principal investigator (PI). society.org/default.aspx).
These entries are evidence of first-level recognition as
a researcher. Industry
Industry may be a good source of funding for nursing
National Nursing Organizations studies, particularly if one of their company’s products
Many specialty nursing organizations provide support is involved in the study. For example, if a particular
for studies relevant to their specialty, including nurse type of equipment is being used during an experimen-
practitioner groups. These organizations often provide tal treatment, the company that developed the equip-
guidance to budding new researchers who need assis- ment may be willing to provide equipment for the
tance in beginning the process of planning and seeking study or may be willing to fund the study. If a compari-
funding for research. To determine the resources pro- son study examining outcomes of one type of dressing
vided by a particular nursing organization, search the versus another is to be conducted, the company that
organization’s website or contact the organization by produces one of the products might provide the product
email, letter, or phone. Table 29-3 provides informa- or fund the study. Industry-supported research is being
tion on 10 large specialty nursing organizations that scrutinized because of publicized incidents in which
provide grant funding. possible conflicts of interest resulted in harm to
Two national nursing organizations that provide a subject or may have prevented the publication of
small grants not linked to a specialty are the American unfavorable findings (Fry-Revere & Malmstrom,
Nurses Foundation and Sigma Theta Tau International. 2009). The ethics of seeking such funding should be

carefully con­sidered because there is sometimes a risk
that the researcher might not be unbiased in interpret-
ing study results.
Foundations
Many foundations in the United States provide funding
for research, but the problem is to determine which
foundations have interests in a particular field of study.
The board of a foundation may evaluate the founda-
tion’s priorities each year, resulting in different priori-
ties each year. You must learn the characteristics of
the foundation, such as what it will fund. A foundation
may fund only studies by female researchers, or it may
be interested only in studies of low-income groups. A
foundation may fund only studies being conducted in
a specific geographical region. The average amount of
money awarded for a single grant and the ranges of
awards need to be determined for each foundation. If
the average award of a particular foundation is $2500
and if $30,000 is needed, that foundation is not the
most desirable source of funds. Identify foundations
that match your research topic, geographical location,
and funding needs. Review carefully the foundation’s
guidelines for submitting funding requests. Making a
personal visit to the foundation or contacting the staff
person responsible for funding is desirable in some
cases. You can increase your likelihood of funding by
revising your proposal to align with the foundation’s
priorities.
Several publications list foundations and their
interests. If you are affiliated with a university that is
a subscriber, a computerized information system, the
Sponsored Programs Information Network, can assist
you in locating the most appropriate funding sources
to support your research interests. The database con-
tains approximately 2000 programs that provide infor-
mation on federal agencies, private foundations, and
corporate foundations. Another useful resource is the
Investigator- Agency-
Initiated Initiated
Research Research
Program Request for
Request for
Announcement Applications
Proposals (RFP)
(PA) (RFA)
Reminder of Narrow focus, Well-defined area
Agency specific desired with objectives to
Priorities outcome be met
CHAPTER 29  Seeking Funding for Research 669
Foundation Directory, which is available online to
subscribers at http://fconline.fdncenter.org. Check
with your hospital or university administrators to find
out if you have access to this resource.
Government Funding
The largest source of grant monies is the federal
government—so much so that the federal government
influences what is studied and what is not. Information
on funding agencies can be obtained from a document
compiled by the federal government, The Catalog
of Federal Domestic Assistance, which is available
online at www.cfda.gov. The National Institutes of
Health (NIH), particularly the National Institute for
Nursing Research, and Agency for Healthcare Research
and Quality are interested in receiving nursing propos-
als. Each agency has areas of focus and priorities for
funding that change over time. It is important to know
this information and prepare proposals within these
areas to obtain funding. This information is available
online at www.grants.gov, a searchable listing of all
federal research funding opportunities.
Two approaches can be used to seek federal funding
for research. As the researcher, you can identify a
significant problem, develop a study to examine it, and
submit a proposal for the study to the appropriate
federal funding agency. This type of proposal is called
an investigator-initiated research proposal. Alterna-
tively, an agency within the federal government can
identify a significant problem, develop a plan by
which the problem can be studied, and publish a
request for proposals (RFP) or a request for appli-
cations (RFA) from researchers (Figure 29-2).
Investigator-Initiated Proposals
If the study is initiated by the investigator, it is useful
for the researcher to contact an official within the
government agency early in the planning process to
Figure 29-2  Types of federal research proposals.
670 UNIT FIVE  Proposing and Seeking Funding for Research
inform the agency of the intent to submit a proposal.
Each agency has established dates, usually three times
a year, when proposals are reviewed. You will need to
start preparing your proposal months ahead of this
deadline, and some agencies are willing to provide
assistance and feedback to the researcher during
development of the proposal. This assistance may
occur through email, telephone conversations, or feed-
back on a draft of the proposal. Proposals submitted
in response to a program announcement (PA) are
considered investigator-initiated proposals. An agency
or group of agencies may release a PA to remind
researchers of priority areas and generate interest in a
priority area.
Requests for Proposals and Applications
The NIH issues an RFP when scientists advising the
institutes have identified a specific need to move an
area of knowledge forward. An RFA may be broader
than an RFP but will still have a focus and a list of
objectives that an institute or center within the NIH
has identified. An RFA will have a single application
deadline. The amount that has been budgeted for the
successful applications is indicated, and the RFA is
usually open for several funding cycles.
Submitting a Proposal
for a Federal Grant
Ensuring a Unique Proposal
During your review of the literature, you may have
read the findings of funded studies, but the literature
does not include recently completed or ongoing funded
studies. Early in the process of planning a study for
which you intend to seek federal funding, it is wise to
determine what studies on your topic of interest have
been funded previously and what funded studies are
currently in process. This information is available on
the NIH Research Portfolio Online Reporting Tools—
Expenditures and Results (RePORTER), which is
maintained by the Office of Extramural Research at
the NIH (http://projectreporter.nih.gov/reporter.cfm).
The institutes and agencies that fund studies and proj-
ects and are included in the RePORTER are listed in
Table 29-4. You can search the database by state,
subject, type of grant, funding agency, or investigator.
Reviewing proposals that are funded by a particular
agency can be helpful. Although the agency cannot
provide these proposals, researchers can sometimes
obtain them by contacting the PI of the study person-
ally. In some cases, the researcher may travel to
TABLE 29-4  Funding Agencies Included in
the National Institutes of Health
Research Portfolio Online
Reporting Tools—Expenditures  
and Results (RePORTER)
Agency for Health Care Research and Quality (AHRQ)
Centers for Disease Control and Prevention (CDC)
Food and Drug Administration (FDA)
Health Resources and Services Administration (HRSA)
National Institutes of Health (NIH)
Substance Abuse and Mental Health Services (SAMHSA)
Veterans Affairs
Washington to meet with an agency representative.
This type of contact allows the researcher to modify
the proposal to fit more closely within agency guide-
lines, increasing the probability of funding. In many
cases, proposals will fit within the interests of more
than one government agency at the time of submis-
sion. It is permissible and perhaps desirable to request
that the proposal be assigned to two agencies for
review and potential funding.
Verifying Institutional Support
Grant awards are most commonly made to institutions
rather than to individuals. It is important to determine
the willingness of the institution to receive the grant
and support the study. This willingness needs to be
documented in the proposal. Supporting the study
involves appropriateness of the study topic; adequacy
of facilities and services; availability of space needed
for the study; contributions that the institution is
willing to make to the study, such as staff time, equip-
ment, or data processing; and provision for overseeing
the rights of human subjects. The study’s budget will
include a category called indirect costs to pay the
institution’s expenses. For federal grants, indirect
costs may by as high as 50% of the direct costs, the
funds necessary to conduct the study. Direct costs are
used to pay a portion of the researcher’s salary, the
salaries of data collectors or other research assistants,
obtain equipment for the study, and provide a small
payment to study participants to acknowledge their
time and effort.
Making Time to Write
Set aside sufficient time to develop your proposal
carefully, beginning with a thorough literature review.
See Chapter 28 for how to write a proposal. Read the
funding agency’s guidelines carefully and completely
before starting to write. Keep the guidelines nearby as
you write so that you can easily refer back to them.

Strictly adhere to the page limitations and required
font sizes. The sections of the proposal may be
uploaded separately. Be sure that all the sections agree
with each other on details such as names of instru-
ments and inclusion criteria for subjects.
Writing your first proposal on a tight deadline is
unwise. Proposals require refining the idea and method
and rewriting the text several times. Plan on 6 to 12
months for proposal development from the point of
early development of your research ideas. As soon as
you have a complete draft, ask a peer or mentor to read
the proposal to check for errors in logic. As people
review your proposal informally, recognize their
questions as indications that an idea was not clearly
presented and may need to be rewritten. Before sub-
mission, it is highly recommended that you have a
content expert or other researcher who is not at your
institution critique the proposal (Office of Manage-
ment and Budget, 2011).
Understanding the Review Process
The Center for Scientific Review has the administra-
tive responsibility for ensuring a fair, equitable review
of all proposals submitted to NIH or other Public
Health Services agencies. After submission, the staff
person assigned to your grant will determine which
integrated review group will review your proposal for
its technical and scientific merit. Within the integrated
review group, each grant is assigned to a study section
for scientific evaluation. The study section comprises
active funded researchers. The study sections have no
alignment with the funding agency. Thus, staff persons
in the agencies have no influence on the committee’s
work of judging the scientific merit of the proposal.
The proposal is given to two or more reviewers who
are considered qualified to evaluate the proposal and
have no conflicts of interest. The reviewers rate the
proposal on the core criteria and overall impact and
submit a written critique of the study. Each member
may have 50 to 100 proposals to read in a 1- to
2-month period. A meeting of the full study section is
then held. The persons who critiqued the proposal
discuss each application, and other members comment
or ask questions before recording their scores.
Proposals are assigned a numerical score used to
develop a priority rating for funding. A study that is
scored is not necessarily funded. The PI may review
the progress of the proposal through the stages of
review by accessing an online system, called the Elec-
tronic Research Administration (eRA) Commons.
Funding begins with the proposal that has the highest
rank order and continues until available funds are
depleted. This process can take 6 months or longer.
CHAPTER 29  Seeking Funding for Research 671
Because of this process, researchers may not receive
grant money for up to 1 year after submitting the
proposal.
Often, researcher-initiated proposals are rejected
(or scored but not funded) after the first submission.
The critique of the scientific committee, called a
summary statement, is available to the researcher via
his or her eRA Commons account. Frequently, the
agency staff encourages the researcher to rewrite the
proposal with guidance from the comments and resub-
mit it to the same agency. The probability of funding
is often greater the second time if the researcher has
followed the suggestions.
Responding to Rejected Grant Proposals
The researcher’s reaction to a rejected proposal is
usually anger and then depression. The frustrated
researcher may abandon the proposal, stuff it in a
bottom drawer somewhere, and forget it. There seems
to be no way to avoid the anger and depression after
a rejection because of the significant emotion and time
invested in writing it. However, after a few weeks it
is advisable to examine the rejection letter again. The
comments can be useful in rewriting the proposal for
resubmission. The learning experience of rewriting the
proposal and evaluating the comments will provide a
background for seeking funding for another study.
A skilled grant writer will have approximately one
proposal funded for every five submitted. The average
is far less than this. Thus, the researcher needs to be
committed to submitting proposals repeatedly to
achieve grant funding.
Grant Management
Receiving notice that a grant proposal is funded is one
of the highlights in a researcher’s career and warrants
a celebration. However, begin work on the study as
soon as possible. You included a detailed plan of activ-
ities in the proposal that is ready to be implemented.
To avoid problems, you need to consider managing the
budget, hiring and training research personnel, main-
taining the promised timetable, and coordinating
activities of the study.
Managing the Budget
Although the supporting institution is ultimately
responsible for dispensing and controlling grant
monies, the PI is responsible for monitoring budget
expenditures and making decisions about how the
money is to be spent (Devine, 2009). If this grant is
the first one received, a PI who has no previous admin-
istrative experience may need guidance in how to keep
672 UNIT FIVE  Proposing and Seeking Funding for Research
records and make reasonable budget decisions. If
funding is through a federal agency, the PI will be
required to provide interim reports as well as updates
on the progress of the study.
Training Research Personnel
When a new grant is initiated, set aside time to inter-
view, hire, and train grant personnel (Martin &
Fleming, 2010). The personnel who will be involved
in data collection need to learn the process, and then
data collection needs to be refined to ensure that each
data collector is consistent with the other data collec-
tors. This process helps evaluate interrater reliability.
The PI needs to set aside time to oversee the work of
personnel hired for the grant.
Maintaining the Study Schedule
The timetable submitted with the proposal needs to be
adhered to whenever possible, which requires careful
planning. Otherwise, other work activities are likely
to take precedence and delay the grant work. Unex-
pected events do happen. However, careful planning
can minimize their impact. The PI needs to refer back
to the timetable constantly to evaluate progress. If the
project falls behind schedule, action needs to be taken
to return to the original schedule or to readjust the
timetable.
Coordinating Activities
During a large study with several investigators and
other grant personnel, coordinating activities can be a
problem. Arrange meetings of all grant workers at
intervals to share ideas and solve problems. Keep
records of the discussions at these meetings. These
actions can lead to a more smoothly functioning team.
Submitting Reports
Federal grants require the submission of interim
reports according to preset deadlines. The notice of a
grant award sent as a PDF (Portable Document
Format) document via email will include guidelines
for the content of the reports, which will consist of a
description of grant activities. Set aside time to prepare
the report, which usually requires compiling figures
and tables. In addition to the written reports, it is often
useful to maintain contact with the appropriate staff at
the federal agency.
Planning Your Next Grant
The researcher should not wait until funding from the
first grant has ended to begin seeking funds for a
second study because of the length of time required to
obtain funding. It may be wise to have several ongoing
studies in various stages of implementation. For
example, you could be planning a study, collecting
data on a second study, analyzing data on a third study,
and writing papers for publication on a fourth study.
A full-time researcher could have completed one
funded study, be in the last year of funding for a
second study, be in the first year of funding for a third
study, and be seeking funding for a fourth study. This
scenario may sound unrealistic, but with planning, it
is not. This strategy not only provides continuous
funding for research activities but also facilitates a
rhythm of research that prevents time pressures and
makes use of lulls in activity in a particular study. To
increase the ease of obtaining funding, the studies
need to be within the same area of research, each
building on previous studies.
KEY POINTS
• Building a program of research requires conducting
a series of studies on a topic, with each study build-
ing on the findings of the previous one.
• The ideal topic around which to build a research
program can be identified by considering topics for
which the researcher has or can gain the expertise
to conduct studies (capacity), funding is available
(capital), and the potential exists for the researcher
to make a difference (contribution).
• Writing a grant proposal for funding requires a
commitment to working extra hours.
• To receive funding, researchers need to learn
grantsmanship skills.
• The first studies are usually conducted with per-
sonal funding or small grants.
• Nongovernmental sources of funding include
private donors, local organizations, nursing organi-
zations, and foundations.
• Before submitting a proposal to seek federal
funding, the researcher should successfully com-
plete two or more small studies and disseminate the
findings.
• The researcher identifies a significant problem,
develops a study to examine it, and submits a pro-
posal for the study to an appropriate federal funding
agency.
• The PI is responsible for keeping within the budget,
training research personnel, maintaining the sched-
ule, and coordinating activities.
• Grants require the submission of interim and
final reports of expenditures, activities, and
achievements.

• A researcher should not wait until funding from
one grant ends before seeking funds for the next
grant.
REFERENCES
Adegbola, M. (2011). Soar like geese: Building developmental
network relationships for scholarship. Nursing Education Per-
spectives, 32(1), 51–53.
Devine, E. B. (2009). The art of obtaining grants. American Journal
of Health-System Pharmacy, 66(6), 580–587.
Finkeissen, E. (2008). Steps from erratic projects toward structured
programs of research. Foundations of Science, 13(2), 143–148.
Fry-Revere, S. & Malmstrom, D. B. (2009). More regulation
of industry-supported biomedical research: Are we asking the
CHAPTER 29  Seeking Funding for Research 673
right questions? Journal of Law, Medicine, and Ethics, 29(3),
420–430.
Martin, C. J. H. & Fleming, V. (2010). A 15-step model for writing
a research proposal. British Journal of Midwifery, 18(12),
791–798.
Maas, M. L., Conn, V., Buckwalter, K. C., Herr, K., & Tripp-
Reimer, T. (2009). Increasing nurse faculty research: The Iowa
Gerontological Nurse Research and Regional Research Consor-
tium Strategies. Journal of Nursing Scholarship, 41(4),
411–419.
Office of Management and Budget (2011). Appendix VI: Develop-
ing and writing grant proposals (pp. 2700–2702). In: Catalog of
Federal Domestic Assistance. Retrieved from https://www.cfda
.gov/downloads/CFDA_2011.pdf.
 Appendix A
Z Values Table

z Score From Mean to z (%) z Score From Mean to z (%)

.00 .00 .46 17.72
.01 .40 .47 18.08
.02 .80 .48 18.44
.03 1.20 .49 18.79
.04 1.60 .50 19.15
.05 1.99 .51 19.50
.06 2.39 .52 19.85
.07 2.79 .53 20.19
.08 3.19 .54 20.54
.09 3.59 .55 20.88
.10 3.98 .56 21.23
.11 4.38 .57 21.57
.12 4.78 .58 21.90
.13 5.17 .59 22.24
.14 5.57 .60 22.57
.15 5.96 .61 22.91
.16 6.36 .62 23.24
.17 6.75 .63 23.57
.18 7.14 .64 23.89
.19 7.53 .65 24.22
.20 7.93 .66 24.54
.21 8.32 .67 24.86
.22 8.71 .68 25.17
.23 9.10 .69 25.49
.24 9.48 .70 25.80
.25 9.87 .71 26.11
.26 10.26 .72 26.42
.27 10.64 .73 26.73
.28 11.03 .74 27.04
.29 11.41 .75 27.34
.30 11.79 .76 27.64
.31 12.17 .77 27.94
.32 12.55 .78 28.23
.33 12.93 .79 28.52
.34 13.31 .80 28.81
.35 13.68 .81 29.10
.36 14.06 .82 29.39
.37 14.43 .83 29.67
.38 14.80 .84 29.95
.39 15.17 .85 30.23
.40 15.54 .86 30.51
.41 15.91 .87 30.78
.42 16.28 .88 31.06
.43 16.64 .89 31.33
.44 17.00 .90 31.59
.45 17.36 .91 31.86

674
 APPENDIX A  Z Values Table 675

z Score From Mean to z (%) z Score From Mean to z (%)

.92 32.12 1.48 43.06
.93 32.38 1.49 43.19
.94 32.64 1.50 43.32
.95 32.89 1.51 43.45
.96 33.15 1.52 43.57
.97 33.40 1.53 43.70
.98 33.65 1.54 43.82
.99 33.89 1.55 43.94
1.00 34.13 1.56 44.06
1.01 34.38 1.57 44.18
1.02 34.61 1.58 44.29
1.03 34.85 1.59 44.41
1.04 35.08 1.60 44.52
1.05 35.31 1.61 44.63
1.06 35.54 1.62 44.74
1.07 35.77 1.63 44.84
1.08 35.99 1.64 44.95
1.09 36.21 1.65 45.05
1.10 36.43 1.66 45.15
1.11 36.65 1.67 45.25
1.12 36.86 1.68 45.35
1.13 37.08 1.69 45.45
1.14 37.29 1.70 45.54
1.15 37.49 1.71 45.64
1.16 37.70 1.72 45.73
1.17 37.90 1.73 45.82
1.18 38.10 1.74 45.91
1.19 38.30 1.75 45.99
1.20 38.49 1.76 46.08
1.21 38.69 1.77 46.16
1.22 38.88 1.78 46.25
1.23 39.07 1.79 46.33
1.24 39.25 1.80 46.41
1.25 39.44 1.81 46.49
1.26 39.62 1.82 46.56
1.27 39.80 1.83 46.64
1.28 39.97 1.84 46.71
1.29 40.15 1.85 46.78
1.30 40.32 1.86 46.86
1.31 40.49 1.87 46.93
1.32 40.66 1.88 46.99
1.33 40.82 1.89 47.06
1.34 40.99 1.90 47.13
1.35 41.15 1.91 47.19
1.36 41.31 1.92 47.26
1.37 41.47 1.93 47.32
1.38 41.62 1.94 47.38
1.39 41.77 1.95 47.44
1.40 41.92 1.96 47.50
1.41 42.07 1.97 47.56
1.42 42.22 1.98 47.61
1.43 42.36 1.99 47.67
1.44 42.51 2.00 47.72
1.45 42.65 2.01 47.78
1.46 42.79 2.02 47.83
1.47 42.92 2.03 47.88
Continued
 676 APPENDIX A  Z Values Table

z Score From Mean to z (%) z Score From Mean to z (%)

2.04 47.93 2.53 49.43
2.05 47.98 2.54 49.45
2.06 48.03 2.55 49.46
2.07 48.08 2.56 49.48
2.08 48.12 2.57 49.49
2.09 48.17 2.58 49.51
2.10 48.21 2.59 49.52
2.11 48.26 2.60 49.53
2.12 48.30 2.61 49.55
2.13 48.34 2.62 49.56
2.14 48.38 2.63 49.57
2.15 48.42 2.64 49.59
2.16 48.46 2.65 49.60
2.17 48.50 2.66 49.61
2.18 48.54 2.67 49.62
2.19 48.57 2.68 49.63
2.20 48.61 2.69 49.64
2.21 48.64 2.70 49.65
2.22 48.68 2.71 49.66
2.23 48.71 2.72 49.67
2.24 48.75 2.73 49.68
2.25 48.78 2.74 49.69
2.26 48.81 2.75 49.702
2.27 48.84 2.76 49.711
2.28 48.87 2.77 49.720
2.29 48.90 2.78 49.728
2.30 48.93 2.79 49.736
2.31 48.96 2.80 49.744
2.32 48.98 2.81 49.752
2.33 49.01 2.82 49.760
2.34 49.04 2.83 49.767
2.35 49.06 2.84 49.774
2.36 49.09 2.85 49.781
2.37 49.11 2.86 49.788
2.38 49.13 2.87 49.795
2.39 49.16 2.88 49.801
2.40 49.18 2.89 49.807
2.41 49.20 2.90 49.813
2.42 49.22 2.91 49.819
2.43 49.25 2.92 49.825
oasis-ebl|Rsalles|1476813894

2.44 49.27 2.93 49.831

2.45 49.29 2.94 49.836
2.46 49.31 2.95 49.841
2.47 49.32 2.96 49.846
2.48 49.34 2.97 49.851
2.49 49.36 2.98 49.856
2.50 49.38 2.99 49.861
2.51 49.40 3.00 49.865
2.52 49.41
 Appendix B
Critical Values for Student’s t Distribution

Level of Significance (α), One-Tailed Test

.001 .005 .01 .025 .05 .10

Level of Significance (α), Two-Tailed Test

df .002 .01 .02 .05 .10 .20
2 22.327 9.925 6.965 4.303 2.920 1.886
3 10.215 5.841 4.541 3.182 2.353 1.638
4 7.173 4.604 3.747 2.776 2.132 1.533
5 5.893 4.032 3.365 2.571 2.015 1.476
6 5.208 3.707 3.143 2.447 1.943 1.440
7 4.785 3.499 2.998 2.365 1.895 1.415
8 4.501 3.355 2.896 2.306 1.860 1.397
9 4.297 3.250 2.821 2.262 1.833 1.383
10 4.144 3.169 2.764 2.228 1.812 1.372
11 4.025 3.106 2.718 2.201 1.796 1.363
12 3.930 3.055 2.681 2.179 1.782 1.356
13 3.852 3.012 2.650 2.160 1.771 1.350
14 3.787 2.977 2.624 2.145 1.761 1.345
15 3.733 2.947 2.602 2.131 1.753 1.341
16 3.686 2.921 2.583 2.120 1.746 1.337
17 3.646 2.898 2.567 2.110 1.740 1.333
18 3.610 2.878 2.552 2.101 1.734 1.330
19 3.579 2.861 2.539 2.093 1.729 1.328
20 3.552 2.845 2.528 2.086 1.725 1.325
21 3.527 2.831 2.518 2.080 1.721 1.323
22 3.505 2.819 2.508 2.074 1.717 1.321
23 3.485 2.807 2.500 2.069 1.714 1.319
24 3.467 2.797 2.492 2.064 1.711 1.318
25 3.450 2.787 2.485 2.060 1.708 1.316
26 3.435 2.779 2.479 2.056 1.706 1.315
27 3.421 2.771 2.473 2.052 1.703 1.314
28 3.408 2.763 2.467 2.048 1.701 1.313
29 3.396 2.756 2.462 2.045 1.699 1.311
30 3.385 2.750 2.457 2.042 1.697 1.310
31 3.375 2.744 2.453 2.040 1.696 1.309
32 3.365 2.738 2.449 2.037 1.694 1.309
33 3.356 2.733 2.445 2.035 1.692 1.308
34 3.348 2.728 2.441 2.032 1.691 1.307
35 3.340 2.724 2.438 2.030 1.690 1.306
36 3.333 2.719 2.434 2.028 1.688 1.306
37 3.326 2.715 2.431 2.026 1.687 1.305
38 3.319 2.712 2.429 2.024 1.686 1.304
39 3.313 2.708 2.426 2.023 1.685 1.304
40 3.307 2.704 2.423 2.021 1.684 1.303
45 3.281 2.690 2.412 2.014 1.679 1.301

Continued

677
678 APPENDIX B  Critical Values for Student’s t Distribution

Level of Significance (α), Two-Tailed Test—cont’d

df .002 .01 .02 .05 .10 .20
50 3.261 2.678 2.403 2.009 1.676 1.299
55 3.245 2.668 2.396 2.004 1.673 1.297
60 3.232 2.660 2.390 2.000 1.671 1.296
65 3.220 2.654 2.385 1.997 1.669 1.295
70 3.211 2.648 2.381 1.994 1.667 1.294
75 3.202 2.643 2.377 1.992 1.665 1.293
80 3.195 2.639 2.374 1.990 1.664 1.292
85 3.189 2.635 2.371 1.988 1.663 1.292
90 3.183 2.632 2.368 1.987 1.662 1.291
95 3.178 2.629 2.366 1.985 1.661 1.291
100 3.174 2.626 2.364 1.984 1.660 1.290
200 3.131 2.601 2.345 1.972 1.653 1.286
300 3.118 2.592 2.339 1.968 1.650 1.284
∞ 3.1 2.58 2.33 1.96 1.65 1.28
 Appendix C

Critical Values of r for Pearson Product Moment Correlation Coefficient

Level of Significance (α), One-Tailed Test
.05 .025 .01 .005 .05 .025 .01 .005

Level of Significance (α), Two-Tailed Test

df = N – 2 .10 .05 .02 .01 df = N – 2 .10 .05 .02 .01
1 .9877 .9969 .9995 .9999 39 .2605 .3081 .3621 .3978
2 .9000 .9500 .9800 .9900 40 .2573 .3044 .3578 .3932
3 .8054 .8783 .9343 .9587 41 .2542 .3008 .3536 .3887
4 .7293 .8114 .8822 .9172 42 .2512 .2973 .3496 .3843
5 .6694 .7545 .8329 .8745 43 .2483 .2940 .3458 .3801
6 .6215 .7067 .7887 .8343 44 .2455 .2907 .3420 .3761
7 .5822 .6664 .7498 .7977 45 .2429 .2876 .3384 .3721
8 .5493 .6319 .7155 .7646 46 .2403 .2845 .3348 .3683
9 .5214 .6021 .6851 .7348 47 .2377 .2816 .3314 .3646
10 .4973 .5760 .6581 .7079 48 .2353 .2787 .3281 .3610
11 .4762 .5529 .6339 .6835 49 .2329 .2759 .3249 .3575
12 .4575 .5324 .6120 .6614 50 .2306 .2732 .3218 .3542
13 .4409 .5140 .5923 .6411 55 .2201 .2609 .3074 .3385
14 .4259 .4973 .5742 .6226 60 .2108 .2500 .2948 .3248
15 .4124 .4821 .5577 .6055 65 .2027 .2404 .2837 .3126
16 .4000 .4683 .5426 .5897 70 .1954 .2319 .2737 .3017
17 .3887 .4555 .5285 .5751 75 .1888 .2242 .2647 .2919
18 .3783 .4438 .5155 .5614 80 .1829 .2172 .2565 .2830
19 .3687 .4329 .5034 .5487 85 .1775 .2108 .2491 .2748
20 .3598 .4227 .4921 .5368 90 .1726 .2050 .2422 .2673
21 .3515 .4132 .4815 .5256 95 .1680 .1996 .2359 .2604
22 .3438 .4044 .4716 .5151 100 .1638 .1946 .2301 .2540
oasis-ebl|Rsalles|1476813899

23 .3365 .3961 .4622 .5052 120 .1496 .1779 .2104 .2324

24 .3297 .3882 .4534 .4958 140 .1386 .1648 .1951 .2155
25 .3233 .3809 .4451 .4869 160 .1297 .1543 .1827 .2019
26 .3172 .3739 .4372 .4785 180 .1223 .1455 .1723 .1905
27 .3115 .3673 .4297 .4705 200 .1161 .1381 .1636 .1809
28 .3061 .3610 .4226 .4629 250 .1039 .1236 .1465 .1620
29 .3009 .3550 .4158 .4556 300 .0948 .1129 .1338 .1480
30 .2960 .3494 .4093 .4487 350 .0878 .1046 .1240 .1371
31 .2913 .3440 .4031 .4421 400 .0822 .0978 .1160 .1283
32 .2869 .3388 .3973 .4357 450 .0775 .0922 .1094 .1210
33 .2826 .3338 .3916 .4297 500 .0735 .0875 .1038 .1149
34 .2785 .3291 .3862 .4238 600 .0671 .0799 .0948 .1049
35 .2746 .3246 .3810 .4182 700 .0621 .0740 .0878 .0972
36 .2709 .3202 .3760 .4128 800 .0581 .0692 .0821 .0909
37 .2673 .3160 .3712 .4076 900 .0548 .0653 .0774 .0857
38 .2638 .3120 .3665 .4026 1000 .0520 .0619 .0735 .0813

679
 Appendix D
Critical Values of F for α = 0.05 and α = 0.01

Please see page 681

680
 Critical Values of F for α = 0.05

df Degrees of Freedom (df) Numerator

Denominator 1 2 3 4 5 6 7 8 9 10 12 15 20 24 30 40 60 120 ∞
1 161.4 199.5 215.7 224.6 230.2 234.0 236.8 238.9 240.5 241.9 243.9 245.9 248.0 249.1 250.1 251.1 252.2 253.3 254.3
2 18.51 19.00 19.16 19.25 19.30 19.33 19.35 19.37 19.38 19.40 19.41 19.43 19.45 19.45 19.46 19.47 19.48 19.49 19.50
3 10.13 9.55 9.28 9.12 9.01 8.94 8.89 8.85 8.81 8.79 8.74 8.70 8.66 8.64 8.62 8.59 8.57 8.55 8.53
4 7.71 6.94 6.59 6.39 6.26 6.16 6.09 6.04 6.00 5.96 5.91 5.86 5.80 5.77 5.75 5.72 5.69 5.66 5.63
5 6.61 5.79 5.41 5.19 5.05 4.95 4.88 4.82 4.77 4.74 4.68 4.62 4.56 4.53 4.50 4.46 4.43 4.40 4.36
6 5.99 5.14 4.76 4.53 4.39 4.28 4.21 4.15 4.10 4.06 4.00 3.94 3.87 3.84 3.81 3.77 3.74 3.70 3.67
7 5.59 4.74 4.35 4.12 3.97 3.87 3.79 3.73 3.68 3.64 3.57 3.51 3.44 3.41 3.38 3.34 3.30 3.27 3.23
8 5.32 4.46 4.07 3.84 3.69 3.58 3.50 3.44 3.39 3.35 3.28 3.22 3.15 3.12 3.08 3.04 3.01 2.97 2.93
9 5.12 4.26 3.86 3.63 3.48 3.37 3.29 3.23 3.18 3.14 3.07 3.01 2.94 2.90 2.86 2.83 2.79 2.75 2.71
10 4.96 4.10 3.71 3.48 3.33 3.22 3.14 3.07 3.02 2.98 2.91 2.85 2.77 2.74 2.70 2.66 2.62 2.58 2.54
11 4.84 3.98 3.59 3.36 3.20 3.09 3.01 2.95 2.90 2.85 2.79 2.72 2.65 2.61 2.57 2.53 2.49 2.45 2.40
12 4.75 3.89 3.49 3.26 3.11 3.00 2.91 2.85 2.80 2.75 2.69 2.62 2.54 2.51 2.47 2.43 2.38 2.34 2.30
13 4.67 3.81 3.41 3.18 3.03 2.92 2.83 2.77 2.71 2.67 2.60 2.53 2.46 2.42 2.38 2.34 2.30 2.25 2.21
14 4.60 3.74 3.34 3.11 2.96 2.85 2.76 2.70 2.65 2.60 2.53 2.46 2.39 2.35 2.31 2.27 2.22 2.18 2.13
15 4.54 3.68 3.29 3.06 2.90 2.79 2.71 2.64 2.59 2.54 2.48 2.40 2.33 2.29 2.25 2.20 2.16 2.11 2.07
16 4.49 3.63 3.24 3.01 2.85 2.74 2.66 2.59 2.54 2.49 2.42 2.35 2.28 2.24 2.19 2.15 2.11 2.06 2.01
17 4.45 3.59 3.20 2.96 2.81 2.70 2.61 2.55 2.49 2.45 2.38 2.31 2.23 2.19 2.15 2.10 2.06 2.01 1.96
18 4.41 3.55 3.16 2.93 2.77 2.66 2.58 2.51 2.46 2.41 2.34 2.27 2.19 2.15 2.11 2.06 2.02 1.97 1.92
19 4.38 3.52 3.13 2.90 2.74 2.63 2.54 2.48 2.42 2.38 2.31 2.23 2.16 2.11 2.07 2.03 1.98 1.93 1.88
20 4.35 3.49 3.10 2.87 2.71 2.60 2.51 2.45 2.39 2.35 2.28 2.20 2.12 2.08 2.04 1.99 1.95 1.90 1.84
21 4.32 3.47 3.07 2.84 2.68 2.57 2.49 2.42 2.37 2.32 2.25 2.18 2.10 2.05 2.01 1.96 1.92 1.87 1.81
22 4.30 3.44 3.05 2.82 2.66 2.55 2.46 2.40 2.34 2.30 2.23 2.15 2.07 2.03 1.98 1.94 1.89 1.84 1.78
23 4.28 3.42 3.03 2.80 2.64 2.53 2.44 2.37 2.32 2.27 2.20 2.13 2.05 2.01 1.96 1.91 1.86 1.81 1.76
24 4.26 3.40 3.01 2.78 2.62 2.51 2.42 2.36 2.30 2.25 2.18 2.11 2.03 1.98 1.94 1.89 1.84 1.79 1.73
25 4.24 3.39 2.99 2.76 2.60 2.49 2.40 2.34 2.28 2.24 2.16 2.09 2.01 1.96 1.92 1.87 1.82 1.77 1.71
26 4.23 3.37 2.98 2.74 2.59 2.47 2.39 2.32 2.27 2.22 2.15 2.07 1.99 1.95 1.90 1.85 1.80 1.75 1.69
27 4.21 3.35 2.96 2.73 2.57 2.46 2.37 2.31 2.25 2.20 2.13 2.06 1.97 1.93 1.88 1.84 1.79 1.73 1.67
28 4.20 3.34 2.95 2.71 2.56 2.45 2.36 2.29 2.24 2.19 2.12 2.04 1.96 1.91 1.87 1.82 1.77 1.71 1.65
29 4.18 3.33 2.93 2.70 2.55 2.43 2.35 2.28 2.22 2.18 2.10 2.03 1.94 1.90 1.85 1.81 1.75 1.70 1.64
30 4.17 3.32 2.92 2.69 2.53 2.42 2.33 2.27 2.21 2.16 2.09 2.01 1.93 1.89 1.84 1.79 1.74 1.68 1.62
40 4.08 3.23 2.84 2.61 2.45 2.34 2.25 2.18 2.12 2.08 2.00 1.92 1.84 1.79 1.74 1.69 1.64 1.58 1.51
60 4.00 3.15 2.76 2.53 2.37 2.25 2.17 2.10 2.04 1.99 1.92 1.84 1.75 1.70 1.65 1.59 1.53 1.47 1.39
120 3.92 3.07 2.68 2.45 2.29 2.17 2.09 2.02 1.96 1.91 1.83 1.75 1.66 1.61 1.55 1.50 1.43 1.35 1.25
∞ 3.84 3.00 2.60 2.37 2.21 2.10 2.01 1.94 1.88 1.83 1.75 1.67 1.57 1.52 1.46 1.39 1.32 1.22 1.00

From Merrington, M., and Thompson, C.M. (1943). Tables of percentage points of the inverted beta (F) distribution. Biometrika, 33(1), 73–78.
 Critical Values of F for α = 0.01

df df Numerator
Denominator 1 2 3 4 5 6 7 8 9 10 12 15 20 24 30 40 60 120 ∞
1 4052 4999.5 5403 5625 5764 5859 5928 5982 6022 6056 6106 6157 6209 6235 6261 6287 6313 6339 6366
2 98.50 99.00 99.17 99.25 99.30 99.33 99.36 99.37 99.39 99.40 99.42 99.43 99.45 99.46 99.47 99.47 99.48 99.49 99.50
3 34.12 30.82 29.46 28.71 28.24 27.91 27.67 27.49 27.35 27.23 27.05 26.87 26.69 26.60 26.50 26.41 26.32 26.22 26.13
4 21.20 18.00 16.69 15.98 15.52 15.21 14.98 14.80 14.66 14.55 14.37 14.20 14.02 13.93 13.84 13.75 13.65 13.56 13.46
5 16.26 13.27 12.06 11.39 10.97 10.67 10.46 10.29 10.16 10.05 9.89 9.72 9.55 9.47 9.38 9.29 9.20 9.11 9.02
6 13.75 10.92 9.78 9.15 8.75 8.47 8.26 8.10 7.98 7.87 7.72 7.56 7.40 7.31 7.23 7.14 7.06 6.97 6.88
7 12.25 9.55 8.45 7.85 7.46 7.19 6.99 6.84 6.72 6.62 6.47 6.31 6.16 6.07 5.99 5.91 5.82 5.74 5.65
8 11.26 8.65 7.59 7.01 6.63 6.37 6.18 6.03 5.91 5.81 5.67 5.52 5.36 5.28 5.20 5.12 5.03 4.95 4.86
9 10.56 8.02 6.99 6.42 6.06 5.80 5.61 5.47 5.35 5.26 5.11 4.96 4.81 4.73 4.65 4.57 4.48 4.40 4.31
10 10.04 7.56 6.55 5.99 5.64 5.39 5.20 5.06 4.94 4.85 4.71 4.56 4.41 4.33 4.25 4.17 4.08 4.00 3.91
11 9.65 7.21 6.22 5.67 5.32 5.07 4.89 4.74 4.63 4.54 4.40 4.25 4.10 4.02 3.94 3.86 3.78 3.69 3.60
12 9.33 6.93 5.95 5.41 5.06 4.82 4.64 4.50 4.39 4.30 4.16 4.01 3.86 3.78 3.70 3.62 3.54 3.45 3.36
13 9.07 6.70 5.74 5.21 4.86 4.62 4.44 4.30 4.19 4.10 3.96 3.82 3.66 3.59 3.51 3.43 3.34 3.25 3.17
14 8.86 6.51 5.56 5.04 4.69 4.46 4.28 4.14 4.03 3.94 3.80 3.66 3.51 3.43 3.35 3.27 3.18 3.09 3.00
15 8.68 6.36 5.42 4.89 4.56 4.32 4.14 4.00 3.89 3.80 3.67 3.52 3.37 3.29 3.21 3.13 3.05 2.96 2.87
16 8.53 6.23 5.29 4.77 4.44 4.20 4.03 3.89 3.78 3.69 3.55 3.41 3.26 3.18 3.10 3.02 2.93 2.84 2.75
17 8.40 6.11 5.18 4.67 4.34 4.10 3.93 3.79 3.68 3.59 3.46 3.31 3.16 3.08 3.00 2.92 2.83 2.75 2.65
18 8.29 6.01 5.09 4.58 4.25 4.01 3.84 3.71 3.60 3.51 3.37 3.23 3.08 3.00 2.92 2.84 2.75 2.66 2.57
19 8.18 5.93 5.01 4.50 4.17 3.94 3.77 3.63 3.52 3.43 3.30 3.15 3.00 2.92 2.84 2.76 2.67 2.58 2.49
20 8.10 5.85 4.94 4.43 4.10 3.87 3.70 3.56 3.46 3.37 3.23 3.09 2.94 2.86 2.78 2.69 2.61 2.52 2.42
21 8.02 5.78 4.87 4.37 4.04 3.81 3.64 3.51 3.40 3.31 3.17 3.03 2.88 2.80 2.72 2.64 2.55 2.46 2.36
22 7.95 5.72 4.82 4.31 3.99 3.76 3.59 3.45 3.35 3.26 3.12 2.98 2.83 2.75 2.67 2.58 2.50 2.40 2.31
23 7.88 5.66 4.76 4.26 3.94 3.71 3.54 3.41 3.30 3.21 3.07 2.93 2.78 2.70 2.62 2.54 2.45 2.35 2.26
24 7.82 5.61 4.72 4.22 3.90 3.67 3.50 3.36 3.26 3.17 3.03 2.89 2.74 2.66 2.58 2.49 2.40 2.31 2.21
25 7.77 5.57 4.68 4.18 3.85 3.63 3.46 3.32 3.22 3.13 2.99 2.85 2.70 2.62 2.54 2.45 2.36 2.27 2.17
26 7.72 5.53 4.64 4.14 3.82 3.59 3.42 3.29 3.19 3.09 2.96 2.81 2.66 2.58 2.50 2.42 2.33 2.23 2.13
27 7.68 5.49 4.60 4.11 3.78 3.56 3.39 3.26 3.15 3.06 2.93 2.78 2.63 2.55 2.47 2.38 2.29 2.20 2.10
28 7.64 5.45 4.57 4.07 3.75 3.53 3.36 3.23 2.12 3.03 2.90 2.75 2.60 2.52 2.44 2.35 2.26 2.17 2.06
29 7.60 5.42 4.54 4.04 3.73 3.50 3.33 3.20 3.09 3.00 2.87 2.73 2.57 2.49 2.41 2.33 2.23 2.14 2.03
30 7.56 5.39 4.51 4.02 3.70 3.47 3.30 3.17 3.07 2.98 2.84 2.70 2.55 2.47 2.39 2.30 2.21 2.11 2.01
40 7.31 5.18 4.31 3.83 3.51 3.29 3.12 2.99 2.89 2.80 2.66 2.52 2.37 2.29 2.20 2.11 2.02 1.92 1.80
60 7.08 4.98 4.13 3.65 3.34 3.12 2.95 2.82 2.72 2.63 2.50 2.35 2.20 2.12 2.03 1.94 1.84 1.73 1.60
120 6.85 4.79 3.95 3.48 3.17 2.96 2.79 2.66 2.56 2.47 2.34 2.19 2.03 1.95 1.86 1.76 1.66 1.53 1.38
∞ 6.63 4.61 3.78 3.32 3.02 2.80 2.64 2.51 2.41 2.32 2.18 2.04 1.88 1.79 1.70 1.59 1.47 1.32 1.00

From Merrington, M., and Thompson, C.M. (1943). Tables of percentage points of the inverted beta (F) distribution. Biometrika, 33(1), 73–78.
 Appendix E
Critical Values of the χ2 Distribution

Alpha (α) Level

df .10 .05 .025 .01 .001
1 2.7055 3.8415 5.0239 6.6349 10.8276
2 4.6052 5.9915 7.3778 9.2103 13.8155
3 6.2514 7.8147 9.3484 11.3449 16.2662
4 7.7794 9.4877 11.1433 13.2767 18.4668
5 9.2364 11.0705 12.8325 15.0863 20.5150
6 10.6446 12.5916 14.4494 16.8119 22.4577
7 12.0170 14.0671 16.0128 18.4753 24.3219
8 13.3616 15.5073 17.5345 20.0902 26.1245
9 14.6837 16.9190 19.0228 21.6660 27.8772
10 15.9872 18.3070 20.4832 23.2093 29.5883
11 17.2750 19.6751 21.9200 24.7250 31.2641
12 18.5493 21.0261 23.3367 26.2170 32.9095
13 19.8119 22.3620 24.7356 27.6882 34.5282
14 21.0641 23.6848 26.1189 29.1412 36.1233
15 22.3071 24.9958 27.4884 30.5779 37.6973
16 23.5418 26.2962 28.8454 31.9999 39.2524
17 24.7690 27.5871 30.1910 33.4087 40.7902
18 25.9894 28.8693 31.5264 34.8053 42.3124
19 27.2036 30.1435 32.8523 36.1909 43.8202
20 28.4120 31.4104 34.1696 37.5662 45.3147
21 29.6151 32.6706 35.4789 38.9322 46.7970
22 30.8133 33.9244 36.7807 40.2894 48.2679
23 32.0069 35.1725 38.0756 41.6384 49.7282
24 33.1962 36.4150 39.3641 42.9798 51.1786
25 34.3816 37.6525 40.6465 44.3141 52.6197
26 35.5632 38.8851 41.9232 45.6417 54.0520
27 36.7412 40.1133 43.1945 46.9629 55.4760
28 37.9159 41.3371 44.4608 48.2782 56.8923
29 39.0875 42.5570 45.7223 49.5879 58.3012
30 40.2560 43.7730 46.9792 50.8922 59.7031
31 41.4217 44.9853 48.2319 52.1914 61.0983
32 42.5847 46.1943 49.4804 53.4858 62.4872
33 43.7452 47.3999 50.7251 54.7755 63.8701
34 44.9032 48.6024 51.9660 56.0609 65.2472
35 46.0588 49.8018 53.2033 57.3421 66.6188
36 47.2122 50.9985 54.4373 58.6192 67.9852
37 48.3634 52.1923 55.6680 59.8925 69.3465
38 49.5126 53.3835 56.8955 61.1621 70.7029
39 50.6598 54.5722 58.1201 62.4281 72.0547
40 51.8051 55.7585 59.3417 63.6907 73.4020
41 52.9485 56.9424 60.5606 64.9501 74.7449
42 54.0902 58.1240 61.7768 66.2062 76.0838
43 55.2302 59.3035 62.9904 67.4593 77.4186
44 56.3685 60.4809 64.2015 68.7095 78.7495
45 57.5053 61.6562 65.4102 69.9568 80.0767

From Corty, E. (2007). A Practical Text for the Health, Behavioral and Social Sciences. St. Louis, MO: Mosby.

683
 Appendix F
Statistical Power Tables (Δ = Effect Size)

5% Level, One-Tailed Test

Power
Δ 99 95 90 80 70 60 50 40 30 20 10
0.01 157695 108215 85634 61823 47055 36031 27055 19363 12555 6453 1321
0.02 39417 27050 21405 15454 11763 9007 6764 4841 3139 1614 331
0.03 17514 12019 9511 6867 5227 4003 3006 2152 1396 718 148
0.04 9848 6578 5348 3861 2939 2251 1691 1210 785 404 84
0.05 6299 4323 3421 2470 1881 1440 1082 775 503 259 54
0.06 4372 3000 2375 1715 1305 1000 751 538 349 180 38
0.07 3209 2203 1744 1259 959 734 552 395 257 133 29
0.08 2455 1685 1334 963 734 562 422 303 197 102 23
0.09 1938 1330 1053 761 579 444 334 239 156 81 18
0.10 1568 1076 852 616 469 359 270 194 126 66 15
0.11 1294 889 704 508 387 297 223 160 104 54 13
0.12 1086 746 590 427 325 249 188 135 88 46 11
0.13 924 635 503 363 277 212 160 115 75 39 10
0.14 796 546 433 313 238 183 138 99 65 34 *
0.15 692 475 376 272 207 159 120 86 56 30 *
0.16 607 417 330 239 182 140 105 76 50 27 *
0.17 537 369 292 211 161 124 93 67 44 24 *
0.18 478 328 260 188 144 110 83 60 39 21 *
0.19 428 294 233 169 129 99 75 54 35 19 *
0.20 385 265 210 152 116 89 67 49 32 17 *
0.22 317 218 173 125 96 74 56 40 27 15 *
0.24 265 182 144 105 80 62 47 34 23 12 *
0.26 224 154 122 89 68 52 40 29 19 11 *
0.28 192 132 105 76 58 45 34 25 17 10 *
0.30 166 114 91 66 51 39 30 22 15 * *
0.32 145 100 79 58 44 34 26 19 13 * *
0.34 127 88 70 51 39 30 23 17 12 * *
0.36 113 78 62 45 35 27 20 15 10 * *
0.38 100 69 55 40 31 24 18 14 * * *
0.40 89 62 49 36 28 21 16 12 * * *
0.45 69 48 38 28 21 17 13 10 * * *
0.50 54 37 30 22 17 13 10 * * * *
0.55 43 30 24 17 14 11 * * * * *
0.60 34 24 19 14 11 * * * * * *
0.65 28 19 16 12 * * * * * * *
0.70 23 16 13 10 * * * * * * *
0.75 18 13 10 * * * * * * * *
0.80 15 10 * * * * * * * * *
0.85 12 * * * * * * * * * *
0.90 * * * * * * * * * * *

From Kraemer, H.C., and Thiemann, S. (1987). How Many Subjects: Statistical Power Analysis in Research. Newbury Park, CA: Sage.

684
 APPENDIX F  Statistical Power Tables (Δ = Effect Size) 685

1% Level, One-Tailed Test

Power
Δ 99 95 90 80 70 60 50 40 30 20 10
0.01 216463 157695 130162 100355 81264 66545 54117 42972 32469 22044 10917
0.02 54106 39417 32535 25085 2031 16634 13528 10742 8117 5511 2730
0.03 24040 17514 14456 11146 9026 7391 6011 4773 3607 2449 1214
0.04 13517 9848 8128 6267 5075 4156 3380 2684 2029 1378 683
0.05 8646 6299 5200 4009 3247 2659 2163 1718 1298 882 437
0.06 6000 4372 3609 2783 2254 1846 1501 1192 901 612 304
0.07 4405 3209 2649 2043 1655 1355 1102 876 662 450 224
0.08 3369 2455 2027 1563 1266 1037 843 670 507 344 171
0.09 2660 1938 1600 1234 999 819 666 529 400 272 136
0.10 2152 1568 1295 998 809 663 539 428 324 220 110
0.11 1776 1294 1069 824 668 547 445 354 268 182 91
0.12 1490 1086 897 692 560 459 374 297 225 153 77
0.13 1268 924 763 589 477 391 318 253 191 130 65
0.14 1092 796 657 507 411 337 274 218 165 112 56
0.15 949 692 571 441 357 293 238 190 144 98 49
0.16 833 607 501 387 314 257 209 166 126 86 43
0.17 736 537 443 342 277 277 185 147 112 76 39
0.18 655 478 395 305 247 202 165 131 100 68 34
0.19 587 428 353 273 221 181 148 118 89 61 31
0.20 528 385 318 246 199 163 133 106 81 55 29
0.22 434 317 262 202 164 135 110 87 66 46 24
0.24 363 265 219 169 137 113 92 73 56 38 21
0.26 307 224 185 143 116 95 78 62 47 33 18
0.28 263 192 159 123 100 82 67 53 41 29 16
0.30 227 166 137 106 86 71 58 46 35 25 14
0.32 198 145 120 93 75 62 51 41 31 22 12
0.34 174 127 105 82 66 55 45 36 28 20 11
0.36 154 113 93 72 59 48 40 32 25 18 10
0.38 137 100 83 64 52 43 35 29 22 16 *
0.40 122 89 74 57 47 39 32 26 20 15 *
0.45 94 69 57 44 36 30 25 20 16 12 *
0.50 73 54 45 35 29 24 20 16 13 * *
0.55 58 43 36 28 23 19 16 13 11 * *
0.60 47 34 29 23 19 16 13 11 * * *
0.65 38 28 23 18 15 13 11 * * * *
0.70 30 23 19 15 12 11 * * * * *
0.75 25 18 15 12 10 * * * * * *
0.80 20 15 12 10 * * * * * * *
0.85 16 12 10 * * * * * * * *
0.90 12 * * * * * * * * * *

From Kraemer, H.C., and Thiemann, S. (1987). How Many Subjects: Statistical Power Analysis in Research. Newbury Park, CA: Sage.
 Glossary
A agencies, and others not directly involved in providing
absolute zero point  Point at which a value of zero indi-
cates the absence of the property being measured. Ratio-
level measurements, such as weight scales, vital signs,
and laboratory values, have an absolute zero point.
abstract  Clear, concise summary of a study, usually
limited to 100 to 250 words.
abstract thinking  Oriented toward the development of an
idea without application to or association with a particular
instance, and independent of time and space. Abstract
thinkers tend to look for meaning, patterns, relationships,
and philosophical implications.
abstract thought processes  Enable both science and the-
ories to be blended into a cohesive body of knowledge,
guided by a philosophical framework and applied to clini-
cal practice.
acceptance rate  Number or percentage of the subjects who
agree to participate in a study. The percentage is calculated
by dividing the number of subjects agreeing to participate
by the number of subjects approached. For example, if
100 subjects are approached and 90 agree to participate,
the acceptance rate is 90% (90 ÷ 100 × 100% = 90%).
accessible population  Portion of a target population to
which the researcher has reasonable access.
accidental or convenience sampling  Nonprobability
sampling technique in which subjects are included in the
study because they happened to be in the right place at
the right time. Available subjects are simply entered into
the study until the desired sample size is reached.
accuracy  The closeness of the agreement between the
measured value and the true value of the quantity being
measured.
accuracy in physiological measures  Comparable to
validity in that it addresses the extent to which the instru-
ment measured the domain that is defined in the study.
accuracy of a screening test  Test used to confirm a diag-
nosis, evaluating it in terms of its ability to correctly
assess the presence or absence of a disease or condition
in comparison with a gold standard.
adjusted  Term used when each hazard ratio (HR) has been
adjusted for every other predictor in the regression model.
administrative databases  Databases with standardized
sets of data for enormous numbers of patients and provid-
ers that are created by insurance companies, government
686
patient care.
Agency for Healthcare Research and Quality (AHRQ) 
Federal government agency created in 1989 to carry out
research, demonstration projects, evidence-based guide-
line development, training, and research dissemination
activities with respect to healthcare services and systems.
Focus of this agency is to promote evidence-based health
care; see the website for details at www.ahrq.gov. AHRQ
was previously named Agency for Health Care Policy and
Research, with a name change in 1999.
alpha (α)  Level of significance or cutoff point used to
determine whether the samples being tested are members
of the same population (nonsignificant) or different popu-
lations (significant); alpha is commonly set at 0.05, 0.01,
or 0.001. Alpha is also the probability of making a type I
error.
alternate-forms reliability  Also referred to as parallel
forms reliability, which involves comparing the equiva-
lence of two versions of the same paper-and-pencil
instruments.
analysis of covariance (ANCOVA)  Statistical procedure
designed to reduce the error term (or variance within
groups) by partialing out the variance resulting from a
confounding variable by performing regression analysis
before performing analysis of variance (ANOVA).
analysis of sources  Process of determining the value of a
reference for a particular study. The source is critically
appraised and then compared with that of other studies to
determine the existing body of knowledge in relation to
the research problem.
analysis of variance (ANOVA)  Statistical technique used
to examine differences among two or more groups by
comparing the variability between the groups with the
variability within the groups.
ancestry searches  Use of citations in relevant studies to
identify additional studies. Especially important in con-
ducting a systematic review of research.
anonymity  In research, the condition in which a subject’s
identity cannot be linked, even by the researcher, with his
or her individual responses.
applied or practical research  Scientific investigations
conducted to generate knowledge that will directly influ-
ence or improve practice.

apprenticeship  A volunteer position in which one works
closely with an experienced researcher in order to develop
one’s research skills.
assent  A child’s affirmative agreement to participate in
research.
associative hypothesis  Identifies or predicts the relation-
ship between or among variables that occur or exist
together in the real world. It is usually developed to guide
a correlational study.
associative relationship  Identifies concepts that occur or
exist together in the real world; thus, when one concept
changes, the other concept changes. These relationships
are part of theory and can be tested through research.
assumptions  Statements taken for granted or considered
true, even though they have not been scientifically tested.
asymmetrical relationship  If A occurs (or changes), then
B will occur (or change), but there may be no indication
that if B occurs (or changes), A will occur (or change).
delete A_B; A—B.
attrition rate  The number and percentage of subjects or
study participants who drop out of a study before comple-
tion, which creates a threat to the study’s internal
validity.
authority  Person with expertise and power who is able to
influence opinion and behavior.
autoethnography  The formal study of one’s own culture
or social context.
B variables.
background for a research problem  Part of the research
problem that indicates what is known or the key research
that has been done in the problem area to be studied.
bar graphs  Figures or illustrations that provide a picture
of the results from a study. These graphs can be horizontal
or vertical bars that represent the size or amount of the
group or variable studied.
basic or pure research  Scientific investigations for the
pursuit of knowledge for knowledge’s sake or for the
pleasure of learning and finding truth.
being  A term in phenomenological research whereby a
person’s experiences of the world are unique to that
person.
being-in-time  Term from phenomenological research that
indicates that a person experiences life situations within
the framework of time and that the past and the future
influence the now and thus are part of being-in-time.
beneficence, principle of  Encourages the researcher to do
good and, above all, to do no harm.
benefit-risk ratio  Means by which researchers and
reviewers of research judge the potential benefits and
risks in a study to promote the conduct of ethical research.
best interest standard  In determining whether an indi-
vidual should participate in a study, the researcher needs
Glossary 687
to do what is best for the individual on the basis of balanc-
ing risks and benefits in a study.
best research evidence  The strongest empirical knowl-
edge available that is generated from the synthesis of
quality study findings to address a practice problem.
between-group variance  Variance of the group means
around the grand mean (the mean of the total sample) that
is examined in analysis of variance (ANOVA).
bias  Any influence or action in a study that distorts the
findings or slants them away from the true or expected.
biased coin design  Technique used to randomly assign
subjects to groups in which selection of the group to
which a particular subject will be assigned is biased in
favor of groups that have smaller sample sizes at the point
of the assignment of that subject.
bibliographical database  Database that either consists of
citations relevant to a specific discipline or is a broad
collection of citations from a variety of disciplines.
bimodal  Distribution of scores from a study has two
modes (or most frequently occurring scores), which
usually means that the researcher has not adequately
defined the study population.
bivariate analysis  Statistical procedures that involve
comparison of summary values from two groups of 
the same variable or from two variables within a single
group.
bivariate correlation analysis  Analysis techniques that
measure the extent of the linear relationship between two
Bland and Altman plot  Analysis technique for examining
the extent of agreement between two physiological mea-
surement techniques. Generally used to compare a new
technique with an established one.
blinding  Structure of a design whereby either the patient
or those providing care to the patient are unaware whether
the patient is in the experimental group or the control
group.
blocking  Part of the randomized block design, in which
the subjects are rank ordered in relation to the blocking
variable to control the effects of this variable, thus
improving the validity of study findings.
body of knowledge  Information, principles, theories, and
empirical evidence that are organized by the beliefs
accepted in a discipline at a given time.
Bonferroni procedure  Parametric analysis technique that
controls for escalation of significance and that can be used
if various t-tests must be performed on different aspects
of the same data.
borrowing  Appropriation and use of knowledge from
other disciplines to guide nursing practice.
bracketing  Qualitative research technique of suspending
or laying aside what is known about an experience being
studied.
688 Glossary
breach of confidentiality  Accidental or direct action that
allows an unauthorized person to have access to raw study
data or subject identity information.
C citation bias  Occurs when certain studies are cited more
calculated variable  A variable used in the analysis is not
collected but calculated from other variables.
canonical correlation  Extension of multiple regression
with more than one dependent variable.
care bundles  Combinations of interrelated nursing
actions.
care maps  Type of map developed in intervention research
of the intervention theory that illustrates the elements of
the intervention and the causal links among them, and that
should show all the causal pathways described in the
intervention theory.
carryover effect  Application of one treatment can influ-
ence the response to following treatments.
case-control design  Involves a matching procedure
whereby a control subject is matched to each case, so that
the cases and controls are different people matched
demographically.
case study design  A design that guides the intensive
exploration of a single unit of study, such as a person,
family, group, community, or institution.
causal connection  The link between the independent vari-
able (cause) and the dependent variable (outcome or
effect) that is examined in quasi-experimental, experi-
mental, and intervention research.
causal explanation  The description or explanation of the
effect(s) of the independent variable (cause or interven-
tion) on the dependent variable (outcome).
causal hypothesis or relationship  Relationship between
two variables in which one variable (independent variable)
is thought to cause or determine the presence of the other
variable (dependent variable). Some causal hypotheses
include more than one independent or dependent variable.
causality  Has three conditions: (1) there must be a
strong relationship between the proposed cause and
effect, (2) the proposed cause must precede the effect in
time, and (3) the cause has to be present whenever the
effect occurs.
cell  Intersection between the row and column in a table
where a specific numerical value is inserted.
censored data  Survival times that are known only to
exceed a certain value.
central limit theorem  States that even when statistics,
such as means, come from a population with a skewed
(asymmetrical) distribution, the sampling distribution
developed from multiple means obtained from that skewed
population will tend to fit the pattern of the normal curve.
chi-square test of independence (χ2)  Used to analyze
nominal data to determine significant differences between
observed frequencies within the data and frequencies that
were expected.
citation  Act of quoting a source, using it as an example,
or presenting it as support for a position taken.
often than others and are more likely to be identified in
database searches.
classical hypothesis testing  Refers to the process of
testing a hypothesis to infer the reality of an effect.
cleaning data  Checking raw data to determine errors in
data recording, coding, or entry.
client treatment matching  Also called treatment match-
ing; is used in intervention research to compare the rela-
tive effectiveness of various treatments when the following
conditions are met: (1) there is no clearly superior treat-
ment for individuals with a problem, (2) a number of
treatments have some proven efficacy for undifferentiated
subjects, and (3) there is evidence of differential outcomes
with and among treatments for defined subtypes of
patients.
clinical databases  Databases of patient, provider, and
healthcare agency information that are developed by
healthcare agencies and sometimes providers to document
care delivery and outcomes.
clinical expertise  Consists of a practitioner’s knowledge,
skills, and past experience in accurately assessing, diag-
nosing, and managing an individual’s health needs.
clinical guidelines  Standardized, current national and
international guidelines for the assessment, diagnosis, and
management of patient conditions that are developed by
clinical guideline panels or professional groups to improve
the outcomes of care and promote evidence-based health
care (see www.guideline.gov).
clinical judgment  The use of clinical expertise to make
sound decisions in the provision of evidence-based health
care.
clinical pathways  Critical pathways or guidelines devel-
oped by healthcare agencies to define the expected care
activities and outcomes of care in specific patient care
situations. These pathways are developed on the basis of
previous research, agency data, and clinical experience to
provide evidence-based practice.
clinical trial  A study conducted to determine the effect of
a selected intervention (such as a drug or nursing or
medical procedure) on identified patient outcomes using
a structured experimental design. Produces a strong type
of research evidence that can be synthesized with other
study findings to determine the current best research evi-
dence in a practice area.
cluster sampling  A sampling frame is developed that
includes a list of all the states, cities, institutions, or orga-
nizations (clusters) that could be used in a study, and a
randomized sample is drawn from this list. Cluster

sampling is used when it is not possible to use simple
random sampling or the individual elements of the popu-
lation are unknown, preventing development of the sam-
pling frame.
code  A symbol or abbreviation used to label words or
phrases in the data.
codebook  Identifies and defines each variable in a study
and includes an abbreviated variable name, a descriptive
variable label, and the range of possible numerical values
of every variable entered into a computer file.
coding  In quantitative studies, the process of transforming
qualitative data into numerical symbols that can be com-
puterized. In qualitative studies, the process of labeling
phrases and quotations to identify themes and patterns.
coefficient of determination (R2)  Computed from a
matrix of correlation coefficients; provides important
information on multicollinearity. This value indicates the
degree of linear dependencies among the variables.
coefficient of stability  Result of a correlational analysis
of the scores of two educational tests or scales given two
to four weeks apart.
coercion  Overt threat of harm or excessive reward inten-
tionally presented by one person to another to obtain
compliance, such as offering subjects a large sum of
money to participate in a dangerous research project.
cognitive application of research  Use of research-based
knowledge to affect a person’s way of thinking about,
approaching, and observing situations.
cohorts  Samples in time-dimensional studies within the
field of epidemiology.
communicating research findings  Developing a research
report and disseminating it through presentations and
publications to a variety of audiences.
comparative analysis  Involves examining methodology
and findings across studies for similarities and differ-
ences. The frequency of similar findings might be
recorded.
comparative descriptive design  Used to describe differ-
ences in variables in two or more groups in a natural
setting.
comparative evaluation  Has four parts: (1) substantiation
of the evidence, (2) fit of the evidence with the healthcare
setting, (3) feasibility of using research findings, and  ing and is established through concept analysis, concept
(4) concerns with current practice.
comparative experimental design  Less rigorous experi-
mental design in which random sampling is difficult if not
impossible. The studies include convenience samples
with random assignment to groups.
comparison group  A group of subjects that is not selected
through random sampling and usually does not receive a
treatment. There are four types of comparison groups:  conceptual model  Set of highly abstract, related con-
(1) groups that receive no treatment, (2) groups that receive
a placebo treatment, (3) groups that receive standard or
Glossary 689
usual healthcare, and (4) groups that receive a second
experimental treatment or a different treatment dose for
comparison with the first experimental treatment.
complete IRB review  Extensive review by an institu-
tional review board (IRB) for studies with greater than
minimal risk.
complete observation  The researcher is passive and has
no direct social interaction in the setting.
complete participation  The researcher becomes a mem­
ber of the group and conceals the researcher role.
complex hypothesis  Predicts the relationship (associative
or causal) among three or more variables; thus, the
hypothesis could include two (or more) independent and
two (or more) dependent variables.
comprehending a source  Involves reading the entire
source carefully and focusing on understanding the 
major concepts and the logical flow of ideas within the
source.
computer searches  Conducted to scan the citations in dif-
ferent databases and identify sources relevant to a research
problem.
computerized database  Structured compilation of infor-
mation that can be scanned, retrieved, and analyzed by
computer and can be used for decisions, reports, and
research.
concept  Term that abstractly describes and names an
object or phenomenon, thus providing it with a separate
identity or meaning.
concept analysis  Strategy through which a set of attributes
or characteristics essential to the connotative meaning or
conceptual definition of a concept are identified.
concept derivation  Process of extracting and defining
concepts from theories in other disciplines. May require
a concept analysis that examines the use of the concept
in the nursing literature, compares the results with the
existing conceptual definition, and, if the two are differ-
ent, modifies the definition to be consistent with nursing
usage.
concept synthesis  Process of describing and naming a
previously unrecognized concept.
conceptual definition  Provides a variable or concept with
connotative (abstract, comprehensive, theoretical) mean­
derivation, or concept synthesis. The conceptual defini-
tion of a variable in a study is often developed from the
study framework and is the link between the study frame-
work and the operational definition of the variable.
conceptual map  Strategy for expressing a framework of
a study that diagrammatically shows the interrelationships
of the concepts and statements.
structs that broadly explains phenomena of interest,
expresses assumptions, and reflects a philosophical stance.
690 Glossary
conclusions  Synthesis and clarification of the meaning of
study findings.
concrete thinking  Thinking that is oriented toward and
limited by tangible things or events observed and experi-
enced in reality.
concurrent relationship  Relationship in which both vari-
ables and concepts occur simultaneously.
concurrent triangulation strategy  This design model is
selected when a researcher wishes to use quantitative and
qualitative methods in an attempt to confirm, cross-vali-
date, or corroborate findings within a single study.
condensed proposal  A brief or shortened proposal devel-
oped for review by clinical agencies and funding
institutions.
confidence interval  Range in which the value of the popu-
lation parameter is estimated to be.
confidentiality  Management of private data in research so
that subjects’ identities are not linked with their responses.
confirmatory data analysis  Use of inferential statistics to
confirm expectations regarding the data that are expressed
as hypotheses.
confirmatory studies  Conducted only after a large body
of knowledge has been generated with exploratory
studies; are expected to have large samples and to use
random sampling techniques.
confounding variables  Variables that have the potential
to affect the outcome of a study, which are recognized
before the study is initiated but that cannot be controlled,
or variables not recognized until the study is in process.
consensus knowledge building  Outcomes design that
requires critical appraisal and synthesis of an extensive
international search of the literature on the topic of
concern, including unpublished studies, studies in prog-
ress, dissertations, and theses.
consent form  Written form, audio recording, or video
recording used to document a subject’s agreement to par-
ticipate in a study.
concurrent validity  Ability to predict the current value of
one measure on the basis of the value obtained on the
measure of another concept.
construct validity  Examines the fit between conceptual
and operational definitions of variables and determines
whether the instrument actually measures the theoretical
construct that it purports to measure.
constructs  Concepts at very high levels of abstraction that
have general meanings.
content analysis  Qualitative analysis technique used to
classify words in a text into a few categories chosen
because of their theoretical importance.
content expert  A clinician or researcher who is known for
his or her work in a specific area.
content validity  Examines the extent to which the mea-
surement method includes all the major elements
relevant to the construct being measured. Evidence for
content-related validity is obtained from the literature,
representatives of the relevant populations, and content
experts.
content validity ratio  Calculated by researchers for each
item on a scale by rating it a 0 (not necessary), 1 (useful),
or 3 (essential).
content validity index  Developed to obtain a numerical
value that reflects the level of content-related validity
evidence for a measurement method.
contingent relationship  Occurs only if a third variable or
concept is present.
continuous variable  Variable in which higher numbers
represent more of that variable and the lower numbers
represent less of that variable.
control  Imposing of rules by the researcher to decrease
the possibility of error and increase the probability that
the study’s findings are an accurate reflection of reality.
control group  Group of elements or subjects not exposed
to the experimental treatment. The term control group is
used in studies with random sampling methods.
convenience sampling  See accidental sampling.
convergent validity  Type of measurement validity
obtained by using two instruments to measure the same
variable, such as depression, and correlating the results
from these instruments. Evidence of validity from examin-
ing convergence is achieved if the data from the two instru-
ments have a moderate to strong positive correlation.
correlational analysis  Statistical procedure conducted to
determine the direction (positive or negative) and magni-
tude (or strength) of the relationship between two
variables.
correlational coefficient  Indicates the degree of rela­
tionship between two variables; coefficients range in
value from +1.00 (perfect positive relationship) to 0.00
(no relationship) to −1.00 (perfect negative or inverse
relationship).
correlational matrix  A table of the bivariate correlations
of every pair of variables in a data set. Along the diagonal
through the matrix the variables are correlated with them-
selves, with the left and right sides of the table being
mirror images of each other.
correlational research  Systematic investigation of rela-
tionships between two or more variables to explain the
type (positive or negative) and strength of relationships
in the world and not to examine cause and effect.
correlational study designs  Variety of study designs
developed to examine relationships among variables.
counterbalancing  Administration of various treatments in
random order rather than consistently in the same
sequence.
covert data collection  Occurs when subjects are unaware
that research data are being collected.

criterion-referenced testing  Comparison of a subject’s
score with a criterion of achievement that includes the
definition of target behaviors. When the subject has mas-
tered the behaviors, he or she is considered proficient in
these behaviors, such as being proficient in the behaviors
of a nurse practitioner.
critical appraisal of research  Systematic, unbiased,
careful examination of all aspects of a study to judge the
merits, limitations, meaning, and significance based on
previous research experience and knowledge of the topic.
critical appraisal process for qualitative research  Eval-
uating the quality of a qualitative study by examining its
philosophical congruence, methodological coherence,
intuitive comprehension, and intellectual contribution.
critical appraisal process for quantitative research  data use agreement  Limits how the data set for a study
Examination of the quality of a quantitative study using
the following three steps: (1) identifying the steps of  Insurance Portability and Accountability Act (HIPAA)
the research process, (2) determining the study’s
strengths and weaknesses, and (3) evaluating the credi-
bility and meaning of a study to nursing knowledge and
practice.
critical cases  Cases that make a point clearly or are
extremely important in understanding the purpose of the
study and are identified through purposive sampling.
critical pathways  See clinical pathways.
critical value  In quantitative data analysis, the value at
which statistical significance is achieved in a study.
crossover or counterbalanced design  Includes the
administration of more than one treatment to each subject,
and the treatments are provided sequentially rather than
concurrently; comparisons are then made of the effects of
the different treatments on the same subject.
cross-sectional designs  Used to examine groups of sub-
jects in various stages of development simultaneously
with the intent of inferring trends over time.
cultural immersion  The spending of extended periods in
the culture one is studying using ethnographic methods to
gain increased familiarity with such things as language,
sociocultural norms, and traditions in a culture.
curvilinear relationship  A relationship between two vari-
ables that varies depending on the relative values of the
variables. The graph of the relationship is a curved line
rather than a straight one.
cutoff point  In factor analysis, variables’ factor loading
must be at least 0.30 to explain a meaningful portion of
the variance within a factor, and these variables are
included as elements of the factor.
D denominator  The number on the bottom part of a
data  (plural) Pieces of information that are collected
during a study (singular: datum).
data analysis  Conducted to reduce, organize, and give
meaning to data.
Glossary 691
data coding sheet  A sheet for organizing and recording
data for rapid entry into a computer.
data collection  Precise, systematic gathering of informa-
tion relevant to the research purpose or the specific objec-
tives, questions, or hypotheses of a study.
data collection forms  Forms developed or modified for a
study to use for recording demographic data, information
from patient records, observations, or values from physi-
ological measures.
data collection plan  Details on how a study will be
implemented.
data saturation  In qualitative research, occurs when addi-
tional sampling provides no new information, only redun-
dancy of previous collected data.
may be used and how it will be protected to meet Health
requirements.
datum  Single piece of information collected for research.
debriefing  Complete disclosure of the study purpose and
results at the end of a study.
deception  Misinforming subjects for research purposes.
decision making  Cognitive process of assessing a situa-
tion and deciding on a course of action, which is impor-
tant for conducting research and providing health care.
Declaration of Helsinki  Ethical code based on the
Nuremberg Code that differentiated therapeutic from non-
therapeutic research.
deductive reasoning  Reasoning from the general to the
specific, or from a general premise to a particular situation.
deductive thinking  Thinking that begins with a theory
or abstract principle that guides the selection of methods
to gather data to support or refute the theory or
principle.
degrees of freedom (df )  Freedom of a score’s value to
vary given the other existing scores’ values and the estab-
lished sum of these scores (df = N − 1).
de-identifying health data  Involves removing the 18 ele-
ments that could be used to identify an individual or his
or her relatives, employer, or household members; this
term is part of the Health Insurance Portability and
Accountability Act (HIPAA).
Delphi technique  Method of measuring the judgments of
a group of experts for assessing priorities or making
forecasts.
demographic or attribute variables  Specific variables
such as age, gender, and ethnicity that are collected in a
study to describe the sample.
fraction
denotative definition  Dictionary definition of a word.
dependent groups  Groups in which the subjects or obser-
vations selected for data collection are in some way
692 Glossary
related to the selection of other subjects or observations.
For example, if subjects serve as their own control by
using the pretest as a control, the observations (and there-
fore the groups) are dependent. Use of twins in a study or
matching subjects on a selected variable, such as medical
diagnosis or age, results in dependent groups.
dependent variable  Response, behavior, or outcome that
is predicted and measured in research; changes in the
dependent variable are presumed to be caused by the
independent variable.
description  Involves identifying and understanding the
nature and attributes of nursing phenomena and some-
times the relationships among these phenomena. This is
an outcome of research.
descriptive correlational design  Used to describe vari-
ables and examine relationships that exist in a study
situation.
descriptive design  Used to identify a phenomenon of
interest, identify variables within the phenomenon,
develop conceptual and operational definitions of vari-
ables, and describe variables in a study situation.
descriptive research  Provides an accurate portrayal or
account of the characteristics of a particular individual,
event, or group in real-life situations for the purpose of
discovering new meaning, describing what exists, deter-
mining the frequency with which something occurs, and
categorizing information.
descriptive statistics  Summary statistics that allow the
researcher to organize the data in ways that give meaning
and facilitate insight, such as frequency distributions and
measures of central tendency and dispersion.
descriptive study designs  Variety of designs developed to
gain more information about characteristics within a par-
ticular field of study and to provide a picture of situations
as they naturally happen.
descriptive theory  Describes the causal process occurring
in intervention research.
design  Blueprint for conducting a study that maximizes
control over factors that could interfere with the validity
of the findings.
design validity  Strength of a design to produce accurate
results, which is determined by examining statistical con-
clusion validity, internal validity, construct validity, and
external validity.
deterministic relationship  Causal statement of what
always occurs in a particular situation, such as a scientific
law.
deviation score  Difference score, which is obtained by
subtracting the mean from each score; indicates the extent
to which a score deviates from the mean.
dialectic reasoning  Involves the holistic perspective, in
which the whole is greater than the sum of the parts, and
examining factors that are opposites and making sense of
them by merging them into a single unit or idea that is
greater than either alone.
diary  Record of events kept by a subject over time that is
collected and analyzed by a researcher.
difference scores  See deviation score.
diminished autonomy  Describes subjects with decreased
ability to voluntarily give informed consent because of
legal or mental incompetence, terminal illness, or confine-
ment to an institution.
direct costs  Specific costs for materials and equipment to
conduct a study that are identified in a grant proposal.
direct measurement  Measurement object and measure-
ment strategies are specific and straightforward, such as
those for measuring the concrete variables of height,
weight, or temperature.
direction of a relationship  The direction of a relationship
can be positive or negative. With a positive relationship,
the two variables change in the same direction (increase
or decrease together). With a negative relationship, the
variables change in opposite directions; thus, as one vari-
able increases, the other decreases.
directional hypothesis  Predicts the specific nature of the
interaction or relationship between two or more variables.
disproportionate sampling  In stratification, when each
stratum has an equivalent number of subjects in the
sample, versus each stratum having a number of subjects
in proportion to its occurrence in the population, which is
proportionate sampling.
dissemination of research findings  Diffusion or com­
munication of research findings by presentations and pub-
lications to a variety of audiences, such as nurses, other
health professionals, policy developers, and consumers.
dissertation  Extensive, usually original, research project
that is completed as the final requirement for a doctoral
degree.
distribution-free  No assumption has been made for a
normal distribution of values in the population from
which the sample was taken.
distribution  Spread of scores in a study or database.
divergent validity  Type of measurement validity obtained
by finding an instrument that measures the opposite of the
concept or variable being studied and correlating the data
collected with an instrument that measures the concept or
variable studied. For example, if the focus of a study is
to measure hope, one instrument would measure hope and
another would measure despair.
dose-intensity  The amount of the intervention delivered
in terms of the (1) components of the intervention, (2)
duration of a single session for the intervention, (3) fre-
quency with which the intervention is delivered (e.g.,
daily, times per week, times per month), and (4) cumula-
tive intervention intensity (number of treatments received
and duration).

double-blinding in a study design  Structure of a design
in which neither the patient nor the caregivers are aware
of the group (experimental or control) to which the patient
is assigned during a study.
dummy variables  Categorical or dichotomous variables
used in regression analysis.
duplicate publication bias  Practice of publishing the
same article or major portions of the article in two or more
print or electronic media without notifying the editors or
referencing the other publication in the reference list.
duration of an intervention  The time required to deliver
a treatment in a study, which includes examining the time
for each individual treatment and the total time the treat-
ment is delivered during the study.
dwelling with the data  Immersion in the data as part of
the process of data management and reduction in
phenomenology.
E the scores in a sample. Much of inferential statistical
ebooks  Books available in a digital or electronic format.
effect size  Degree to which the phenomenon is present in
the population or to which the null hypothesis is false.
effectiveness  A treatment or intervention indicates that it
is capable of producing positive results in a usual or
routine care condition.
element  Person (subject or participant), event, behavior,
or any other single unit of a study.
eligibility criteria  See sampling criteria.
eliminative induction  Qualitative data analysis technique
that is part of a process referred to as analytic induction
and requires that the hypothesis generated from the analy-
sis be tested against alternatives.
embodied  Heideggerian phenomenologist’s belief that the
person is a self within a body; thus the person is referred
to as embodied.
embodiment  The unity of body and mind that eliminates
the idea of a subjective and objective world.
emic view  Anthropological research approach of studying
behavior from within the culture.
empirical generalizations  Statements that have been
repeatedly tested through research and have not been dis-
proved. Scientific theories have empirical generalizations.
empirical literature  Includes relevant studies published in
journals, in books, and online, as well as unpublished
studies, such as master’s theses and doctoral dissertations.
empirical world  Experienced through our senses; the
concrete portion of our existence.
endogenous variables  Variables whose variations are
explained within the theoretical model as part of a study
with a model-testing design.
environmental variable  Type of extraneous variable
related to the setting in which a study is conducted.
epistemology  A view of knowing and knowledge.
Glossary 693
equivalence reliability  Compares two versions of the
same paper-and-pencil instrument or two observers mea-
suring the same event.
error score  Amount of random error in the measurement
process.
errors in physiological measures  Sources of erroneous
measurement with physiological instruments; include
environment, user, subject, machine, and interpretation
errors.
escalates significance  During analysis of study data, there
is an increased identification of significant findings that
might not be an accurate reflection of reality and might
be a type I error (saying something is significant when it
is not). For example, performing multiple t-tests to
analyze study data can cause an increase or escalation of
significant findings and an increased incidence of type I
error.
estimator  Statistic that produces a value as a function of
analysis involves the use of point estimation to evaluate
the fit between the estimator (a statistic) and the popula-
tion parameter.
ethical principles  Principles of respect for persons, benef-
icence, and justice relevant to the conduct of research.
ethnographic research  Qualitative research methodology
developed within the discipline of anthropology for inves-
tigating cultures that involves collection, description, and
analysis of data to develop a theory of cultural behavior.
ethnographies  The written reports of a culture from the
perspective of insiders. These reports were initially the
products of anthropologists who studied primitive,
foreign, or remote cultures.
ethnography  A word derived by combining the Greek
roots of ethno (folk or people) and graphos (picture or
portrait).
ethnonursing research  Emerged from Leininger’s theory
of transcultural nursing and focuses mainly on observing
and documenting interactions with people to determine
how daily life conditions and patterns are influencing
human care, health, and nursing care practices.
etic approach  Anthropological research approach of
studying behavior from outside the culture and examining
similarities and differences across cultures.
evaluation step of critical appraisal  Determining the
validity, credibility, significance, and meaning of the
study by examining the links among the study process,
study findings, and previous studies.
evaluation apprehension  Shown when subject’s respons-
es in the experiment are due to apprehension rather than
the effects of the independent variable.
event-partitioning designs  Merger of the longitudinal
and trend designs to increase sample size and avoid the
effects of history on the validity of findings.
694 Glossary
evidence-based practice (EBP)  Conscientious integra-
tion of best research evidence with clinical expertise and
patient values and needs in the delivery of quality, cost-
effective health care.
evidence-based practice centers  Universities and health-
care agencies identified by the Agency for Healthcare
Research and Quality (AHRQ) as centers for the conduct,
communication, and synthesis of research knowledge in
selected areas to promote evidence-based health care.
evidence-based practice guidelines  Rigorous, explicit
clinical guidelines developed on the basis of the best
research evidence available (such as findings from sys-
tematic reviews, metaanalyses, mixed-methods system-
atic reviews, meta-syntheses, and extensive clinical trials);
supported by consensus from recognized national experts
and affirmed by outcomes obtained by clinicians.
exclusion sampling criteria  Sampling requirements iden-
tified by the researcher that prevent an element or subject
from being in a sample.
execution errors  Errors that occur because of a defect in
the data collection procedure.
exempt from review  Studies that have no apparent risks
for the research subjects might be designated as exempt
from review by the chair of the institutional review board
(IRB).
existence statement  Declares that a given concept exists
or that a given relationship occurs.
exogenous variables  Variables within the theoretical
model that are caused by factors from outside the model;
these variables are examined in a study with a model-
testing design.
expedited IRB review  Review process for studies that
have some risks, which are minimal or no greater than
those ordinarily encountered in daily life or during the
performance of routine physical or psychological exami-
nations.
experimental group  Subjects who are exposed to the
experimental treatment or intervention.
experimental research  Objective, systematic, controlled
investigation to examine probability and causality among
selected independent and dependent variables for the
purpose of predicting and controlling phenomena.
experimental study designs  Designs that provide the
greatest amount of control possible to examine causality
more closely.
experimenter expectancy  Researcher’s belief or projec-
tion of the outcome of a study.
explanation  Achieved when research clarifies the rela-
tionships among phenomena and identifies why certain
events occur.
exploratory data analysis  Examining the data descrip-
tively to become as familiar as possible with the nature
of the data and to search for hidden structures and
models.
exploratory-descriptive qualitative research  Research
conducted to address an issue or problem in need 
of a solution and/or understanding using qualitative 
methodology.
exploratory factor analysis  Similar to stepwise regres-
sion, in which the variance of the first factor is partialed
out before analysis is begun on the second factor. It is
performed when the researcher has few prior expectations
about the factor structure.
exploratory regression analysis  Used when the research-
er may not have sufficient information to determine which
independent variables are effective predictors of the de-
pendent variable; thus, many variables may be entered into
the analysis simultaneously. This type is the most com-
monly used regression analysis strategy in nursing studies.
exploratory studies  Designed to increase the knowledge
of a field of study and not intended for generalization to
large populations. Exploratory studies provide the basis
for confirmatory studies.
external criticism  Method of determining the validity of
source materials in historical research that involves
knowing where, when, why, and by whom a document
was written.
external validity  Extent to which study findings can be
generalized beyond the sample used in the study.
extraneous variables  Exist in all studies and can affect
the measurement of study variables and the relationships
among these variables.
F
F statistic  Value or result obtained from conducting a type
of analysis of variance.
fabrication in research  Type of misconduct in research
that involves making up study results and recording or
reporting them.
face validity  Verifies that the instrument looked like or
gave the appearance of measuring the content desired for
a study.
factor  Hypothetical construct created by factor analysis
and includes the clustering of like variables that are
named based on the focus of the clustered variables.
factor analysis  Analysis that examines interrelationships
among large numbers of variables and disentangles those
relationships to identify clusters of variables that are most
closely linked together. Two types of factor analysis are
exploratory and confirmatory.
factor loading  The regression coefficient for the variable
on the factor determined by factor analysis.
factor scores  Variables included in a factor are identified,
and the scores on these variables are summed for each
subject; thus, each subject will have a score for each
factor in the instrument.
factorial analysis of variance  Mathematically, the analy-
sis technique is simply a specialized version of multiple

regression; a number of types of factorial ANOVA have
been developed to analyze data from specific experimen-
tal designs.
factorial design  Study design that includes two or more
different characteristics, treatments, or events that are
independently varied within a study.
fair treatment  Ethical principle that promotes fair selec-
tion and treatment of subjects during the course of a study.
false negative  Result of a diagnostic or screening test that
indicates a disease is not present when it is.
false positive  Result of a diagnostic or screening test that
indicates a disease is present when it is not.
falsification of research  Type of research misconduct that
involves either manipulating research materials, equip-
ment, or processes or changing or omitting data or results
such that the research is not accurately represented in the
research record.
fatigue effect  When a subject becomes tired or bored with
a study; can affect the findings from the study.
feasibility of a study  Determined by examining the time
and money commitment; the researcher’s expertise; avail-
ability of subjects, facility, and equipment; cooperation of
others; and the study’s ethical considerations.
field notes  Notes made during and immediately following
the observations.
field work  Data collection process for a qualitative study.
findings  Translated and interpreted results from a study.
fishing  Conducting multiple statistical analyses of rela-
tionships or differences searching for significant study
findings; can increase the risk for type I error.
fixed-effect model  Model used when each study is esti-
mating the exact same quality
focus groups  Groups that are designed to obtain partici-
pants’ perceptions in a specific (or focused) area in a
setting that is permissive and nonthreatening.
forced choice  Response set for items in a scale that have
an even number of choices, such as four or six, in which
the respondents cannot choose an uncertain or neutral
response and must indicate support for or against the topic
measured.
forest plots  Particular figures that are very important in
the presentation of results in meta-analysis reports that
synthesize the results of many studies.
framework  The abstract, logical structure of meaning that
guides development of the study and enables the researcher
to link the findings to the body of knowledge for nursing.
frequency distribution  Statistical procedure that involves
listing all possible measures of a variable and tallying
each datum on the listing. The two types of frequency
distributions are ungrouped and grouped.
frequency table  A way of organizing the data by listing
every possible value in the first column of numbers and
the frequency (tally) of each value in the second column
of numbers.
Glossary 695
funnel plot  Provides graphic representations of possible
effect sizes (ESs) or odds ratios (ORs) for interventions
in selected studies
G
general proposition  Highly abstract statement of the rela-
tionship between two or more concepts that is found in a
conceptual model.
generalization  Extends the implications of the findings
from the sample that was studied to the larger population
or from the situation studied to a larger situation.
geographical analyses  Used to examine variations in
health status, health services, patterns of care, or patterns
of use by geographical area.
going native  In ethnographic research, when the researcher
becomes part of the culture and loses all objectivity and,
with it, the ability to observe clearly.
gold standard  Most accurate means of currently diagnos-
ing a particular disease; serves as a basis for comparison
with newly developed diagnostic or screening tests.
government report  Document with facts that can be used
to develop the significance and background section of a
research proposal or report.
grant  Research funding from private or public institutions
to support the conduct of a study.
grantsmanship  Expertise and skill in successfully devel-
oping proposals to obtain funding for selected studies.
grey literature  Studies that have limited distributions,
such as theses and dissertations, unpublished research
reports, articles in obscure journals, some online jour-
nals, conference papers and abstracts, conference pro-
ceedings, research reports to funding agencies, and
technical reports.
grounded theory research  Qualitative, inductive research
technique based on symbolic interaction theory that is
conducted to discover what problems exist in a social
scene and the processes people use to handle them. The
research process involves formulation, testing, and rede-
velopment of propositions until a theory is developed.
grouped frequency distribution  Presents a count of
patient characteristics that are divided by subsets. For
example, instead of providing number of subjects for all
ages, provides number of subjects from ages 20 to 29 and
30 to 39.
Grove Model for Implementing Evidence-Based Guide-
lines in Practice  Model developed by one of the text-
book authors (Grove) to promote the use of national,
standardized evidence-based guidelines in clinical
practice.
H
Hawthorne effect  Psychological response in which sub-
jects change their behavior simply because they are sub-
jects in a study, not because of the research treatment.
696 Glossary
hazard  A neutral word intended to describe the risk of
event occurrence.
hazard ratio (HR)  Interpreted almost identically to an
odds ratio (OR), with the exception that the HR represents
the risk of the event’s occurring sooner.
heterogeneity  Researcher’s attempt to obtain subjects
with a wide variety of characteristics to reduce the risk of
bias in studies not using random sampling.
hierarchical statement sets  Composed of a general prop-
osition, specific proposition, and hypothesis or research
question, moving from abstract statements to more con-
crete statements proposition and a hypothesis or research
question.
highly controlled settings  Artificially constructed envi-
ronments that are developed for the sole purpose of con-
ducting research, such as laboratories, experimental
centers, and research medical units.
HIPAA Privacy Rule  Federal regulations implemented in
2003 to protect an individual’s health information. The
HIPAA Privacy Rule affects not only the healthcare envi-
ronment but also the research conducted in this
environment.
historical research  Qualitative research method that
includes a narrative description or analysis of events that
occurred in the remote or recent past.
history  An event that is not related to the planned study
but that occurs during the time of the study.
history effect  Event that is not related to the planned study
but occurs during the time of the study and could influ-
ence the responses of subjects to the treatment.
homogeneity  Degree to which objects are similar or a
form of equivalence, such as limiting subjects to only one
level of an extraneous variable to reduce its impact on the
study findings.
homogeneity reliability  Type of reliability testing used
with paper-and-pencil tests that addresses the correlation
of various items within the instrument. Also referred to as
internal consistency reliability.
homoscedastic  Data are evenly dispersed both above and
below the regression line, which indicates a linear rela-
tionship on a scatter diagram (plot).
horizontal axis  The x axis in a scatterplot or graph of a
regression line.
human rights  Claims and demands that have been justi-
fied in the eyes of an individual or by the consensus of a
group of individuals and are protected in research.
hypothesis  Formal statement of the expected relationship(s)
between two or more variables in a specified population.
hypothesis guessing  Occurs when subjects within a study
guess the hypothesis of the researcher.
hypothetical population  A population that cannot be
defined according to sampling theory rules, which require
a list of all members of the population.
I
immersed  Being fully invested in the data and spending
extensive amounts of time reading and thinking about the
data.
immersion in the data  Initial phase of qualitative data
analysis when researchers become very familiar with the
data by reading and rereading notes and transcripts,
recalling observations and experiences, listening to audio
tapes, and viewing videos.
implications of research findings for nursing  Meaning
of research conclusions for the body of knowledge,
theory, and practice in nursing.
inclusion sampling criteria  Sampling requirements iden-
tified by the researcher that must be present for the
element or subject to be included in the sample.
incomplete disclosure  Subjects are not completely
informed about the purpose of a study because that
knowledge might alter the subjects’ actions. After the
study, the subjects must be debriefed about the complete
purpose of the study and the findings.
independent groups  Groups in which the selection of one
subject is totally unrelated to the selection of other sub-
jects. An example is when subjects are randomly selected
and assigned to the treatment and control groups.
independent samples t-test  The most common paramet-
ric analysis technique used in nursing studies to test for
significant differences between two independent samples.
independent variable  Treatment, intervention, or exper-
imental activity that is manipulated or varied by the
researcher to create an effect on the dependent
variable.
indirect costs  Expenses related to a research project but
not specifically part of the implementation of the steps of
the study, such as administrative costs. Grants that fund
indirect costs provide researchers greater freedom to
conduct studies.
indirect measurement  Used with abstract concepts; the
concepts are not measured directly, but instead, indicators
or attributes of the concepts are used to represent the
abstraction, such as a perception of pain scale to measure
chronic pain.
individually identifiable health information (IIHI)  Any
information collected from an individual, including
demographic information, that is created or received by
healthcare providers, a health plan, or a healthcare clear-
inghouse, that is related to the past, present, or future
physical or mental health or condition of an individual,
and that identifies the individual.
inductive reasoning  Reasoning from the specific to the
general in which particular instances are observed and
then combined into a larger whole or general statement.
inference  Use of inductive reasoning to move from a spe-
cific case to a general truth. Thus, statistics are used to

infer from the specific study results to a general statement
about the larger population.
inferential statistics  Statistics designed to allow inference
from a sample statistic to a population parameter; com-
monly used to test hypotheses of similarities and differ-
ences in subsets of the sample under study.
inferred causality  Cause-and-effect relationship is identi-
fied from numerous studies conducted over time to deter-
mine risk factors or causal factors in selected situations.
informants  People interviewed to learn more about the
meaning of observations.
information-rich cases  Cases selected during the purpo-
sive sampling process from which qualitative researchers
can learn a great deal about the central focus of their
study.
informed consent  Prospective subject’s agreement to
voluntarily participate in a study, which is reached after
the subject assimilates essential information about the
study.
inherent variability  Data can be naturally expected to
have a few random observations included in the extreme
ends of the tails.
institutional review  Process of examining studies for
ethical concerns by a committee of peers to determine if
the study can be conducted in a selected agency.
institutional review board (IRB)  Committee that reviews
research to ensure that the investigator is conducting the
research ethically. Universities, hospital corporations, and
many managed care centers have IRBs to promote the
conduct of ethical research and to protect the rights of
prospective subjects at their institutions.
instrumentation  A component of measurement that
involves the application of specific rules to develop a
measurement device or instrument.
integration  Making connections between ideas, theories,
and experience.
intention to treat  An analysis based on the principle that
participant data are analyzed according to the groups into
which they were randomly assigned regardless of what
happens to them in the study.
interaction effects  Influence on the design validity by the
interaction of different facets of the study, such as selec-
tion of subjects and treatment, setting and treatment, or
history and treatment.
interaction of different treatments  A threat to construct
validity; occurs when subjects receive more than one
treatment in a study.
intercept  The point where the regression line crosses (or
intercepts) the y axis; is represented by the letter a.
intermediate end points  Events or markers that act as
precursors to the final outcome.
intermediate outcome  Mediating variables in theory-
guided intervention research.
Glossary 697
intermediate mediation  Level of casual assertion that
considers causal factors operating between molar and
micro levels.
internal consistency  Maintaining the integrity of the dif-
ferent steps of the research process, such as sample selec-
tion, measurement of study variables, implementation of
study treatment, and data collection.
internal consistency reliability  See homogeneity reli-
ability.
internal criticism  Involves examination of the reliability
of historical documents.
internal validity  Extent to which the effects detected in
the study are a true reflection of reality rather than being
the result of the effects of extraneous variables.
interpretation  Process of information that occurs in the
mind of the reader.
interpretation of research outcomes  Involves examining
the results of data analysis, forming conclusions, consid-
ering the implications for nursing, exploring the signifi-
cance of the findings, generalizing the findings, and
suggesting further studies.
interrater reliability  Degree of consistency between two
raters who are independently assigning ratings to a vari-
able or attribute being investigated.
interrupted time-series designs  Designs similar to
descriptive time designs except that a treatment is applied
at some point in the observations.
interval data  Numerical information collected during a
study that has equal distances between a continuum of
values and the data also follows the rules of mutually
exclusive categories, exhaustive categories, and rank
ordering, which influences the type of statistical analyses
that can be conducted.
interval estimate  Researcher identifies a range of values
on a number line where the population parameter is
thought to be.
interval level of measurement  Interval scales have equal
numerical distances between intervals or values of the
scale in addition to following rules of mutually exclusive
categories, exhaustive categories, and rank ordering, such
as temperature.
intervening variable  Mediating variable that can
affect the occurrence, strength, or direction of a
relationship.
intervention fidelity  Reliable and competent implementa-
tion of an experimental treatment that includes two core
components: (1) adherence to the delivery of the pre-
scribed treatment behaviors, session, or course, and 
(2) competence in the researcher or interventionalist’s
skillfulness in delivery of the intervention.
intervention research  Methodology for investigating the
effectiveness of a nursing intervention in achieving the
desired outcome or outcomes in a natural setting.
698 Glossary
intervention taxonomy  An organized categorization of
interventions performed by nurses.
intervention theory  This theory includes a careful
description of the problem that the intervention will
address, intervening actions that must be implemented to
address the problem, moderating variables that might
change the impact of the intervention, mediating variables
that might alter the effect of the intervention, and expected
outcomes of the intervention.
interventionist  In intervention research, a person who
has been formally prepared to provide a particular inter-
vention and is accountable for the fidelity of the interven-
tion.
interventions  Treatments, therapies, procedures, or ac-
tions that are implemented by researchers to determine
their outcomes in a study, and if effective, are imple-
mented by healthcare professionals to and with patients,
in a particular situation, to move the patients’ conditions
toward desired health outcomes that are beneficial to  kurtosis  Degree of peakedness (platykurtic, mesokurtic,
them.
intervention reliability  Ensures that the research treat-
ment is standardized and applied consistently each time
it is administered in a study.
interview  Structured or unstructured verbal communica-
tion between the researcher and subject during which
information is obtained for a study.
introspection  Process of turning your attention inward
toward your own thoughts to provide increased awareness
and understanding of the flow and interplay of feelings
and ideas.
intuition  Insight or understanding of a situation or event
as a whole that usually cannot be logically explained.
intuitive comprehension  Reader recognizes the findings
of a qualitative study as being a credible representation
of reality.
invasion of privacy  When private information is shared
without an individual’s knowledge or against his or her
will.
investigator-initiated research proposal  Researcher
identifies a significant problem, develops a study to
examine it, and submits a proposal for the study to the
appropriate federal funding agency.
inverse linear relationship  Indicates that as one variable
or concept changes, the other variable or concept changes
in the opposite direction; also referred to as a negative
linear relationship.
Iowa Model of Evidence-Based Practice  Model devel-
oped in 1994 and revised in 2001 by Titler and colleagues
to promote evidence-based practice in clinical agencies.
J level of significance  See alpha (α).
justice, principle of  States that human subjects should be
treated fairly.
K
key informants  Participants in qualitative studies who
provide quality information during the conduct of the
study.
keywords  Major concepts or variables that must be
included in your literature search. Keywords or terms can
be identified by determining the concepts relevant to your
study, the populations of particular interest in your study,
interventions to be implemented, and measurement
methods to be used in the study, or possible outcomes for
the study.
knowledge  Essential content or body of information for a
discipline that is acquired through traditions, authority,
borrowing, trial and error, personal experience, role-mod-
eling and mentorship, intuition, reasoning, and research.
Kolmogorov-Smirnov two-sample test  Nonparametric
test used to determine whether two independent samples
have been drawn from the same population.
or leptokurtic) of the curve shape that is related to the
spread or variance of scores.
L
landmark studies  Major projects that generate knowl-
edge that influences a discipline and sometimes society in
general and marks an important stage of development or
a turning point in a field of research.
language bias  Can occur if searches focus just on
studies in English and important studies exist in other
languages.
latent transition analysis (LTA)  Outcomes research strat-
egy used in situations in which stages or categories of
recovery have been defined and transitions across stages
can be identified. To use this analysis method, each
member of the population is placed in a single category
or stage for a given point of time.
least-squares  The fact that when deviations from the
mean are squared, the sum is smaller than the sum of
squared deviations from any other value in a sampling
distribution.
least-squares regression line  A line drawn through a scat-
terplot that represents the smallest deviation of each value
from the line.
legally authorized representative  Individual or other
body authorized under applicable law to consent on behalf
of a prospective subject to the subject’s participation in
the procedures involved in the research.
leptokurtic  Term used to describe an extremely peaked-
shape distribution of a curve, which means that the scores
in the distribution are similar and have limited variance.
levels of measurement  The rules for assigning numbers
to objects so that a hierarchy in measurement was

established, and the levels of measurement from lower to
higher are nominal, ordinal, interval, and ratio.
Likert scale  Instrument designed to determine the opinion
or attitude of a subject; it contains a number of declarative
statements with a scale after each statement.
limitations  Theoretical and methodological restrictions or
weaknesses in a study that may decrease the generaliz-
ability of the findings.
line graphs  Figures or illustrations that are used to repre-
sent the results from studies. Often line graphs are used
to show changes in different groups over time.
line of best fit  The use of a regression equation to develop
the line that allows the highest degree of prediction  measurement error  Difference between what exists in
possible in a predictive correlational study, in which  reality and what is measured by a research instrument.
independent variables are used to predict the dependent
variable.
linear relationship  Relationship between two variables or
concepts will remain consistent regardless of the values
of each of the variables or concepts.
literature review  See review of relevant literature.
location bias  Can occur if studies are published in lower-
impact journals and indexed in less searched databases.
logic  Science that involves valid ways of relating ideas to
promote human understanding; includes abstract and con-
crete thinking and logistic, inductive, and deductive
reasoning.
logical positivism  Branch of philosophy that operates on
strict rules of logic, truth, laws, axioms, and predictions.
Quantitative research emerged from logical positivism.
logistic reasoning  Used to break the whole into parts that
can be carefully examined, as can the relationships among
the parts; one can understand the whole by examining the
parts.
longitudinal designs  Panel designs used to examine
changes in the same subjects over an extended period.
low statistical power  When the strength or power of a
study to detect relationships between variables of differ-
ences between groups is below the acceptable standard
power (0.8) needed to conduct a study. Low statistical
power increases the likelihood of a type II error.
M meta-analysis  Involves the statistical pooling of the
manipulation  Implementation or controlled movement of
a treatment or an independent variable in a study to deter-
mine its effect on the study dependent variable.
Mann-Whitney U test  Used to analyze ordinal data with
95% of the power of the t-test to detect differences
between groups of normally distributed populations.
matching  Technique used when an experimental subject
is randomly selected and a subject similar in relation to
important extraneous variables is randomly selected for
inclusion in the control or comparison group. This process
results in dependent or related groups.
Glossary 699
maturation  Unplanned and unrecognized changes experi-
enced during a study, such as subjects growing older,
wiser, stronger, hungrier, or more tired, that can influence
the findings of a study.
mean  Value obtained by summing all the scores and divid-
ing that total by the number of scores being summed.
mean deviation  The average difference score, using the
absolute values.
mean difference  A standard statistic that is calculated to
determine the absolute difference between two groups.
measurement  Process of assigning numbers to objects,
events, or situations in accord with some rule.
measures of central tendency  Statistical procedures
(mode, median, and mean) for determining the center of
a distribution of scores.
measures of dispersion  Statistical procedures (range, dif-
ference scores, sum of squares, variance, and standard
deviation) for examining how scores vary or are dispersed
around the mean.
median  Score at the exact center of the ungrouped fre-
quency distribution.
mediator variables  Variables that bring about the effects
of the intervention after it has occurred and thus influence
the outcomes of an intervention study.
memo  Developed by the researcher to record insights or
ideas related to notes, transcripts, or codes during qualita-
tive data analysis.
mentee  The protégé who is guided by a mentor in the
mentorship process.
mentor  Someone who serves as a teacher, sponsor, guide,
exemplar, or counselor for a novice or protégé. For
example, an expert nurse serves as a guide or role model
for a novice nurse or mentee.
mentorship  Intense form of role-modeling in which an
expert nurse serves as a teacher, sponsor, guide, exemplar,
or counselor for a novice nurse.
mesokurtic  Term that describes a normal curve with an
intermediate degree of kurtosis and intermediate variance
of scores.
results from several previous studies into a single quanti-
tative analysis that provides one of the highest levels of
evidence for an intervention’s efficacy.
metasummary, qualitative  Synthesis or summing of
findings across qualitative reports to determine the current
knowledge in an area.
meta-synthesis, qualitative  Synthesis of qualitative stud­
ies that provides a fully integrated, novel description or
explanation of a target event or experience versus a sum­
mary view of that event or experience. Meta-synthesis
requires more complex, integrative thought in developing
700 Glossary
a new perspective or theory based on the findings of previ-
ous qualitative studies.
method of least squares  Procedure in regression analysis
for developing the line of best fit.
methodological designs  Used to develop the validity and
reliability of instruments to measure research concepts
and variables.
methodological limitations  Restrictions or weaknesses in
the study design that limit the credibility of the findings
and the population to which the findings can be
generalized.
metric ordinal scale  Scale that has unequal intervals; its
use to collect data during a study results in ordinal data.
micromediation  Examines causal connections at the level
of small particles such as atoms; is part of multicausality
theory.
middle-range theories  Theories that are less abstract and
address more specific phenomena than grand theories,
that are directly applicable to practice, and that focus on
explanation and implementation.
minimal risk  Risk of harm anticipated in the proposed
research is not greater, with regard to probability and
magnitude, than that ordinarily encountered in daily life
or during the performance of routine physical or psycho-
logical examinations.
mixed-method approach  Offers investigators the oppor-
tunity to utilize the strengths of both qualitative and quan-
titative research designs in the conduct of a study.
mixed-method systematic review  Synthesis of any of a
variety of study designs, such as qualitative research and
quasi-experimental, correlational, and/or descriptive
study, to determine the current knowledge in an area.
mixed results  Study results including a combination of
significant and nonsignificant outcomes, which are prob-
ably the most common results of studies.
modal percentage  Appropriate for nominal data; indi-
cates the relationship of the number of data scores repre-
sented by the mode with the total number of data scores.
mode  Numerical value or score that occurs with the great-
est frequency in a distribution; however, it does not neces-
sarily indicate the center of the data set.
model-testing designs  Used to test the accuracy of a
hypothesized causal model or map.
moderator variable  Variable that occurs with the inter-
vention (independent variable) and alters the causal rela-
tionship between the intervention and outcomes. It
includes characteristics of the subjects and the person
implementing the intervention.
molar  Causal laws related to large and complex objects
that are part of the theory of multicausality.
monographs  Books, booklets of conference proceedings,
or pamphlets, which are usually written once and may be
updated with a new edition.
monomethod bias  More than one measure of a variable
is used in a study, but all measures use the same method
of recording, such as using two paper-and-pencil scales
to measure depression in a study.
mono-operation bias  Occurs when only one method of
measurement is used to measure a variable or concept in
a study, such as the use of one paper-and-pencil scale to
measure chronic pain.
multicausality  Recognition that a number of interrelat-
ing variables can be involved in having a particular
effect.
multicollinearity  Occurs when the independent variables
in a regression equation are strongly correlated.
multidimensional scaling  A measurement method that
was developed to examine many aspects or elements of a
concept or variable.
multilevel analysis  Used in epidemiology to study how
environmental factors and individual attributes and
behavior interact to influence individual-level health
behavior and disease risk.
multilevel synthesis  Involves synthesizing the findings
from quantitative studies separate from qualitative studies
and then integrating the findings from these two syntheses
in the final report.
multimethod-multitrait technique  When a variety of
data collection methods, such as interview and observa-
tion, are used and different measurement methods are
used for each concept in a study.
multitrait-multimethod  The approach of combining con-
vergent and divergent validity testing of instruments.
multimodal  A distribution of scores that has more than
two modes or most frequently occurring scores.
multiple regression analysis  Extension of simple linear
regression with more than one independent variable
entered into the analysis.
multistage cluster sampling  Type of cluster sampling
when the random selection of the sample continues
through several stages.
N
narrative analysis  Qualitative means of formally analyz-
ing text including stories.
natural settings  Field settings or uncontrolled, real-life
settings where research is conducted, such as subjects’
homes, worksites, or schools.
necessary relationship  One variable or concept must
occur for the second variable or concept to occur.
negative likelihood ratio  Ratio of true-negative results
to false-negative results; is calculated as follows: 
Negative likelihood ratio = 100% − Sensitivity ÷ Speci­
ficity.
negative linear relationship  See inverse linear relation-
ship.

negative results  Unpredicted nonsignificant or inclusive
results from a study that are often the most difficult to
explain.
negatively skewed  Data are not normally distributed in a
study; characterized by a curve in which the largest
portion of data is above the mean.
nested design  Design that allows the researcher to con-
sider the effect of variables that are found only at some
levels of the independent variables being studied.
nested variables  Variables found only at certain levels of
the independent variable, such as gender, race, socioeco-
nomic status, and education.
naturalistic inquiry  encompasses studies designed to
study people and situations in their natural states.
network sampling  Nonprobability sampling method that
includes a snowballing technique that takes advantage of
social networks and the fact that friends tend to hold
characteristics in common. Subjects meeting the sample
criteria are asked to assist in locating others with similar
characteristics.
networking  Process of developing channels of communi-
cation between people with common research interests
throughout the country.
nominal data  Lowest level of data that can only be orga-
nized into categories that are exclusive and exhaustive,
but the categories cannot be compared or rank ordered
and can only be analyzed by the lowest level of statistical
analyses.
nominal level of measurement  Lowest level of measure-
ment that is used when data can be organized into catego-
ries that are exclusive and exhaustive, but the categories
cannot be compared or rank ordered, such as gender, race,
marital status, and diagnoses.
noncoercive disclaimer  A statement that participation is
voluntary and refusal to participate will involve no penalty
or loss of benefits to which the subject is entitled.
nondirectional hypothesis  States that a relationship
exists but does not predict the exact nature of the
relationship.
nonequivalent control group designs  Designs in which
the control group is not selected by random means, such
as the one-group posttest-only design, posttest-only
design with nonequivalent groups, and one-group pretest-
posttest design.
nonparametric statistical analysis  Statistical techniques
used when the assumptions of parametric statistics are not
met, and most commonly used to analyze nominal- and
ordinal-level data.
nonprobability sampling  Nonrandom sampling in which
not every element of the population has an opportunity
for selection in the sample, such as convenience (acciden-
tal) sampling, quota sampling, purposive sampling, and
network sampling.
Glossary 701
nonsignificant results  Negative results or results contrary
to the researcher’s hypotheses that can be an accurate
reflection of reality or can be caused by study weaknesses.
Also see negative results.
nontherapeutic research  Research conducted to generate
knowledge for a discipline and in which the results from
the study might benefit future patients but will probably
not benefit those acting as research subjects.
normal curve  A symmetrical, unimodal bell-shaped curve
that is a theoretical distribution of all possible scores, but
no real distribution exactly fits the normal curve.
normally distributed  Data points that follow the spread
or distribution of a normal curve.
norm-referenced testing  Test performance standards that
have been carefully developed over years with large, rep-
resentative samples through the use of standardized tests
with extensive reliability and validity.
null hypothesis  States that there is no relationship between
the variables being studied; a statistical hypothesis used
for statistical testing and interpreting statistical outcomes.
number needed to treat  A metric that is defined as the
number of patients who would have to be treated with the
new intervention to avoid one event that might have
occurred with standard treatment.
Nuremberg Code  Ethical code of conduct to guide inves-
tigators when conducting research.
numerator  The number on the top part of a fraction.
Nursing Care Report Card  Evaluation of hospital
nursing care using 10 indicators (2 structure indicators, 2
process indicators, and 6 outcome indicators). This report
card could facilitate benchmarking or setting a desired
standard that would allow comparisons of hospitals in
terms of their nursing care quality.
nursing interventions  Deliberative cognitive, physical,
or verbal activities performed with or on behalf of indi-
viduals and their families that are directed toward accom-
plishing particular therapeutic objectives relative to
individuals’ health and well-being.
nursing research  Scientific process that validates and
refines existing knowledge and generates new knowledge
that directly and indirectly influences the delivery of
evidence-based nursing practice.
nursing-sensitive patient outcomes  Patient outcomes
that are influenced by or associated with nursing.
O
oblique rotation  Type of rotation in factor analysis used
to accomplish the best fit (best-factor solution) and in
which the factors are allowed to be correlated.
observation  Collection of data through listening, smelling,
touching, and seeing, with an emphasis on what is seen.
observational checklist  Form used to record whether a
behavior occurred.
702 Glossary
observational measurement  Use of structured and
unstructured observation to measure study variables.
observed level of significance  The actual level of signifi-
cance that is achieved or observed in a study.
observed score  Actual score or value obtained for a
subject on a measurement tool.
observer as participant  Researcher’s time is spent mainly
observing and interviewing subjects and less in the par-
ticipation role.
odds ratio (OR)  The ratio of the odds of an event occur-
ring in one group, such as the treatment group, with the
odds of it occurring in another group, such as the standard
care group.
one-group posttest-only design  Preexperimental design
with numerous threats to validity that is inadequate for
making causal references.
one-group pretest-posttest design  Quasi-experimental
design in which the pretest scores serve as the comparison
group and the posttest scores after the treatment serve as
the experimental group.
one-tailed test of significance  Analysis used with direc-
tional hypotheses in which extreme statistical values of
interest are thought to occur in a single tail of the curve.
one-way chi-square  A statistic that compares different
levels of one variable only measured at the nominal level.
open-label extension  A term that usually applies to ran-
domized controlled trials or other clinical trials whereby
participants in the active treatment group are offered an
opportunity to continue treatment and those in the placebo
or control group are given the option to transition to the
active treatment at the conclusion of the intervention
period.
operational definition  Description of how variables or
concepts will be measured or manipulated in a study.
operational reasoning  Involves identification and dis-
crimination among many alternatives or viewpoints and
focuses on the process of debating alternatives.
operationalizing a variable or concept  Development of
the conceptual definition of a concept or variable to link
it to the study framework and the operational definition
of a concept or variable so it can be measured or manipu-
lated in a study.
operator  Permits grouping of ideas, selection of places to
search in a database record, and ways to show relation-
ships within a database record, sentence, or paragraph.
The most common operators are Boolean, locational, and
positional.
operator, Boolean  The three words AND, OR, and NOT
are used with the researcher’s identified concepts in con-
ducting searches of databases.
operator, locational  Search operator that identifies terms
in specific areas or fields of a record, such as article title,
author, and journal name.
operator, positional  Search operator used to look for
requested terms within certain distance of one another.
Common positional operators are NEAR, WITH, and
ADJ.
ordinal data  Data that can be ranked, but the intervals
between the ranked data are not necessarily equal, such
as military ranks. This level of data is analyzed by non-
parametric statistical techniques.
ordinal level measurement  Measurement that yields data
that can be ranked, but the intervals between the ranked
data are not necessarily equal, such as levels of coping.
outcomes of care  The dependent variables or clinical
results of health care that are measured to determine the
impact of the process of care management techniques.
The outcomes from the Medical Outcomes Study Frame-
work include clinical end points, functional status, general
well-being, and satisfaction with care.
outcome reporting bias  Occurs when study results are
not reported clearly and with complete accuracy.
outcomes research  Important scientific methodology that
was developed to examine the end results of patient care.
The strategies used in outcomes research are a departure
from traditional scientific endeavors and incorporate eval-
uation research, epidemiology, and economic theory
perspectives.
outliers  Extreme scores or values in a set of data that are
exceptions to the overall findings.
out-of-pocket costs  Those expenses incurred by the
patient, family, or both that are not reimbursable by the
insurance company; they might include costs of buying
supplies, dressings, selected medications, or special foods.
P
paired or dependent samples  Samples that are related or
matched in some way. See dependent groups.
paradigm  Particular way of viewing a phenomenon in the
world.
parallel-forms reliability  See alternate-forms reliability.
parallel synthesis  Involves the separate synthesis of
quantitative and qualitative studies, but the findings from
the qualitative synthesis are used in interpreting the syn-
thesized quantitative studies.
parameter  Measure or numerical value of a population.
parametric statistical analyses  Statistical techniques
used when three assumptions are met: (1) the sample was
drawn from a population for which the variance can be
calculated, and the distribution is expected to be normal
or approximately normal, (2) the level of measurement
should be interval or ratio with an approximately normal
distribution, and (3) the data can be treated as though they
were obtained from random samples.
paraphrasing  Involves expressing clearly and concisely
the ideas of an author in your own words.

partially controlled setting  Environment that the re-
searcher manipulates or modifies in some way when it is
used as a setting for a study.
participant observervation  Special form of observation
in which researchers immerse themselves in the setting so
they can hear, see, and experience the reality as the par-
ticipants do. However, the participants are aware of the
dual roles of the researcher (participant and observer).
participants  Individuals who participate in qualitative
and quantitative research; also referred to as subjects in
quantitative research.
path coefficient  Effect of the independent variable on the
dependent variable that is determined through path
analysis.
patient  Someone who has already gained access to care.
pattern  Analysis of qualitative data to determine the
trends and links among the facets of the data that can
become the meaningful findings from the study.
Pearson’s product-moment correlation coefficient (r)  to develop or refine the methodology, such as the treat-
Parametric test used to determine the relationship between
two variables.
percentage distribution  Indicates the percentage of the
sample with scores falling within a specific group or
range.
percentage of variance  Amount of variability explained
by a linear relationship; the value is obtained by squaring
Pearson’s correlation coefficient (r). For example, if an r
= 0.5 in a study, the percentage of variance explained is
r2 = 0.25, or 25%.
periodicals  Subset of serials with predictable publication
dates, such as journals that are published over time and
are numbered sequentially for the years published.
permission to participate in a study  Agreement of
parents or guardians to the participation of their child or
ward in research.
person  Someone who may or may not have gained access
to care.
personal experience  Gaining of knowledge by being
personally involved in an event, situation, or circum-
stance.
phenomena  experiences that cannot be explained by
examining causal relations but need to be studied as the
very things they are.
phenomenological research  Inductive, descriptive quali-
tative methodology developed from phenomenological
philosophy for the purpose of describing experiences as
they are lived by the study participants.
phenomenological researcher  Researcher who focuses
on describing experiences as they are lived by the study
participants rather than large samples with generalizable
findings.
philosophy  Broad, global explanation of the world that
gives meaning to the world of nursing and provides a
Glossary 703
framework within which thinking, knowing, with doing
occur.
physiological measures  Techniques used to measure
physiological variables either directly or indirectly, such
as techniques to measure heart rate or mean arterial
pressure.
philosophical perspective  The worldview of the re-
searcher that guides the questions asked and the methods
selected for conducting a specific study.
PICOS Format  An abbreviation that stands for Popula-
tion or participants of interest; Intervention needed for
practice; Comparisons of the intervention with control,
placebo, standard care, variations of the same interven-
tion, or different therapies; Outcomes needed for practice;
and Study design. PICOS is one of the most common
formats used to develop a relevant clinical question to
guide a systematic review of research.
pilot study  Smaller version of a proposed study conducted
ment, instrument, or data collection process.
placebo  An intervention intended to have no effect;
however, it is like the test intervention or is experienced
like the real study intervention to account for how study
participants would respond without actually receiving the
real intervention.
plagiarism  Type of research misconduct that involves the
appropriation of another person’s ideas, processes, results,
or words without giving appropriate credit, including
those obtained through confidential review of others’
research proposals and manuscripts.
platykurtic  Term that indicates a relatively flat curve with
the scores having large variance among them.
population  All elements (individuals, objects, events, or
substances) that meet the sample criteria for inclusion
in a study; sometimes referred to as a target popula-
tion.
population-based studies  Important type of outcomes
research that involves studying health conditions in the
context of the community rather than the context of the
medical system.
population parameter  A true but unknown numerical
characteristic of a population. Parameters of the popula-
tion are estimated with statistics.
population studies  Studies that target the entire popula-
tion.
position papers  Papers disseminated by professional
organizations and government agencies to promote a par-
ticular viewpoint on a debatable issue.
positive likelihood ratio  Likelihood ratio calculated to
determine the likelihood that a positive test result 
is a true positive and a negative test result is a true 
negative. Positive Likelihood Ratio = Sensitivity ÷ (100%
− Specificity).
704 Glossary
positive linear relationship  Indicates that as one variable
changes (value of the variable increases or decreases),  outcome in a given situation that can be achieved through
the second variable will also change in the same
direction.
positively skewed  Data from a study having a non-normal
distribution that is characterized by a curve that has the
largest portion of data below the mean.
post hoc tests  Statistical tests developed specifically to
determine the location of differences in studies with more
than two groups, such as when ANOVA results are sig-
nificant in a study that has three or more groups. Fre-
quently used post hoc tests are Bonferroni’s procedure,
the Newman-Keuls test, the Tukey HSD test, the Scheffé
test, and Dunnett’s test.
poster session  Visual presentation of a study by using
pictures, tables, and illustrations on a display board.
posttest-only design with comparison group  Preexperi-
mental design conducted to examine the difference
between the experimental group that receives a treatment
and the comparison group that does not.
power  Probability that a statistical test will detect a sig-
nificant difference or relationship that exists, which is the
capacity to correctly reject a null hypothesis. Standard
power of 0.8 is used to conduct power analysis to deter-
mine the sample size for a study.
power analysis  Used to determine the risk of a type II
error so that the study can be modified to decrease the
risk, if necessary. Conducting a power analysis involves
alpha, effect size, and standard power of 0.8 to determine
the sample size for a study. Power analysis is also con-
ducted when nonsignficant results are obtained to deter-
mine the power of the analysis conducted.
practical or clinical significance  Associated with its
importance to the body of knowledge that applies to
nursing; sometimes referred to as clinical significance.
practice effect  Occurs when subjects improve as they
become more familiar with the experimental protocol.
practice pattern  Usual care that is provided on a unit or
in a setting.
practice pattern profiling  Epidemiological technique
used in outcomes research that focuses on patterns of care
rather than individual occurrences of care.
practice styles  Particular ways to implement health care
that affect health outcomes and are examined in outcomes
research. Practice styles are part of the construct process
of care from Donabedian’s theory of health care.
practice theories  Another name for middle-range theories
that are usefulness in practice.
precision  Accuracy with which the population parameters
have been estimated within a study. Also used to describe
the degree of consistency or reproducibility of measure-
ments with physiological instruments.
prediction  Ability to estimate the probability of a specific
research.
prediction equation  Outcome of regression analysis
whereby a formula or equation is developed to predict a
dependent variable.
predictive design  Developed to predict the value of the
dependent variable on the basis of values obtained from
the independent variables; one approach to examining
causal relationships between variables.
predictive validity  Type of construct instrument validity
in which future performance or attitudes are proposed or
predicted on the basis of an instrument’s scores—for
example, measuring health-related behaviors with an
instrument to predict future health status of indivi­duals.
premise  Statement that identifies the proposed relation-
ship between two or more variables or concepts.
preproposal  Short document (four to five pages plus
appendices) written to explore the funding possibilities
for a research project.
prescriptive theory  Specifies what must be done to
achieve the desired effects, including (1) the components,
intensity, and duration required, (2) the human and mate-
rial resources needed, and (3) the procedures to be fol-
lowed to produce the desired outcomes.
presentation  The sharing of research findings verbally by
delivering a research report and responding to questions,
or by displaying a poster of a study at a conference or
meeting.
pretest and posttest design with a comparison group 
Type of quasi-experimental design frequently imple-
mented to determine the effect of a treatment by compar-
ing the experimental group (treatment group) with the
comparison group.
pretest-posttest control group design  Classic experi-
mental design in which two randomized groups—one
receiving the experimental treatment and one receiving no
treatment, a placebo treatment, or usual or standard
care—are examined for differences to determine the
impact of a treatment.
primary prevention studies  Specially designed studies
that attempt to measure things that do not happen. Changes
in a community are examined and inferred to be a conse-
quence of the effectiveness of the prevention program
(treatment).
primary source  Source that is written by the person who
originated or is responsible for generating the ideas
published.
primary theoretical source  Source that is written by the
theorist who developed the theory or conceptual content.
primordial cell  Concept from Donabedian’s framework
that is the physical-physiological function of the

individual patient being cared for by the individual
practitioner.
principal component analysis  Second step in exploratory
factor analysis that provides preliminary information that
the researcher needs so that decisions can be made before
the final factoring.
principal investigator (PI)  In a research grant, the indi-
vidual who will have primary responsibility for adminis-
tering the grant and interacting with the funding agency.
privacy  The freedom an individual has to determine the
time, extent, and general circumstances under which
private information will be shared with or withheld from
others.
probability  Addresses relative, rather than absolute, cau-
sality. Probability of a relationship and significance of
interventions are examined in research.
probability distributions  Distributions of values for dif-
ferent statistical analysis techniques, such as tables of r
values for Pearson Product Moment Correlation, t values
for t-test, or F values for analysis of variance. Some
common probability distribution tables are found in the
appendices of this text.
probability sampling method  Random sampling tech-
niques in which each member (element) in the population
should have a greater than zero opportunity to be selected
for the sample; examples are simple random sampling,
stratified random sampling, cluster sampling, and system-
atic sampling.
probability statement  Expresses the likelihood that
something will happen in a given situation and addresses
relative rather than absolute causality.
probability theory  Theory that addresses relative rather
than absolute causality. Thus, from a probability perspec-
tive, a cause will not produce a specific effect each time
that particular cause occurs, but the probability value indi-
cates how frequently the effect might occur with the
cause.
probing  Technique interviewers use to obtain more infor-
mation in a specific area of the interview.
problem statement  Single statement that follows the sig-
nificance and background of a problem and identifies the
gap in the knowledge base needed for practice.
problematic reasoning  Involves identifying a problem,
selecting solutions to the problem, and resolving the
problem.
procedural rigor  In critical appraisal of qualitative
studies, the standard of methodological congruence
includes procedural rigor, which involves examining a
study for the researcher’s detail in applying selected pro-
cedures or steps of a qualitative study.
process of care  Construct that includes mechanisms
for delivering health care; is one of three constructs
Glossary 705
(structure, process, and outcomes of care) in Donabedi-
an’s theory of health care.
project grant proposal  Paper written to obtain funding
for the development of new educational programs in
nursing, such as a program designed to teach nurses to
provide a new type of nursing care or as a project to
support nursing students seeking advanced degrees.
proportionate sampling  In stratification, each stratum
should have numbers of subjects selected in proportion to
their occurrence in the population.
proposal, research  Written plan identifying the major ele-
ments of a study, such as the problem, purpose, and
framework, and outlining the methods to conduct the
study; a formal way to communicate ideas about a pro-
posed study to receive approval to conduct the study and
to seek funding.
proposition  Abstract, formal statement of the relationship
between two concepts.
prospective cohort study  Epidemiological study in which
a group of people are identified who are at risk for expe-
riencing a particular event.
protection from discomfort and harm  A right of research
subjects based on the ethical principle of beneficence,
which holds that one should do good and, above all, do
no harm. The levels of discomfort and harm are (1) no
anticipated effects, (2) temporary discomfort, (3) unusual
levels of temporary discomfort, (4) risk of permanent
damage, and (5) certainty of permanent damage.
providers of care  Individuals responsible for delivering
care, such as nurse practitioners and physicians, who are
part of the structure of care of Donabedian’s theory of
health care.
publication bias  Bias that occurs when studies with posi-
tive results are more likely to be published than studies
with negative or inconclusive results.
published research  Studies that are permanently recorded
in hard copies of journals or books or are posted online
for readers to access.
purposive sampling  Judgmental or selective sampling
method that involves conscious selection by the
researcher of certain subjects or elements to include in
a study. Purposive sampling is a type of nonprobability
sampling.
Q
Q-sort methodology  Technique of comparative rating in
which a subject sorts cards with statements on them into
designated piles (usually 7-10 piles in the distribution of
a normal curve) that might range from best to worst.
qualitative research  Systematic, interactive, subjective
approach used to describe life experiences and give them
meaning.
706 Glossary
qualitative research proposal  A document developed by
a researcher of a proposed qualitative study that often
includes an introduction, philosophical base, and method-
ology for conducting the study.
qualitative research synthesis  Process and product of
systematically reviewing and formally integrating the
findings from qualitative studies. Qualitative research
synthesis consists of two categories, metasummary and
meta-synthesis.
qualitative research reports  The intent of a qualitative
research report is to describe the dynamic implementation
of the research project and the unique, creative findings
obtained. The report usually includes introduction,
methods, results, and discussion sections.
quantitative research  Formal, objective, systematic
study process to describe and test relationships and to
examine cause-and-effect interactions among variables.
quantitative research proposal  A document developed
by a researcher of a proposed quantitative study that often
includes the introduction, review of the literature, frame-
work, and methodology proposed for the study.
quantitative research report  Report that includes an
introduction, methods, results, and a discussion of find-
ings for a quantitative study.
quasi-experimental research  Type of quantitative re-
search conducted to explain relationships, clarify why
certain events happen, and examine causality between
selected independent and dependent variables.
quasi-experimental study designs  Designs with limited
control that were developed to provide alternative means
of examining causality in situations not conducive to
experimental controls.
query letter  Letter sent to an editor of a journal to deter-
mine interest in publishing an article, or a letter sent to a
funding agency to determine interest in funding a study.
questionnaire  Printed self-report form designed to elicit
information that can be obtained through written responses
of the subject.
quota sampling  Nonprobability convenience sampling
technique with an added strategy to ensure the inclusion
of subject types likely to be underrepresented in the con-
venience sample, such as women, minority groups, and
the undereducated.
R formula such as the Fog Index. The readability of a scale
random assignment to groups  Procedure used to assign
subjects to treatment or comparison groups in which the
subjects have an equal opportunity to be assigned to either
group.
random-effect model  A method of analyzing studies that
are not all estimating the same intervention effect but
related effect over studies that follows a distribution
across studies.
random error  Error that causes individuals’ observed
scores to vary haphazardly around their true score without
a pattern.
random heterogeneity  Subjects in a treatment or inter-
vention group differing in ways that correlate with the
dependent variable.
random sampling methods  See probability sampling
method.
random variation  Expected difference in values that
occurs when one examines different subjects from the
same sample.
randomization  From a sampling theory point of view,
each individual in the population should have a greater
than zero opportunity to be selected for a sample, which
is achieved by random sampling. The methods of assign-
ing subjects to groups can also be random and promote
randomization in the final study groups.
randomized blocking design  Experimental design in
which the researcher includes subjects with various levels
of an extraneous variable in the sample but controls the
numbers of subjects at each level of the variable and their
random assignment to groups within the study.
randomized controlled trials (RCTs)  Classic means of
examining the effects of various treatments in which the
effects of a treatment are examined by comparing the
treatment group with the no-treatment group.
range  Simplest measure of dispersion, obtained by sub-
tracting the lowest score from the highest score or just
identifying the lowest and highest scores in a distribution
of scores.
rating scales  Crudest form of measure using scaling tech-
niques; ratings are chosen from an ordered series of cat-
egories of a variable assumed to be based on an underlying
continuum—for example, rating acute pain on a scale
from 1 to 10, with 1 being minimal pain and 10 being
extreme pain.
ratio level of measurement  Highest measurement form
that meets all the rules of other forms of measure: mutu-
ally exclusive categories, exhaustive categories, rank
ordering, equal spacing between intervals, and a con­
tinuum of values; also has an absolute zero, such as
weight.
readability  The degree of difficulty of a text to be read
and comprehended; can be determined by a readability
can influence the reliability and validity of a scale used
in a study.
repeated measures design  A research design that repeat-
edly assesses or measures study variables for the same
group of people.
reasoning  Processing and organizing ideas to reach con-
clusions; examples are problematic, operational, dialectic,
and logistic types of reasoning.

recommendations for further research  Ideas that emer­
ged from the present study and previous studies in the
same area that provide directions for future studies.
recruiting research participants  The process of obtain-
ing subjects or participants for a study that includes iden-
tifying potential subjects, approaching them to participate
in the study, and gaining their acceptance to participate.
refereed journal  Publication that uses expert reviewers
(referees) to determine whether a manuscript will be
accepted for publication.
reference group  Group of individuals who constitute the
standard against which individual subjects’ scores are
compared.
referencing  Comparing a subject’s score against a stan-
dard; used in norm-referenced and criterion-referenced
testing.
reflexivity  Self-awareness and critical examination of the
interaction between self and the data during collection and
analysis of qualitative data. May lead the researcher to
explore personal feelings and experiences that influence
the study.
refusal rate  Percentage of potential subjects who decide
not to participate in a study. The refusal rate is calculated
by dividing the number refusing to participate by the
number of potential subjects approached. For example, if
100 subjects are approached and 15 refuse to participate,
the refusal rate is 15 ÷100 = 0.15, or 15%.
regression analysis  Analysis wherein the independent
(predictor) variable or variables influence variation or
change in the value of the dependent variable.
regression coefficient R  Statistic for regression analysis.
regression line  Line that best represents the values of the
raw scores plotted on a scatter diagram. The procedure
for developing the line of best fit is the method of least
squares.
relational statement  Declares that a relationship or link
of some kind (positive or negative) exists between two or
more concepts. Relational statements are also called prop-
ositions in theory and become the focus of testing in
quantitative research.
relative risk  Type of risk associated with conducting any
diagnostic and screening test for determining the health
problems of patients.
relevant literature  Sources that are pertinent or highly
important in providing the in-depth knowledge needed to
make changes in practice or to study a selected problem.
reliability  Represents the consistency of the measure
obtained. Also see reliability testing.
reliability testing  Measure of the amount of random error
in the measurement technique. Reliability testing of mea-
surement methods focuses on the following three aspects
of reliability: stability, equivalence, and internal consis-
tency or homogeneity.
Glossary 707
reliable measure  A measurement method used in research
that provides consistent data from subjects.
replication  Reproducing or repeating a study to determine
whether similar findings will be obtained.
replication, approximate  Operational replication that
involves repeating the original study under similar condi-
tions and following the methods as closely as possible.
replication, concurrent  Involves collection of data for
the original study and simultaneous replication of the data
to provide a check of the reliability of the original study.
Confirmation of the original study findings through repli-
cation is part of the original study’s design.
replication, exact  Involves precise or exact duplication of
the initial researcher’s study to confirm the original
findings.
replication, systematic  Constructive replication that is
done under distinctly new conditions in which the
researchers conducting the replication do not follow 
the design or methods of the original researchers; instead,
the second investigative team begins with a similar
problem statement but formulates new means to verify the
first investigator’s findings.
representativeness of the sample  Sample must be like
the population in as many ways as possible.
request for applications (RFA)  Similar to an RFP except
that the government agency not only identifies the problem
of concern but also describes the design of the study.
Researchers bid for this contract.
request for proposals (RFP)  An opportunity for funding
where an agency within the federal government seeks
proposals from researchers dealing with a specific
problem. Federal Register publishes opportunities for
funding that usually have deadlines that are only a few
weeks after the publication, and researchers need to have
a strong background in the area of concern to submit a
proposal.
research  Diligent, systematic inquiry or investigation to
validate and refine existing knowledge and generate new
knowledge.
research benefit  Something of health-related, psychoso-
cial, or other value to an individual research subject, or
something that will contribute to the acquisition of gen-
eralizable knowledge. Assessing research benefits is part
of the ethical process of balancing benefits and risks for
a study.
research design  See design.
research grant  Funding awarded specifically for conduct-
ing a study.
research hypothesis  Alternative hypothesis to the null
hypothesis, stating that there is a relationship or a differ-
ence between two or more variables.
research methodology  The process or plan for conduct-
ing the specific steps of the study.
708 Glossary
research misconduct  Fabrication, falsification, or plagia-
rism in processing, performing, or reviewing research, or
in reporting research results; it does not include honest
error or differences in opinion.
research objectives (or aims)  Clear, concise, declarative
statements that are expressed to direct a study and are
focused on identification and description of variables  results  Outcomes from data analysis that are generated for
or determination of the relationships among variables,  each research objective, question, or hypothesis.
or both.
research problem  Area of concern where there is a gap
in the knowledge base needed for nursing practice.
Research is conducted to generate essential knowledge to
address the practice concern, with the ultimate goal of
providing evidence-based nursing care.
research proposal  See proposal, research.
research purpose  Concise, clear statement of the specific
goal or aim of the study that is generated from the
problem.
research questions  Concise, interrogative statements
developed to direct studies that are focused on description
of variables, examination of relationships among vari-
ables, determination of differences between two or more
groups, and prediction of dependent variable using inde-
pendent variables.
research topics  Concepts or broad problem areas that
indicate the foci of essential research knowledge needed
to provide evidence-based nursing practice. Research
topics include numerous potential research problems.
research utilization  Process of synthesizing, disseminat-
ing, and using research-generated knowledge to make an
impact on or a change in the existing practices in society.
research variables or concepts  Qualities, properties, or
characteristics identified in the research purpose and
objectives or questions that are observed or measured in
a study. Research variables or concepts are often used in
qualitative studies or descriptive and correlational quan-
titative studies, in which the intent is to observe or
measure variables as they exist in a natural setting without
the implementation of a treatment.
research participants or informants  See subjects.
researcher-participant relationships  In qualitative re-
search, the specific interactions between the researcher
and the study participants to accomplish the purpose of
the study.
residual variables  Type of variable in a model testing
design that indicates the effect of unmeasured variables
not included in the model. These variables explain some
of the variance found in the data but not the variance
within the model.
respect for persons, principle of  Indicates that persons
have the right to self-determination and the freedom to
participate or not participate in research.
responder analysis  An additional technique to identify
variables associated with a beneficial response to an inter-
vention or to determine outcomes of potential importance
to the treatment.
response set  Parameters within which the question or item
is to be answered in a questionnaire.
retention rate  The number and percentage of subjects
completing the study.
retaining research participants  Specific actions taken by
the researcher to keep subjects participating in a study and
to prevent their attrition; provides a more representative
sample for the study and decreases the threats to design
validity.
retrospective study  Epidemiological study in which a
group of people are identified who have experienced a
particular event—for example, studying occupational
exposure to chemicals to determine cause-and-effect
relationships.
review of relevant literature  Analysis and synthesis of
research sources to generate a picture of what is known
and not known about a particular situation or research
problem.
right to self-determination  See self-determination, right
to.
rigor  Striving for excellence in research through the use
of discipline, scrupulous adherence to detail, and strict
accuracy.
risk ratio  The ratio of the risk of those subjects in the
intervention group to the risk of those in the control group
for having a particular health outcome.
rival hypothesis  Alternate explanation of cause in a
study.
robustness  Analysis procedure that will yield accurate
results even if some of the assumptions are violated by
the data being analyzed.
role-modeling  Learning by imitating the behavior of an
exemplar or role model.
S
sample  Subset of the population that is selected for a
study.
sample attrition  See attrition rate.
sample characteristics  Description of the research sub-
jects obtained by analyzing data acquired from the mea-
surement of the demographic variables (e.g., age, gender,
ethnicity, medical diagnoses).
sample size  Number of subjects or participants recruited
and consenting to take part in a study.
sampling  Selecting groups of people, events, behaviors,
or other elements with which to conduct a study.

sampling criteria  List of the characteristics essential for
membership in the target population. Sampling criteria
consist of both inclusion and exclusion criteria.
sampling error  Difference between a sample statistic
used to estimate a population parameter and the actual but
unknown value of the parameter.
sampling frame  Listing of every member of the popula-
tion with membership defined by the sampling criteria.
sampling method  Process of selecting a group of people,
events, behaviors, or other elements that are representa-
tive of the population being studied; includes probability
or random and nonprobability or nonrandom sampling
methods.
sampling plan  Describes the strategies that will be used
to obtain a sample for a study and may include either
probability or nonprobability sampling methods.
scale  Self-report form of measurement composed of
several items that are thought to measure the construct
being studied, in which the subject responds to each item
on the continuum or scale provided, such as a pain percep-
tion scale or state anxiety scale.
scatter diagrams or scatterplots  Figures that provide
useful preliminary information about the nature of  lowed by the collection and analysis of qualitative data.
the relationship between variables that should be devel-
oped and examined before correlational analysis is per-
formed.
science  Coherent body of knowledge composed of
research findings, tested theories, scientific principles,
and laws for a discipline.
scientific method  All procedures that scientists have used,
currently use, or may use in the future to pursue knowl-
edge, such as quantitative, qualitative, outcomes, and
intervention research.
scientific theory  Theory with valid and reliable methods
of measuring each concept and relational statements that
have been repeatedly tested through research and demon-
strated to be valid.
secondary analysis design  Involves studying data previ-
ously collected in another study, but using different
methods of organization of the data and different statisti-
cal analyses to reexamine the data.
secondary source  Source that summarizes or quotes
content from primary sources.
seeking approval to conduct a study  Process that
involves submission of a research proposal to a selected
group for review and often verbally defending that
proposal.
selection  The process by which subjects are chosen to take
part in a study and how subjects are grouped within a study.
selectivity in physiological measures  Element of accu-
racy that involves the ability to identify the signal under
study correctly to distinguish it from other signals.
Glossary 709
self-determination, right to  Based on the ethical princi-
ple of respect for persons, which states that humans are
capable of controlling their own destiny. The right to self-
determination is violated through the use of coercion,
covert data collection, and deception in the research
process.
semantic differential scale  Instrument that consists of
two opposite adjectives with a seven-point scale between
them. The subject selects one point on the scale that best
describes his or her view of the concept being examined
on the opposite adjectives.
seminal study  Study that prompted the initiation of a field
of research.
sensitivity of physiological measures  Related to the
amount of change of a parameter that can be measured
precisely.
sensitivity of screening or diagnostic test  The accuracy
of a screening or diagnostic test; the proportion of patients
with the disease who have a positive test result or true
positive.
sequential explanatory strategy  Method whereby the
researcher collects and analyzes quantitative data fol-
sequential exploratory strategy  Method in which the
collection and analysis of qualitative data precedes the
collection of quantitative data.
sequential transformative strategy  Strategy in which
either qualitative or quantitative data collection and anal-
ysis can come first. The results are integrated during the
interpretation phase.
sequential relationship  Relationship in which one con­
cept occurs later than the other.
serendipitous results  Unexpected results from a study
that were not the original focus of the study.
serendipity  Accidental discovery of something valuable
or useful during the conduct of a study.
serials  Literature published over time or in multiple
volumes; serials do not necessarily have a predictable
publication date.
setting  Location for conducting research; may be natural,
partially controlled, or highly controlled.
sham  Interventions are often used with procedures, and
are a variation of a “fake” intervention that omits the
essential therapeutic element of the intervention.
Shapiro-Wilk’s W test  A formal test of normality that
assesses whether a variable’s distribution is skewed and/
or kurtotic.
significance of a problem  Part of the research problem that
indicates the importance of the problem to nursing and to
the health of individuals, families, and communities.
significant results  Results that are in keeping with those
identified by the researcher.
710 Glossary
simple hypothesis  Predicts the relationship (associative or
causal) between two variables.
simple interrupted time-series design  Similar to the
descriptive time-series design, with the addition of a treat-
ment that occurs or is applied (interrupts the time series)
at a given point in time.
simple linear regression  Parametric analysis technique
that provides a means to estimate the value of a dependent
variable based on the value of an independent variable.
simple random sampling  Selection of elements at random
from the sampling frame for inclusion in a study. Each
study element has a probability greater than zero of being
selected for inclusion in the study.
situated  The way a person is in a specific context and time
that shapes his or her experiences, paradoxically freeing
and constraining the person’s ability to establish meanings
through language, culture, history, purposes, and values.
situated freedom  The amount of freedom a patient has in
healthcare options as a result of his or her specific
circumstance.
skewed  A curve that is asymmetrical (positively or nega-
tively skewed) because of an asymmetrical distribution of
scores from a study.
skimming a source  Quickly reviewing a source to gain a
broad overview of the content.
slope  Determines the direction and angle of the regression
line within the graph. The value is represented by the
letter b.
small area analyses  Geographical analyses used to
examine variations in health status, health services, pat-
terns of care, or patterns of use by geographical areas.
Spearman rank-order correlation coefficient  Nonpara-
metric analysis technique for ordinal data that is an adap-
tation of the Pearson’s product-moment correlation used
to examine relationships among variables in a study.
specific propositions  Statements found in theories that are
at a moderate level of abstraction and provide the basis
for the generation of hypotheses to guide a study.
specificity of a screening or diagnostic test  Proportion of
patients without disease who have a negative test result
or true negative. Specificity indicates the accuracy of a
screening or diagnostic test.
split-half reliability  Process used to determine the homo-
geneity of an instrument’s items; the items are split in
half, and a correlational procedure is performed between
the two halves.
spurious correlations  Relationships between variables
that are not logical. In some cases, these significant rela-
tionships are a consequence of chance and have no
meaning.
stability reliability  Concerned with the consistency of
repeated measures of the same attribute with the use of
the same scale or instrument over time.
stage-based interventions  Interventions tailored to a spe-
cific phase of recovery, response to treatment, or change
in behavior.
standard deviation (SD)  Measure of dispersion that is
calculated by taking the square root of the variance.
standard of care  The national designation of the type of
care that patients should receive from healthcare agencies
and providers.
standard scores  Used to express deviations from the
mean (difference scores) in terms of standard deviation
units, such as Z scores, in which the mean is 0 and the
standard deviation is 1.
standardized mean difference  Calculated in a meta-anal-
ysis when the same outcome, such as depression, is mea-
sured by different scales or methods.
standardized mortality ratio  Observed number of deaths
divided by the expected number of deaths and multiplied
by 100.
statement synthesis  Development of statements propos-
ing specific relationships among the concepts being
studied. This step is a part of developing a framework for
a study.
statistic  Numerical value obtained from a sample that is
used to estimate the parameters of a population.
statistical conclusion validity  Concerned with whether
the conclusions about relationships and differences drawn
from statistical analyses are an accurate reflection of
reality.
statistical hypothesis  See null hypothesis.
statistical regression  Movement or regression of extreme
scores toward the mean in studies using a pretest-posttest
design.
statistical significance  Results are unlikely to be due to
chance; thus, there is a difference between groups, or
there is a significant relationship between variables.
Stetler model  Model developed by Stetler that provides
a comprehensive framework to enhance the use of re-
search findings by nurses to facilitate evidence-based
practice.
stratification  Used in a design so that subjects are distrib-
uted throughout the sample by using sampling techniques
similar to those used in blocking, but the purpose of the
procedure is even distribution throughout the sample.
stratified random sampling  Used when the researcher
knows some of the variables in the population, which are
critical to achieving representativeness. These identified
variables are used to divide the sample into strata or
groups.
strength of a relationship  Amount of variation explained
by the relationship.
structured interviews  Use of strategies that give the
researcher an increasing amount of control over the
content of the interview.

structured observation  Clearly identifying what is to be
observed and precisely defining how the observations are
to be made, recorded, and coded.
structures of care  The elements of organization and
administration, as well as provider and patient character-
istics, that guide the processes of care.
study validity  Measure of the truth or accuracy of a claim
that is an important concern throughout the research
process.
study variables  Concepts at various levels of abstraction
that are measured, manipulated, or controlled in a study.
subject attrition  See attrition rate.
subjects  Individuals participating in a study.
subject term  Term included in a database thesaurus that
one needs identified.
substantive theory  Theory of social process developed
by the discovery mode to explain a particular social
world.
substitutable relationship  Relationship in which a similar
concept can be substituted for the first concept and the
second concept will occur.
substituted judgment standard  In the ethical conduct of
research, it is a standard concerned with determining the
course of action that incompetent individuals would take
if they were capable of making a choice.
sufficient relationship  States that when the first variable
or concept occurs, the second will occur regardless of the
presence or absence of other factors.
sum of squares  Mathematical manipulation involving
summing the squares of the difference scores that is used
as part of the analysis process for calculating the standard
deviation.
summary statistics  See descriptive statistics.
summated scales  Scales in which various items are
summed to obtain a single score.
survey  Data collection technique in which questionnaires
or personal interviews are used to gather data about an
identified population.
symbolic meanings  In qualitative data analysis, the
meaning attached to particular ideas or clusters of  theoretical sampling  Often used in grounded theory
data.
symmetrical curve  A curve in which the left side is a
mirror image of the right side.
symmetrical relationship  Complex relationship that con-
sists of two statements: If A occurs (or changes), B will
occur (or change); if B occurs (or changes), A will occur
(or change); A ↔ B.
synthesis of sources  Clustering and interrelating ideas
from several sources to form a gestalt or new, complete
picture of what is known and not known in an area.
systematic bias or variation  Consequence of selecting
subjects whose measurement values are different or vary
in some way from the population.
Glossary 711
systematic error  Measurement error that is not random
but occurs consistently, such as a scale that inaccurately
weighs subjects 3 pounds heavier than they are.
systematic review  Structured, comprehensive synthesis
of quantitative studies in a particular healthcare area to
determine the best research evidence available for expert
clinicians to use to promote an evidence-based practice.
systematic sampling  Conducted when an ordered list of
all members of the population is available and involves
selecting every kth individual on the list, starting from a
point that is selected randomly.
T
table  Presentation of the results of a study in columns and
rows for easy review by the reader.
tails  Extremes of the normal curve where significant sta-
tistical values can be found.
target population  Group of individuals who meet the
sampling criteria and to which the study findings will be
generalized.
tentative theory  Theory that is newly proposed, has had
minimal exposure to critical appraisal by the discipline,
and has had little testing.
testable hypothesis  Contains variables that are measur-
able or can be manipulated in the real world.
test-retest reliability  Determination of the stability or
consistency of a measurement technique by correlating
the scores obtained from repeated measures.
textbooks  Monographs written to be used in formal edu-
cational programs
themes  In qualitative data analysis, the concepts or pat-
terns that emerge after researchers have spent extensive
time examining the data.
theoretical limitations  Weaknesses in the study frame-
work and conceptual and operational definitions that
restrict abstract generalization of the findings.
theoretical literature  Concept analyses, maps, theories,
and conceptual frameworks that support a selected
research problem and purpose.
research to advance the development of a theory through-
out the research process. The researcher gathers data from
any individual or group that can provide relevant data for
theory generation.
theory  An integrated set of defined concepts, existence
statements, and relational statements that present a view
of a phenomenon and can be used to describe, explain,
predict, or control that phenomenon.
therapeutic research  Research that provides the patient
an opportunity to receive an experimental treatment that
might have beneficial results.
thesis  Research project completed by a master’s student
as part of the requirements for a master’s degree.
712 Glossary
threats to statistical conclusion validity  The reasons for
the false conclusions that might be obtained from the
results of data analyses.
time-lag bias  A type of publication bias. Occurs because
studies with negative results are usually published later,
sometimes 2 or 3 years later, than studies with positive
results.
time-dimensional designs  Designs used to examine the
sequence and patterns of change, growth, or trends over
time.
total variance  The sum of the within-group variance and
the between-group variance; determined by conducting
analysis of variance (ANOVA).
traditions  Truths or beliefs that are based on customs and
past trends and provide a way of acquiring knowledge.
translation/application  Transforming from one language
to another to facilitate understanding; part of the process
of interpreting research outcomes in which results are
translated and interpreted into findings.
translational research  An evolving concept that is
defined by the National Institutes of Health as the  difference between groups is not significant). The null
translation of basic scientific discoveries into practical
applications.
treatment  Independent variable or intervention that is
manipulated in a study to produce an effect on the depen-
dent variable. The treatment or independent variable is
usually detailed in a protocol to ensure consistent imple-
mentation in the study.
treatment diffusion  Occurs when the control group sub-
jects communicate with the experimental subjects and are
exposed to aspects of the study treatment.
treatment fidelity  The accuracy, consistency, and thor-
oughness in how an intervention is delivered according to
the specified protocol, treatment program, or intervention
model.
treatment standardization  A process of ensuring that the
research treatment is applied consistently each time the
treatment is administered; part of intervention fidelity.
trend designs  Designs used to examine changes in the
general population in relation to a particular phenomenon.
trial and error  An approach with unknown outcomes that
is used in a situation of uncertainty when other sources of
knowledge are unavailable.
triple-blinded  Type of study in which participants,
researchers, and those involved in data management are
unaware of group assignment.
truncated  In correlational designs, if the range of scores
is truncated, the obtained correlational value will be arti-
ficially depressed. The lowest values and the highest
values for a variable either are not measured or are con-
densed and merged with less extreme values.
true negative  Result of a diagnostic or screening test that
indicates accurately when a disease is not present.
true positive  Result of a diagnostic or screening test that
indicates accurate identification of the presence of a
disease.
true score  Score that would be obtained if there were no
error in measurement; some measurement error always
occurs.
t-test  A parametric analysis technique used to determine
significant differences between measures of two samples;
t-test analysis techniques exist for dependent and inde-
pendent groups.
two-tailed test of significance  Type of analysis used for
a nondirectional hypothesis in which the researcher
assumes that an extreme score can occur in either tail.
two-way chi-square  A statistic that tests whether propor-
tions in levels of one variable are significantly different
from proportions of the second variable.
type I error  Error that occurs when the researcher con-
cludes that the samples tested are from different popula-
tions (the difference between groups is significant) when,
in fact, the samples are from the same population (the
hypothesis is rejected when it is true.
type II error  Error that occurs when the researcher
concludes that there is no significant difference between
the samples examined when, in fact, a difference exists.
The null hypothesis is regarded as true when it is
false.
U
unexpected results  Relationships found between vari-
ables or differences between groups that were not hypo­
thesized or predicted from the framework guiding the
study.
ungrouped frequency distribution  Process whereby
researchers list all categories of the variable on which
they have data and tally each datum on the listing.
unimodal  Distribution of scores in a sample has one mode
or most frequently occurring score.
unstructured interview  Interview initiated with a broad
question, after which subjects are usually encouraged to
elaborate on particular dimensions of a topic.
unstructured observations  Spontaneously observing and
recording what is seen with a minimum of planning.
V
validation  Phase in which the research reports are criti-
cally appraised to determine their scientific soundness.
validity of instrument  The extent to which an instrument
actually reflects or is able to measure the construct being
examined.
variables  Qualities, properties, or characteristics of per­
sons, things, or situations that change or (vary) and are
manipulated, measured, or controlled in research.

variance  Measure of dispersion that is the mean or average
of the sum of squares.
variance analysis  Outcomes research strategy to track
individual and group variance from a specific critical
pathway. The goal is to decrease preventable variance in
the process, thus helping patients and their families
achieve optimal outcomes.
vary  To be different; numerical values associated with
variables will change from one instance to another.
verbal presentation  The communication of a research
report at a professional conference or meeting.
vertical axis  The y axis in a graph of a regression line or
scatterplot.
visual analog scale  A line 100 mm in length with right-
angle stops at each end on which subjects are asked to
record their response to a study variable. Also referred to
as magnitude scale.
volunteer sample  Those willing to participate in the
study. All samples with human subjects must be volunteer
samples.
Glossary 713
voluntary consent  Indication that prospective subject has
decided to take part in a study of his or her own volition
without coercion or any undue influence.
W
Wilcoxon matched-pairs signed-ranks test  Nonpara-
metric analysis technique used to examine changes that 
occur in pretest-posttest measures or matched-pairs
measures.
within-group variance  Variance that results when indi-
vidual scores in a group vary from the group mean.
Y
y intercept  Point where the regression line crosses (or
intercepts) the y axis. At this point on the regression line,
x = 0.
Z
Z scores  Standardized scores developed from the normal
curve.
 Index
Page numbers followed by “f” indicate figures, “t” indicate tables, and “b” indicate boxes.
A Agency for Healthcare Research and
AAALAC. See American Association for
Accreditation of Laboratory Animal
Care
AACN. See American Association of
Critical Care Nurses
Absolute zero point, 387
Abstract concept, 151
as qualitative study focus, 151-152
Abstract thinking, 2
Abstract thought process, 4-7
Abstracts, 622, 623b
for conference presentations, critical
appraisal of, 453
Academic committees, as support system,
530
ACNP. See Acute care nurse practitioners
Acquired immunodeficiency syndrome
(AIDS)
coping interventions for, 238
research for, 169
Acute care nurse practitioners (ACNP),
care from, 221
Adjusted hazard ratios, 577-578
Administrative databases, 307
Advanced beginner nurse, 10
Advanced practice nurse (APN)
evidence-based practice and, 1
outcomes classification, 305
outcomes research and, 305-306
role of, 3
Advanced testing, 344-345
Advances in Nursing Science, 20
African Americans
hypertension in, 12f
TM program for BP in, 35, 37-38, 50,
51f
Agency for Health Care Policy and
Research (AHCPR), 22, 300
Agency for Health Services Research
(AHSR), 300
Agency for Healthcare Research and
Quality (AHRQ), 11-12, 22-23, 80,
297, 300
evidence-based guidelines and,
496-499
714
Quality (Continued)
evidence-based practice centers and,
502-503
Aggregate, 294-295
AHCPR. See Agency for Health Care
Policy and Research
AHRQ. See Agency for Healthcare
Research and Quality
AHSR. See Agency for Health Services
Research
AIDS. See Acquired immunodeficiency
syndrome
Alternate-forms reliability, 390
American Association for Accreditation of
Laboratory Animal Care (AAALAC),
191
American Association of Critical Care
Nurses (AACN), 79
American Nurses Association (ANA)
APN, types of, 305
Commission on Nursing Research, 20
mission of, 23
nursing, definition of, 1
Nursing Care Report Card, 303
nursing functions and activities, 19
research conferences by, 19
American Nurses Credentialing Center
(ANCC), 22, 469
American Nurses Foundation, 19
American Organization of Nurse
Executives (AONE), 79-80
American Psychological Association,
Publication Manual of, 107
American Recovery and Reinvestment
Act, 300-301
ANA. See American Nurses Association
Analysis
of change, 315
of data, 45-48, 491
of questionnaires, 429
techniques, 371
of improvement, 315
of ITT, 334
multilevel, 317-318
responder, 334-335
Analysis of variance (ANOVA), 371,
584-587
ANCC. See American Nurses
Credentialing Center; American
Nursing Credentialing Center
Animal subjects, 190-191
Annual Review of Nursing Research,
21
Anonymity, 171-172, 177-178
ANOVA. See Analysis of variance
Anxiety, learning and, 121
AONE. See American Organization of
Nurse Executives
APN. See Advanced practice nurse
Applicability, as outcomes assessment
instrument, 316t
Application, Stetler Model of Research
Utilization to Facilitate Evidence-
Based Practice phase, 495-496
Applied research, 35-36
Approximate replication, 77
ARIMA. See Autoregressive integrated
moving average
Articles for publication, critical appraisal
of research, 453
Assent form, 166-167, 168b
Assistance availability, data collection
and, 516-517
Association of Collegiate Schools of
Nursing, 19
Associative hypothesis
causal hypothesis versus, 144-146
expression of, 144
variables of, 144-145
Associative relationship, 144
Assumption, making explicit, 41-42
Asymmetrical curve, 540
Asymmetrical relationship, 122
Attrition, 37
Authority, 9
Authors, 105
Autoethnography, 64
Autonomous agent, 164
Autonomy
diminished, 165-169
right to, 171-173

Autonomy-enabling structure,
302-303
Autoregressive integrated moving average
(ARIMA), 241-242
B C
Baseline observation carried forward
(BOCF), 334
Basic research, 34-35
Behavioral research, National
Commission for Protection of,
162-163
Being and Time, 60-61
Being-in-time, 61
The Belmont Report, 162
Benefit-risk ratio, 175-176
Benefits, research, assessment of,
175-176
Best interest standard, 168
Best research evidence, 468
definition of, 11-12, 28-30
introduction to, 28-31
levels of, 30-31
Biased coin design, 232
Biases, 197
citation, 484
language, 484
location, 484
for meta-analysis, 484-485
outcome reporting, 484
publication, 484, 485f
systematic, 354-356
for systematic reviews, 484-485
time-lag, 484
Bibliographical database, 103, 104t
Bimodal distribution, 540-541, 540f
Biomedical research, National
Commission for Protection of,
162-163
Bivariate correlational analysis, 560-565
Bland-Altman plots, 566
example of, 565-566, 566t
result interpretation, 566
of systolic blood pressure, 406f
of test-retest data, 566f
Blinding, open label vs., 331-332
Blocking, 206-207
accomplishing effects of, 246
examples using age and ethnicity, 206t
studies using control strategies for, 205t
Blood pressure (BP), TM program for, in
monitoring African Americans, 35,
37-38, 50, 51f
BOCF. See Baseline observation carried
forward
Body of knowledge, 1, 10
Books, published research in, 632
Boolean operator, 105, 105f
Borrowing, 9
BP. See Blood pressure
Bracketing, 60
Burgess Report, 19
Calculated variable, 543
Calculation, 571-573
result interpretation, 573
California Nursing Outcomes Coalition
(CalNOC), 304
CalNOC. See California Nursing
Outcomes Coalition
Carryover effect, 249
Case study, designing, 224
Case-control designs, 583
Category systems, 421-422
Causal connection, 326-327
Causal explanation, 326-327
Causal hypothesis
associative hypothesis versus,
144-146
variables of, 145-147
Causal relationship, 123, 145
Causality, 195
understanding, role of correlation, 565
Causality theory
conditions of, 195
experimental study design examining,
231
nursing philosophy and, 196-197
Causation, probability and, 196
Cause-and-effect relationship, 50
CEA. See Cost-effectiveness analysis
Censored data, 577
Center for Epidemiological Studies
Depression Scale (CES-D), 431,
431b, 432f, 486
Central limit theorem, 538-539
Central tendency, measures of, 538,
552-554
CES-D. See Center for Epidemiological
Studies Depression Scale
CFA. See Confirmatory factor analysis
CFR. See Code of Federal Regulations
Chain sampling. See Network sampling
Change, analysis of, 315
Child
assent form, sample of, 168b
end-of-life study on, 169
as research subject, 166-167
Chi-square test (X2), 371, 587-588
Chronic obstructive pulmonary disease
(COPD), 67-68
CINAHL. See Cumulative Index to
Nursing and Allied Health Literature
CIOMS. See Council for International
Organizations of Medical Sciences
Index 715
Citation, 76
Citation bias, 484
Citation search index, 103
Classic experimental design, 652f
Classic test theory, scale construction
using, 442-444
Classical hypothesis testing, 535-536
Client treatment matching (CTM), 337
Clinical and Translational Science Awards
(CTSA) Consortium, 503
Clinical databases, 306-307
Clinical expertise, 12
Clinical importance, statistical
significance vs., 537
Clinical Laboratory Standards Institute
(CLSI), 395
Clinical practice
research priority for, 79-81
research problem, as source of, 75-76
Clinical trials, 250-251
methodology for, 250
preference, 336-337
CLSI. See Clinical Laboratory Standards
Institute
Cluster randomized controlled trials, 336
Cluster sampling, 360-361
multistage, 360-361
Cochrane, Archie, 20-21
Cochrane Center, 20-21
Cochrane Collaboration, 20-21, 31
Cochrane Nursing Care Field, 471-472
Code of Federal Regulations (CFR),
162-163
Codebook, development of, for data
definitions, 520-523, 522f
Coding plan, 517-523, 522f
Coefficient of determination, 573
Coefficient of stability, 390
Coercion, 164
Cognitive application, 495-496
Cohort, 218
Cohort design, 222
Cohort study
prospective, 309
retrospective, 310-311
Colleague support, 531
Common-Sense Model of Illness
Representation, 142-143, 143f
Comparative descriptive design
example of, 218f
purpose of, 217-218
Comparative evaluation, Stetler Model
of Research Utilization to Facilitate
Evidence-Based Practice phase,
495
Comparative experimental design, 232
Comparative treatment efficacy (CTE)
trial, 251
716 Index
Comparison group, 203, 356-357
nonequivalent design of, 234-240
one-group posttest-only design with,
234, 235f
one-group pretest-posttest design with,
234-236, 236f
posttest-only design with, 234, 235f
pretest and posttest design with,
237-238, 237f-239f
removed, 238-240, 239f
reversed, 240, 240f
selection of, 233-234
Competent nurse, 10
Complex hypothesis, 146-147
Complexity of care, dimension of, 229
Comprehension, as informed consent,
178-180
Comprehensiveness, as outcomes
assessment instrument, 316t
Computer search, 104
Computer-based data collection, 513-514
Computerized databases, 103
Concept analysis, 119
how to conduct, publications on, 119t
published examples of, 119t
Concept derivation, 118-119
Concept synthesis, 118
Concepts, 41, 116
abstract, 151
as abstract, 151-152
definition of, 116, 132
for frameworks, 132
linking to concept, 133
operationalizing, 155-156
research, 151-152
research variables and, 151-152
as variable, 138, 151-152
Conceptual definition, 43
denotative definition versus, 118
importance of, 119-120
Conceptual map
construction of, 132-134
example of, 133f
Conceptual model, 117
Orem’s model of self-care as, 125
Conclusions, 48
formation of, 597-598
Concrete thinking, 2
Concrete-abstract continuum, 2-3
Concurrent relationship, 123
Concurrent replication, 77
Concurrent triangulation strategy, 211,
211f
Condensed proposal, content of, 644
Conduct and Utilization of Research in
Nursing (CURN), 21
Confidence intervals, 541-542, 558
odds ratio and, 575
Confidentiality
assurance of, 177-178
breach of, 172
maintaining in research, 172-173
right to, 171-173
Confirmatory analysis, 545
Confirmatory data analysis, 542
Confirmatory factor analysis (CFA), 398
Confirmatory studies, 370
Confounding variable, 152, 342
Consensus knowledge building, 308
Consent. See also Informed consent
recording of, 182
sample form for, 181f
voluntary, 180
waiving, 180
written documents for, 180-182
Consistency, data collection and, 516
Consolidated Standards for Reporting
Trials (CONSORT), 251-253, 252f,
331, 520
CONSORT. See Consolidated Standards
for Reporting Trials
Construct
confounding and levels of, 201
preoperational clarification of, 200-201
validity, 200-202
Content analysis, 281-282
Content expert, 666
Content validity, 394-397
Content validity index (CVI), 395
Content validity ratio (CVR), 395, 396t
Contingent relationship, 123-124
Continuous outcomes, meta-analysis
results for, 486
Control, 13-14
of environment, 199, 203
of extraneous variables of, 204-207
group, 203, 205t, 233-234, 356-357
of measurement, 204
in quantitative research, 36-38, 37t
in research, 197
settings for, 37
subject, controlling equivalence of, 203
of treatment, 203-204
Convenience sampling, 363-364
Convergent validity, 400-401
COPD. See Chronic obstructive
pulmonary disease
Correlational analyses, 560
Correlational research, 26, 49-50
decision tree for, 258f
purpose of, 87
topics, problems, and purpose of,
88t-89t
Correlational study design
descriptive, 225-226, 225f
focus and purpose statement of, 244t
Correlational study design (Continued)
predictive, 226-227, 226f
purpose of, 224-228
Cost factors, data collection and, 515
Cost-benefit analysis, 312
Cost-effectiveness analysis (CEA),
312-313
Costs of care
to maintain quality, 297-298
studies examining, 298t
Council for International Organizations of
Medical Sciences (CIOMS), 169
Council of Nurse Researchers, 20
Counterbalanced design, 249-250
Counterbalancing, 204, 205t
Covert data collection, 165
Cox proportional hazards regression,
577-578, 577t
Criterion-referenced testing, 388
Critical appraisal
of abstracts for conference
presentations, 453
of published systematic review,
480-482
of research, 451
articles for publication, 453
evolution of, 451-454
implemented in nursing, 452-454
by nurse educators, 453
by nurse researchers, 453
nurses’ expertise in, 454, 454t
by practicing nurses, 452-453
presentations, 453
proposals, 453-454
publications, 453
of studies by students, 452
Critical cases, 365
Critical pathway, 309
Critical social theory
problem and purpose of, 93t
topics, problems, and purpose of, 92t
Critical-Care Pain Observation Tool,
419b
Cronbach’s alpha coefficient, 391-392
Crossover design, 249-250
Crossover randomized controlled trials,
336
Cross-sectional design, 43-44
example of, 220f
purpose of, 220-221
Cross-sectional study, 222f
CTE. See Comparative treatment efficacy
trial
CTM. See Client treatment matching
CTSA. See Clinical and Translational
Science Awards Consortium
Cumulative Index to Nursing and Allied
Health Literature (CINAHL), 103

CURN. See Conduct and Utilization of
Research in Nursing
Curvilinear relationship, 121
example of, 122f
CVI. See Content validity index
CVR. See Content validity ratio
D covert, 165
Damage, permanent
certainty of, 175
protection from, 174-175
risk of, 175
Data
calculations, 543
definitions, codebook development for,
520-523, 522f
deviation exploration using statistics,
554-558
confidence intervals, 558
degrees of freedom, 558
measures of dispersion, 554-558
normal curve, 555-557
sampling error, 557-558
documentation, 543-544
entry period, 532
exploratory analysis of, 544-545
tables and graphs for, 545
management of, 531-532
preparation of, for computer entry,
531-532
quality of, 372
retrieval of, 532
saturation of, 371
scoring, 543
storage of, 532, 543-544
summarizing of, with statistics,
550-554
central tendency measures, 552-554
frequency distributions, 550-552
transformations, 543
Data analysis, 45-48, 491
confirmatory, 542
exploratory, 542
packages, sources of, 537t
practical aspects of, 542-546
preparation, 542-544
cleaning, 543
identification of missing, 543
process, evidence examination from,
593-594
quantitative, stages of, 542
of questionnaires, 429
results
evidence examination from, 594-595
mixed, 594-595
negative, 594
nonsignificant, 594
serendipitous, 595
Data analysis (Continued)
significant and nonpredicted, 594
significant and predicted, 594
unexpected, 595
techniques, 371
Data collection, 45-46, 345-347,
523-530
external influences on subject
responses, 527
factors influencing, 515-517
assistance availability, 516-517
consistency, 516
cost, 515
time, 515-516
flow chart, 521f
forms, 517-519, 519f, 658
maintaining control, 524
methods of, 268-276
modes of, 507-515, 508f
computer-based, 513-514
electronic, 508-515
online, 508-513
participant-administered instruments,
507-508
phones, 514-515
researcher-administered instruments,
507-508
scannable forms, 508, 509f-512f
passive resistance, 527-528
people problems, 524-528
plan, 517-523
detailed, 519-520
problems, 524
with sample selection, 525, 526f
solving, 524
researcher problems in, 528-530
event problems, 529-530
institutional problems, 529
interactions, 528
lack of skill in technique, 528-529
maintaining perspective, 529
role conflict, 528-529
subject as object, 526-527
subject attrition, 525-526, 526f
subject mortality, 525-526
tasks, 523-524
Data collection process, evidence
examination from, 592-593
Data safety and monitoring board, as
support system, 531
Data use agreement, 170-171
Databases
administrative, 307
classification schemes used in, 308
clinical, 306-307
examples of, 307t
measurement using existing, 439-440
Index 717
Databases (Continued)
as sample source, 306-308
saved searches and alerts, 108
searching, 103
terms, arrangement of, 108-109
types emanating from patient care,
306f
Debriefed, subjects, 178
Deception, 165
Decision making, Stetler Model of
Research Utilization to Facilitate
Evidence-Based Practice phase,
495
Decision points, 520
Decision tree
for correlational studies, 258f
for descriptive studies, 258f
for experimental studies, 260f
for quasi-experimental studies, 259f
for statistical procedure, 546, 547f
for type of study selection, 257f
Declaration of Helsinki, 160
Deductive reasoning, 7, 36
Deductive thinker, 148
Deductive thinking, 58
Degrees of freedom (df), 545, 558, 563
De-identifying health data, 170
Delphi technique, 435-437, 436f
Demographic variable, 154, 154t
Denominator, 575
Department of Health, Education, and
Welfare, U.S. (DHEW)
regulations set by, 162
Tuskegee Syphilis Study, 161
Department of Health and Human
Services, U.S. (DHHS), 300
human research subjects, protection of,
163, 164t
IRB/privacy board responsibilities,
188t
Dependent t-test analysis technique,
580
Dependent variable, 145, 151
Depression, guided imagery and, 113
Description, 12-13
Descriptive correlational design, 43-44
example of, 225f
purpose of, 225-226
Descriptive research, 26, 49
purpose of, 87
topics, problems, and purpose of,
88t-89t
Descriptive statistics, 538
Descriptive study design
complexity of, 215-216
example of, 216f
focus and purpose statement of, 244t
purpose of, 215-224
718 Index
Descriptive study design (Continued)
uses for, 215
using questionnaires, 217
Descriptive theory, 327
Designs, 36. See also Research
comparison of, 244t
concepts important to, 195-197
definition of, 214
descriptive correlational, 43-44
development of, 214
nonexperimental, 337-338
purpose of, 218
selection, decision tree for, 257f
studies using control strategies for, 205t
Deterministic relationship, 123
df. See Degrees of freedom
DHEW. See Department of Health,
Education, and Welfare, U.S.
DHHS. See Department of Health and
Human Services, U.S.
Diagrams, scatter, 560
Dialectic reasoning, 6
Diary, 437-439
key points when selecting, 437
sample page of, 438f
Dichotomous outcomes, meta-analysis
results for, 486-488, 488f
Difference scores, as measure of
dispersion, 554, 555t
Digital object identifiers (DOIs), 107
Diminished autonomy, 165-169
Direct measurement, 382-383
Direct physiological measures, 413-414
Direct recall, 310
Directional hypothesis, 148-149
Directory of Nurse Researchers, Sigma
Thata Tau, 440-441
Disclosure, consent to incomplete, 178
Discomfort
protection from, 174-175
temporary, 174-175
unusual levels of, 175
Discriminant analysis, validity from, 401
Dispersion, measures of, 558
difference scores, 554, 555t
range, 554
standard deviation, 555
variance, 554-555, 555t
Disproportionate sampling, 360
Divergent validity, 401
Documentation. See also Informed
consent
written consent form, 180-182
DOIs. See Digital object identifiers
Domain concepts, 126t
Donabedian, Avedis, 294
Dose-intensity, of interventions, 332-333
Double-blinded, 331
Dummy variables, used in regression
analysis, 574-575
E literature review and, 98
eBooks, 100
EBP. See Evidence-based practice
Economic study, 312-314
ECPs. See Emergency contraceptive pills
Eddy, David, 22
Education, nursing, 19
EFA. See Exploratory factor analysis
Effect size (ES), 121, 330, 368-369, 536
Effectiveness
efficacy vs., 329-330, 330t
testing variations in
based on intervener characteristics,
345
based on intervention strength, 345
based on patient characteristics,
344-345
based on setting, 345
Effectiveness and Efficiency: Random
Reflections on Health Services,
20-21
Efficacy, effectiveness vs., 329-330, 330t
Electronic data collection, 508-515
Electronic database, 102
Electronic monitoring
physiologic measures obtained through,
415-417
process of, 415f
Elements, 351-352
Eligibility criteria, 352-353
Emergency contraceptive pills (ECPs),
qualitative study and, 366-367
Empirical generalizations, 599
Empirical literature, 100
discussion of, 111-112
Empirical world, 3
End-of-life study, 169
Endogenous variable
cause of, 227
measurement of, 227-228
Entry period, 532
Environment, control of, 199, 203
Environmental variable, 153-154
Epidemiological study, 218-219
Epistemology, 58
Equivalence, of subjects and groups, 203
Equivalence reliability, 390-391
Error rate, 198
Error score, 383
ES. See Effect size
Essentialist, 196
Estimation, 541-542
Ethical codes
development of, 159-163
in research, 159
Ethical study, 314
Ethics, research and, 159
Ethnographic research, 27, 63-65
methodology for, 286-287
philosophical orientation of, 64-65
problem and purpose of, 92
topics, problems, and purpose of,
90t-91t
Ethnographies, 63-64
nursing science contribution from, 65
types of, 64-65, 64t
Ethnonursing research, 65
Evaluation, Stetler Model of Research
Utilization to Facilitate Evidence-
Based Practice phase, 496
Evaluation apprehension, 201
Event-partitioning design, 222-223
Evidence, examination of, 590-595
from data analysis process, 593-594
from data analysis results, 594-595
from data collection process, 592-593
from measurement, 591-592
from research plan, 590-591
Evidence-based practice (EBP), 297, 468
advanced practice nurse and, 1
barriers related to, 468-471
benefits related to, 468-471
best research evidence for, 28-31
definition of, 17
for elderly African American women
with HTN, 12f
ethical research in, 159
guidelines
development of, 496-499
implementation of, 496-502
introduction to, 31
websites for, 499
models of, 11f-12f, 493-496
nurses promoting, 1
research, significance of, 11-14
resources, 472t
Stetler Model of Research Utilization
to Facilitate, 474, 493-496
Evidence-based practice centers (EPCs),
502-503
Exact replication, 77
Exclusion sampling criteria, 353
Exempt from review, 183-185, 185b
Existing healthcare data, 439-440
Exogenous variable
cause of, 227
measurement of, 227-228
Experience
of competent nurse, 10
of expert nurse, 10
levels of, 10
personal, 10

Experimental designs, 335-337
classic, 652f
Experimental posttest-only comparison,
246f
group design, 246
Experimental research, 26-27
elements of, 244
purpose of, 53
Sir Ronald Fisher and, 34
topics, problems, and purpose of,
88t-89t
variables in, 156
Experimental study, 37
decision tree for, 260f
intervention development in, 228-231
manipulation in, 197
research design for, 43
settings and conditions for, 87
Experimental study design, 244-254
causality, examination of, 231
as classic/original, 218f, 245
focus and purpose statement of, 244t
posttest-only comparison group, 246,
246f
types of, 244t
Experimenter expectancy, 201
Expert nurse, 10
Explanation, 13
causal, 326-327
Exploratory data analysis, 542, 544-545
tables and graphs for, 545
Exploratory factor analysis (EFA), 398
Exploratory studies, 370
Exploratory-descriptive qualitative
research, 27, 66-68
nursing research contribution from,
66-68
philosophical orientation, 66
External validity, 202
Extraneous factors, 342
Extraneous variable, 199, 342
control of, 204-207
types of, 152-154
F human research subjects, protection of,
F statistic, 585
Fabrication, 188
Face validity, 394-397
Faces pain thermometer (FPT), 418,
419b
Factor analysis, 566-568
validity from, 398-400
Factor scores, 568
Factorial design
example of, 247f
overview of, 247
Facts for Action to Stroke Treatment
(FAST), 230-231
Index 719
Fair treatment, 173-174 Framework (Continued)
subject selection, 173-174 relational statement, extracting, 132
of subjects, 174 relational statements, examination of,
False negative, 406 120-124
False positive, 406 statement synthesis and, 132
Falsification, 188 terms relating to, 116-117
FAST. See Facts for Action to Stroke Framingham study, 219
Treatment Freedom from drift, 404
Fatigue countermeasures program for Frequency distributions, 539f, 550-552,
nurses (FCMPN), 145-146, 151, 154, 551f
155b-156b grouped, 551, 551t
Fatigue effect, 249 with outlier, 553f
FCMPN. See Fatigue countermeasures without outlier, 553f
program for nurses of SF-36 Physical Functioning scale,
FDA. See Food and Drug Administration, 556, 556f
U.S. ungrouped, 550-551
Federal grant, proposal submission for, Frequency response, 404
670-671 Frequency table, 550
institutional support verification, 670 of smoking status, 552t
rejection of, 671 Functional Outcomes of Sleep
review process, understanding of, 671 Questionnaire (FOSQ), 335
uniqueness of, 670, 670t Fundamentalists, 387-388
Field setting. See Natural setting Funding, for research
Field work, 286 application requests, 670
Findings apprenticeship, 666
communication of, 49 capacity, 665
determination of, 595, 595t-596t capital, 665
interpretation of, 48 commitment level, 665
Fisher, Sir Ronald, 34 contribution, 664-665
Fishing, 198 course attendance, 666
Fixed-effects model, 486 foundations, 669
Focus group, 274-276 government, 669-670, 669f
assumptions of, 274 grants types for, 667
data collection from, 274 grantsmanship, 666
example study using, 276 industry, 668-669
facilitator role in, 275-276 investigator-initiated proposals,
participants of, recruitment of, 275 669-670
purpose of, 274 local, 667-668
questions for, 276 national nursing organizations, 668,
setting of, 275 668t
Fog formula, 398b, 442 nongovernment, 667-669
Food and Drug Administration, U.S. private, 667-668
(FDA) program building, 663-665, 664f, 664t
Amendments Act of 2007, 163 proposal requests, 670
regional nursing research organizations,
163 667, 667t
IRB/privacy board responsibilities, 188t research committees, serving on, 667
research, regulations in, 164t sources for, 667-670
Forced choice, 430 support of other people, 665-666
Forest plot, 488, 489f workshop attendance, 666
FOSQ. See Functional Outcomes of Sleep Funnel plot, 484, 484f-485f
Questionnaire
FPT. See Faces pain thermometer G
Framework GDS. See Geriatric depression scale
conceptually defining, 132 General proposition, 124
definition of, 116 Generalizations, 25
development of, 41-42 empirical, 599
map, 41, 42f of findings, 598-599
720 Index
Generalizing, 36, 352
Geographical analysis, 311-312
as small area analysis, 311
Geographical information system (GIS),
299
Geriatric depression scale (GDS),
581-582
GI. See Guided imagery
GIS. See Geographical information
system
Going native, 286
Gold standard, 31, 406
Goldmark Report, 19
Google, 106
Government funding, for research,
669-670, 669f
Government reports, 100
Graduate Record Examination (GRE),
388
Grand theories, 125, 125t-126t
Grants. See also Research, funding for
management of, 671-672
budget, 671-672
coordinating activities, 672
research personnel training, 672
study schedule, 672
submitting reports, 672
planning future, 672
GRE. See Graduate Record Examination
Grey literature, 475
Grounded theory research, 27, 62-63
definition of, 62
methodology for, 285-286
nursing research contribution from, 63
pain management theory, 142
philosophical orientation, 63
purpose of, 87-92
topics, problems, and purpose of, 90t-91t
Group, random assignment to, 232-233
Grouped frequency distribution, 551, 551t
Grove Model for Implementing
Evidence-Based Guidelines in
Practice, 500-501, 501f
Guided imagery (GI), 113
depression and, 113
H philosophical orientation, 68
HAPI. See Health and Psychological
Instruments Online
Harm, protection from, 174-175
Harvard Nurse Health Study, 309
Hawthorne effect, 38
Hazard ratio (HR), 577
adjusted, 577-578
Health
concept of, 294
definition of, 295
nurses promoting, 22
Health and Psychological Instruments
Online (HAPI), 440
Health care, theory of, 294
Health Care Finance and Quality Act of
1999, 22
Health Insurance Portability and
Accountability Act (HIPAA), 439
data use agreement, 170-171
enactment of, 159
IRB/privacy board responsibilities,
188t
objective and applicability of, 164t
PHI, authorization to use, 182-183,
184f
Privacy Rule and, 163, 170
Privacy Rule influence on IRB, 187
Health Promotion/Transtheoretical Model,
78-79, 79f
Healthcare data, existing, 439-440
Health-Promoting Lifestyle Profile II
(HPLPII), 45b
Health-related quality of life (HRQoL),
603
self-care program on, 245-246
Healthy People 2000, 22
Healthy People 2010, 22
Healthy People 2020, 22, 80-81
Heidegger, Martin, 60-61
Helsinki, Declaration of, 160
Heterogeneity, 205t, 206, 482, 554
Heterogenous, 554
population, 352-353
HHCC. See Home Health Care
Classification System
Hierarchical statement set, 132
Highly controlled setting, 373-374
for research, 37
Hinshaw, Ada Sue, 21
HIPAA. See Health Insurance Portability
and Accountability Act
Histogram, of smoking status, 552f
Historical research, 27, 68
data analysis of, 289
literature review and, 98
methodology for, 289-290
nursing science contribution from, 68
problem and purpose of, 92
sources for, 289
topics, problems, and purpose of,
90t-91t
History, treatment and, 202
History effect, 199
HIV. See Human immunodeficiency virus
Home Health Care Classification System
(HHCC), 325
Homogeneity, 205t, 206, 391
Homoscedastic, 562
Hospitalized patients, as research subjects,
169
HPLPII. See Health-Promoting Lifestyle
Profile II
HR. See Hazard ratio (HR)
HRQoL. See Health-related quality of life
HTN. See Hypertension
Human Genome Project, U.S., 417
Human immunodeficiency virus (HIV),
coping interventions for, 238
Human rights, protection of, 163-175
Human subjects
AIDS patients as, 169
as autonomous agent, 164
children as, 166-167
coercion of, 164
comprehension of, 178-180
confidentiality of, 171-173
deception of, 165
DHEW, regulations set by, 162
with diminished capacity, 167-169
discomfort, protection from, 163
fair treatment of, 174
harm, protection from, 163
informed consent form, 176-183
legally and mentally incompetent,
165-169
National Commission for Protection of,
162-163
neonate as, 165-166
permanent damage to, 175
pregnant women as, 167
prisoners as, 169
selection of, 173-174
terminally ill as, 169
unethical treatment of, 159-160
Humane Care and Use of Laboratory
Animals Regulations, 191
Husserl, Edmund, 60
Hypertension (HTN)
evidence-based practice model for, 12f
research relating to, 11-12
Hypotheses, 124, 150
associative, 144-146
causal, 144-146
complex, 146-147
development of, 148-149
directional, 147-149
formulation of, 142-149
generating for research, 150t
guessing, 201
nondirectional, 147
null, 147-148
purpose of, 142
research, 147-148
rival, 199
simple, 146-147
sources of, 142-144

Hypotheses (Continued)
statistical, 147
testable, 149
testing, 541-542
types of, 144-148
Hypotheses testable, 149
Hypothetical population, 352
I levels of reviews conducted by,
IIHI. See Individually identifiable health
information
Illness Representation, 143
IM. See Index Medicus
Image, 20
Improvement, analysis of, 315
Incompetence, 167-168
Incomplete disclosure, consent to, 178
Independent samples t-test, 581-583
calculation of, 581-582
nonparametric alternative, 582-583
result interpretation, 582
Independent t-test analysis technique, 580
Independent variable, 145, 151
Index Medicus (IM), 103
Indicator
nursing-sensitive quality indicators, 304
of quality nursing practice, 303-304
Indirect measurement, 383
Indirect physiological measures, 413-414
Indirect recall, 310
Individual, 294-295
Individually identifiable health
information (IIHI), 170
Inductive reasoning, 7
Inductive thinker, 148
Inductive thinking, 58
Industry funding, for research, 668-669
Infer, 537
Inference, 537
Inferential statistics, 538, 542
Informants, 65, 351-352, 352f
Information, gathering of, 339, 339b-340b
Information-rich cases, 365
Informed consent, 659-660
for child research subject, 166, 166t
competence to give, 180
comprehension of, 178-180
documentation of, 180-182
from legally authorized representative,
182-183
from mentally ill, 182
obtaining from subjects, 176-183
simplification of, 179b
voluntary, 180
Ingenta Connect, 106
Institute for Research and Service in
Nursing Education, 19
Institutional review, 183-187
Institutional review board (IRB)
approval for research participation,
168
benefit-risk ratio, 176
complete review, 186
expedited versus complete, 185-186,
186b
HIPAA Privacy Rule influence on, 187
183-187
members of, 163
purpose of, 183
Institutional support, 531
Institutions. See Hospitalized patients;
Prisoners
Instrumentation, 44-45, 199
assessing readability levels of, 442
evaluation of existing, 442
locating existing, 440-441
reliability of, 397-398
selection of existing, 440-442, 441f
Intellectual contribution, 465
Intention-to-treat (ITT), analysis of, 334
Intermediate end points, 314-315
Intermediate mediation, 196
Intermediate outcomes, 342
Internal consistency, 391-393
Internal validity, 199-200
International Journal of Nursing
Terminology and Classifications, 20
International Organization of
Standardization (IOS), 395
Internet. See also Online data collection
literature search using, 106, 106t
research problems and, 76
studies, searching for, 86
Interpretation, of research outcomes, 48,
590
Interrater reliability, 390
Interrupted time-series design
with multiple treatment replications,
242f, 244
with nonequivalent no-treatment
comparison group time series,
242f, 243
purpose of, 241-244
simple, 241f, 242-243
Interval data, 388
Interval level, of measurement, 387
Intervening variable, 123-124
Intervention fidelity, 37, 198-199,
605-606
Intervention label, 229
Intervention reliability, 198-199
Intervention research, 3, 8, 28
focus of, 74-75
information gathering for, 339,
339b-340b
Index 721
Intervention research (Continued)
planning of, 338-339
problems of, 326-329, 327f
project planning in, 338
project team formation, 338
project team work, 338-339
purpose of, 74-75, 93
systematic replication and, 77
Intervention studies, problems examined
by, 325, 326b
Intervention theory, 327
elements of, 328b
Intervention-based research
critical appraisal of, 346b-347b, 347
nursing conducted, 323-324
phases of, 340f
terminology for, 329-335
Interventions
definition of, 324
designing of, 229-231, 339-342,
341b
dose-intensity of, 332-333
formal testing of, 344
in NIC, 229
pilot test of, 343-344
placebo, 330-331
scientific rationale for, 328-329
sham, 330-331
stage-based, 325
taxonomy, 325
testing process of, 339-345
theory-based, 327-328
validation of, 342-343
Interviewers, training of, 424
Interviews, 271-274, 422-425
advantages of, 424-425
conducting an effective, 272-273
designing questions for, 422
disadvantages of, 424-425
learning to, 272
pilot testing protocol for, 424
positive environment for, 272
preparation for, 424
problems with, 273
recording data, 424
structured, 422-424, 425b
Introspection, 5
Intuition, 5-6
Intuitive comprehension, 465
Inverse linear relationship, 561
Investigator-initiated research proposal,
669-670
IOS. See International Organization of
Standardization
Iowa Model of Evidence-Based Practice
to Promote Quality of Care, 493,
496, 497f
IRB. See Institutional review board
722 Index
Item response theory, scale construction
using, 444-445
ITT. See Intention-to-treat
J Limited data set, 170-171
JCAHO. See Joint Commission on
Accreditation of Healthcare
Organizations
Jewish Chronic Disease Hospital Study,
162, 165
Joanna Briggs Institute, 31
The Joint Commission, 304
Joint Commission on Accreditation of
Healthcare Organizations (JCAHO),
22
Joint National Committee on Prevention,
Detection, Evaluation, and Treatment
of High Blood Pressure, 499-502,
500f
Journal article, 628-631
manuscript preparation, 629-630
manuscript revision requests, 630-631
manuscript submission for review, 630
peer review, 630
query letter development, 629
selection of, 628-629
Journal Citation Report, 628
Journal impact factor, 628-629
Journal of Nursing Scholarship, 20
Journals, online, published research in,
631-632
Judgemental sampling. See Purposive
sampling
K summary of, 109, 109t
Kendall’s tau, 564
Keyword, 101-102
Knowledge, 8-10
acquiring, ways to, 8-10
body of, 1, 10
clinical expertise and, 12
Kolmogorov-Smirnov D test, 541
Kurtosis, 541, 541f
L performing complex, 105-106
Laboratory tests, physiological measures
from, 414-415
Landmark study, 100
Language, 106
Language bias, 484
Last observation carried forward (LOCF),
334
Latent transition analysis (LTA), 317
Law, 8
Learning, anxiety and, 121
Legally authorized representative, 168
Leptokurtic, 541, 541f
Levels of measurement, 385-388, 386f
Life Orientation Test (LOT-R), 45b
Likelihood ratios (LRs), 408
Likert scale, 45, 45b, 388, 430-431
Limitations, identification of, 598
Limited recall, 310
Line of best fit, 570-571
Linear regression equation, computation
of, 571t
Linear relationship, 560-561, 570, 571f
example of, 121f
inverse, 561
negative, 561
positive vs., 121
positive, 561
negative vs., 121
value of, 121
Literature
empirical, 100
discussion of, 111-112
searching of, 490
types of, 97
Literature review, 40-41
course papers, writing, 98
definition of, 97-98
example of, 112-114
nursing journals, 98
processing of, 108-110
purpose of, 98-100
qualitative research and, 98
quantitative review and, 98-100, 99t
sources included in, 101-102
stages of, 98, 102-108
strength of evidence, examining, 98
synthesis of study to generate, 104t
systems model of, 102, 102f
time frame for, 101
writing of, 110-112
Literature search, 102
by author, 105
databases for, 103
keyword selection for, 104-105
limitation of, 106
plan development, 102-103, 103t
relevant, location of, 106-107
saved, 108
strategy for, 104-106
using Internet, 106, 106t
written record of, 103, 103t
Local funding, for research, 667-668
Location bias, 484
Locational operator, 105
LOCF. See Last observation carried
forward
Logic, 6
Logical Investigations, 60
Logical positivism, 24
Logistic reasoning, 6-7, 36
Logistic regression, 576-577
calculation, 576
odds ratio in, 576-577
Longitudinal design
example of, 219f
purpose of, 219-220
with treatment partitioning, 223f
LOT-R. See Life Orientation Test
Low statistical power, 198
LRs. See Likelihood ratios
LTA. See Latent transition analysis
Lung damage, smoking and, 7-8
M
MacArthur Competency Assessment Tool
for Clinical Research (MacCAT-CR),
168
MacCAT-CR. See MacArthur Competency
Assessment Tool for Clinical
Research
MACE. See Major adverse cardiovascular
events
Magnet Hospital Designation Program for
Excellence in Nursing Services, 22,
452-453, 469
Magnitude scaling, 433
Major adverse cardiovascular events
(MACE), 577-578, 577t
Manipulation, 197
Manipulation checks, 332, 333b
Mann-Whitney U test, 582-583
Manuscript
preparation of, 629-630
revision requests, 630-631
submission for review, 630
Matching, 205t, 207
Maturation, 199
McNernar test, 587
MD. See Median
Mead, George Herbert, 63
Mean, 538, 539f, 553-554
deviation, 554
standard error of, 557
Measurement error, 383-385, 383f
types of, 383-385
Measurement sensitivity, 370-371
Measurements
control of, 204
direct, 382-383
directness of, 382-383
evidence examination from, 591-592
indirect, 383
interval level of, 387
levels of, 385-388, 386f
controversy over, 387-388
for statistical analyses, 387

Measurements (Continued)
methods, reliability testing of, 544
methods of, 44-45, 314-315
nominal level of, 386
observational, 421-422
ordinal level of, 386-387
ratio level of, 387
reference testing of, 388
reliability of, 198
using existing databases, 439-440
Median (MD), 538, 539f, 553
Mediating variable, 123-124
Mediator variable, 154-155
Medical Outcomes Study (MOS), 301-302
conceptual framework for, 301f
MEDLINE, 103
Memoing, 283
Mental illness, informed consent from
patient, 167-169
Mentor, 666
Mentorship, 10
role-modeling, as form of, 10
Merleau-Ponty, 61
Mesokurtic, 541, 541f
Meta-analysis, 29
biases for, 484-485
clinical question for, 483
conduction, guidelines for synthesizing
research evidence, 479, 481t
for continuous outcomes, 486
for dichotomous outcomes, 486-488,
488f
directions and purpose to direct, 483
facilitation of, reporting for authors,
482b
search criteria for, 483
strategies for, 483
Metasearcher, 106
Metasummary, 489-490
Meta-synthesis, 29
conduction, of qualitative research,
489-491
oasis-ebl|Rsalles|1476813940
definition of, 489-490
exercise, framing of, 490
findings, discussion of, 491, 492f
Method of least squares, 571
Methodological coherence, 465
Methodological congruence, 465
Methodological design, 255
Methodological limitations, 598
Metric ordinal scale, 387
Micromediation, 196
Middle development, 251
Middle-range theories, 117
application of, 125-126
in nursing, 127t-128t
qualitative research and, 125-126
used in nursing, 127t-128t
Midwest Nursing Research Society
(MNRS), 19
Minimal risk, 185-186
Minnesota Multiphasic Personality
Inventory (MMPI), 388
Misconduct, in research, 159
Mixed methods, 208-212
approaches, 208
concurrent procedures, 208-209
sequential procedures, 208
transformative procedures, 209
Mixed results, 594-595
Mixed-methods systematic reviews,
29-30, 492-493
basic structure for, 493
MMPI. See Minnesota Multiphasic
Personality Inventory
MNRS. See Midwest Nursing Research
Society
Modality, 540-541
Mode, 538, 539f, 552
Model-testing design
example of, 227f
requirements for, 227-228
Moderator variable, 154-155
Molar causal law, 196
Monographs, 100
Monomethod bias, 201
Mono-operation bias, 201
Montag, Mildred, 19
Morisky medication adherence scale, 660
MOS. See Medical Outcomes Study
MTQ: Purposeful Action, 395-397
Multicausality, 196
Multicollinearity, 573-574
Multidimensional scaling, scale
construction using, 445
Multilevel synthesis, 492-493
Multimethod-multitrait technique,
384-385
Multiple R, 572
Multiple regression analysis, 573-575,
574t
Multistage cluster sampling, 360-361
N example of, 248f-249f
NANDA. See North American Nursing
Diagnosis Association
Narrative analysis, 282-283
National Association of Neonatal Nurses,
31
National Center for Advancing
Translation Sciences (NCATS), 503
National Center for Nursing Research
(NCNR), 21
National Commission for Protection of
Human Subjects of Biomedical and
Behavioral Research, 162-163
Index 723
National Database of Nursing Quality
Indicators (NDNQI), 303-304,
439-440, 440b
National Guideline Clearinghouse (NGC),
31
National Human Genome Research
Institute (NHGRI), 35
priorities of, 35-36
National Institute of Nursing Research
(NINR), 21, 23
mission of, 23
research priorities for, 80
science of nursing, 80
National League for Nursing (NLN), 20
National Library of Medicine
(MEDLINE), 103
National Nursing Home Survey (NNHS),
362
National nursing organizations, 668,
668t
National Quality Forum (NQF),
304-305
Natural setting, 373
for research, 37
Naturalistic inquiry, 66
Nazi medical experiments, 160
NCATS. See National Center for
Advancing Translation Sciences
NCNR. See National Center for Nursing
Research
NDNQI. See National Database of
Nursing Quality Indicators
Near-infrared spectroscopy, 419b
NEBHE. See New England Board of
Higher Education
Necessary relationship, 123
Need-Driven Dementia-Compromised
Behavior Model, 327
Negative likelihood ratio, 408
Negative linear relationship, 121, 121f,
561
Negative results, 594
Negatively skewed, 540
Neonate, as research subject, 165-166
Nested design
overview of, 248
Nested variable, 248
Network sampling, 366
Networking, 666
New England Board of Higher Education
(NEBHE), 19
NGC. See National Guideline
Clearinghouse
NHGRI. See National Human Genome
Research Institute
NIC. See Nursing Intervention
Classification
724 Index

Nightingale, Florence Nurses (Continued) Nursing research (Continued)

data collection and statistical analyses, critical appraisal of research by, framework for exploring, 2-11, 2f
17 452-453 HIPAA affecting, 170
nursing, influence on, 17-19 expert, 10 historical development of, 17-23
research by, 17 expertise, in critical appraisal of historical events influencing, 18t
NILT. See The Nursing Intervention research, 454, 454t in other countries, 80
Lexicon and Taxonomy intervention-based research conducted prediction of, 13-14
NINR. See National Institute of Nursing by, 323-324 requirements for, 2-3
Research novice, 10 in twenty-first century, 22-23
NLN. See National League for Nursing pain management role of, 327f Nursing Research, 19
NNHS. See National Nursing Home proficient, 10 Nursing science, 7-8
Survey research, levels participating in, 21 Nursing situation
NNT. See Numbers needed to treat Nursing questions, generation of, 81-82
NOC. See Nursing Outcome Classification body of knowledge, 1, 10 study problem formulation, 81-83
Nominal level, of measurement, 386 clinical expertise in, 12 Nursing-sensitive patient outcome
Noncoercive disclaimer, 178 critical appraisal of research (NSPO), 303-305
Nondirectional hypothesis, 147 implemented in, 452-454 Nutting Report, 19
Nonexperimental designs, 337-338 definition of, 1
Nongovernment funding, for research, education for, 19 O
667-669 in empirical world, 3 Objective. See also Research objectives
Nonparametric statistical analysis, 542 goal of, 11-12 formats for developing, 138
Nonparametric statistics, 542 middle-range theories used in, formulation of, 138-140
parametric statistics vs., to determine 127t-128t in qualitative research, 139-140, 150
differences, 580 NIC, work relating to, 230t in quantitative research, 138-139, 150
Nonpredicted results, significant and, 594 NOC, work relating to, 230t Observational checklist, 422
Nonprobability sampling methods outcomes research and, 302-306 Observational measurement, 421-422
in qualitative research, 358t, 364-367 philosophy and, 10-11 Observations
in quantitative research, 362-364 Nursing Care Report Card, 303 physiological measures by, 413
Nontherapeutic research, 160 Nursing Diagnosis Conference, 20 structured, 421-422
Nonverbal Pain Assessment Tool (NPAT), Nursing education, 19 Observed score, 383
413, 423f Nursing History Review, 68 OCS. See Omaha Classification System
Normal curve, 538, 539f, 555-557, 556f Nursing Intervention Classification (NIC), Odds ratio (OR), 487, 575, 575t
Normality, tests of, 541 229, 325 confidence intervals and, 575
Norm-referenced testing, 388 nursing work relating to, 230t interpretation of, 575
North American Nursing Diagnosis The Nursing Intervention Lexicon and in logistic regression, 576-577
Association (NANDA), 20, 325 Taxonomy (NILT), 325 Office for Human Research Protection
Notes on Nursing, 17 Nursing interventions. See also (OHRP), 162-163
Novice nurse, 10 Interventions Office for Protection from Research Risks
NPAT. See Nonverbal Pain Assessment definition of, 324 (OPRR), 191
Tool taxonomies, 325 Office of Laboratory Animal Welfare
NQF. See National Quality Forum variations in, 324-325 (OLAW), 190-191
NRS. See Numerical Rating Scale Nursing knowledge, acquiring, ways to, Office of Research Integrity (ORI),
NSPO. See Nursing-sensitive patient 8-10 188-189
outcome Nursing Outcome Classification (NOC), OHRP. See Office for Human Research
Nucleic acids, genetic advancements in 229, 230t Protection
measuring, 417-418 Nursing research OLAW. See Office of Laboratory Animal
Null hypotheses, 147-148 in 1900s, 19 Welfare
Numbers needed to treat (NNT), 333-334 in 1950s-1960s, 19 Omaha Classification System (OCS), 338
Numerator, 575 in 1970s, 20-21 The Omaha System, 325
Numerical Rating Scale (NRS), 430f in 1980s-1990s, 21-22 Oncology Nursing Society, 31, 305
Nuremberg Code, 160, 161b case study designs in, 223-224 One-group posttest-only design, 234, 235f
Nurse educators, critical appraisal of controlling outcome of, 13-14 One-group pretest-posttest design,
research by, 453 definition of, 1-2 234-236, 236f
Nurse researchers, critical appraisal of description of, 12-13 One-tailed test of significance, 545
research by, 453 effective enterprise, ensuring, 23 Online data collection, 508-513
Nurses evolution of, 17 Open label
advanced beginner, 10 explanation of, 13 blinding vs., 331-332
competent, 10 federal funds for, 21 designs, 337

Open label extension, 331-332
Operational definition, 43
Operational reasoning, 6
Operationalizing variable, 155-156
Operator, 105, 141
OPRR. See Office for Protection from
Research Risks
OR. See Odds ratio
Ordinal level, of measurement, 386-387
Orem’s model of self-care
as conceptual model, 125
theories relating to, 126t
ORI. See Office of Research Integrity
Outcome assessment instruments, 316t
Outcome reporting bias, 484
Outcomes
of care, 296
continuous, meta-analysis results for,
486
evaluation of, 308-309
intermediate, 342
observation process steps and, 342b
proximate, 296
Outcomes research, 3, 27-28. See also
Research
advanced practice nursing and,
305-306
classification of, 305
database usage in, 306-308
disseminating findings, 318
federal government involvement in,
300-302
focus of, 74-75
methodology for, 306-318
nursing practice and, 302-306
overview of, 294
purpose of, 83
statistical methods for, 315-318
strategies for, 308-312
theoretical basis of, 294-298
topics, problems, and purpose of,
92-93, 92t
Out-of-pocket cost, 298
P 90t-91t
Pain management theory, 142
Paired samples, t-tests for, 583-584
Paired t-test analysis technique, 580
Parallel group randomized controlled
trials, 335-336
Parallel synthesis, 492-493
Parallel-forms reliability, 390
Parameter, 537
population, 354
Parametric statistical analysis, 542
Parametric statistics, 542
nonparametric statistics vs., to
determine differences, 580
Paraphrasing, 110-111
Parent theory, 129
Parents and Newborn Screening Theory,
509f-512f
Partially controlled setting, 373
for research, 37
Participant-administered instruments, in
data collection, 507-508
Participant-observation, 65
Path coefficient, 227-228
Patient, definition of, 294-295
Patient outcomes evaluation, 316t
PCA. See Principle components
analysis
PDAs. See Personal digital assistants
Pearson, Karl, 34
Pearson’s product-moment correlation
coefficient, 561-564
calculation of, 562-563, 562t
coefficient of determination for, 564
effect size of, 564
results interpretation, 563-564, 563f
Peer, researcher interaction with, 76
People for Ethical Treatment of Animals
(PETA), 190
Perceived Stress Scale (PSS-10), 45b
Perceived view, 66
Periodical, 100
Permanent damage. See Damage,
permanent
Permission to participate, in study,
166
Person, 294-295
Personal digital assistants (PDAs),
514-515
Personal experience, 10
PETA. See People for Ethical Treatment
of Animals
Phenomena, 60
Phenomenological research, 27, 60-62
nursing science contribution from,
61-62
purpose of, 87
topics, problems, and purpose of,
The Phenomenology of Perception, 61
PHI. See Protected health information
Philosophical analysis
problem and purpose of, 93
topics, problems, and purpose of, 93t
Philosophical congruence, 465
Philosophical perspective, 57
Philosophy, 10-11
describes view of science, 57-58,
58f
guides criteria of rigor, 58-59, 58f
guides methods, 58, 58f
qualitative research and, 57-59
Index 725
Phones, for data collection, 514-515
Physical activity counseling, 237
Physiologic measures
electronic monitoring and, 415-417
new development of, 418-419
obtaining across time, 419-420
selection of, 420-421
Physiological measures, 402, 411-421
accuracy of, 402-406
direct, 413-414
error of, 402-406
indirect, 413-414
from laboratory tests, 414-415
by observation, 413
precision of, 402-406
by self-report, 412-413
sensitivity of, 404
PICOS format, 474-475
Pilot study, 46, 343-344, 523
Pilot testing, of interventions, 343-344
Placebo interventions, 330-331
Plagiarism, 188
Platykurtic, 541, 541f
Population-based study, 311
Populations, 351-352, 352f, 537-538
definition of, 44
heterogenous, 352-353
hypothetical, 352
parameter, 354
studies, 352
target, 351-352, 352f
Position papers, 100
Positional operator, 105-106
Positive likelihood ratio, 408
Positive linear relationship, 120-121,
121f, 561
Positively skewed, 540
Post hoc analysis, 545-546
Poster sessions, for research findings
presentation, 625-627, 626f
Posttest design. See Pretest and posttest
design
Posttest-only design, 234, 235f
Power, 367, 536
Power analysis, 219-220, 367
Practical research, 35
Practicality, as outcomes assessment
instrument, 316t
Practice effect, 249
Practice pattern, 297
profiling, 308-309
Practice style, influencing quality, 297
Practice theories, 125
Pragmatists, 387-388
Precision, 354
Predicted results, significant and, 594
Prediction, 13-14
Predictive correlational design, 227
726 Index
Predictive design
example of, 226f
purpose of, 226-227
Predictive validity, 573-574
Preference clinical trials, 336-337
Preferred Reporting Items for Systematic
Reviews and Meta-Analyses
(PRISMA) Statement, 472-473, 477f
Pregnancy, research during, 167
Premature infant study, 250
Preparation, Stetler Model of Research
Utilization to Facilitate Evidence-
Based Practice phase, 493-495
Preproposal proposal, content of, 644
Prescriptive theory, 327
Presentations, critical appraisal of
research, 453
Pretest and posttest design
with comparison group, 237, 237f
with removed treatment, 238-240, 239f
with reversed treatment, 240, 240f
with two comparison treatments,
237-238, 238f-239f
Pretest sensitization, 246
Pretest-posttest control group design, 245
example of, 245f
Primary prevention study, 254
Primary source, 101
Primary study strategies, 483b
Primordial cell, 295-296
Principle, 8
Principle components analysis (PCA),
566-568
example of, 567-568, 567t
factors generated from, 568
Principle of beneficence, 162
Principle of justice, 162
Principle of respect for persons, 162
PRISMA. See Preferred Reporting Items
for Systematic Reviews and
Meta-Analyses Statement
Prisoners, as research subjects, 169
Privacy, 169-171
invasion of, 169-170
Privacy Rule
de-identifying PHI under, 170
enactment of, 163
individually identifiable health
information and, 170
IRB, influence on, 187
Private funding, for research, 667-668
Probability, 196
Probability distribution, 563, 564f
Probability sampling methods, 357-362,
358t
Probability theory, 535
Probably statement, 123
Probing, 424
Problem. See Research problem Qualitative research (Continued)
Problem statement, 73 goals of, 74-75
Problematic reasoning, 6 introduction to, 57
Process, evaluation of, 296 literature review and, 98
Professional standards review meta-synthesis conduction of,
organizations (PSROs), 22 489-491
Proficient nurse, 10 methods of, 27
Profiling, practice pattern, 308-309 middle-range theory and, 125-126
Project grant proposals, 667 nonprobability sampling methods in,
Project team 358t, 364-367
formation of, 338 objective, formulation of, 139-140
work of, 338-339 objective, questions, hypothesis,
Proportionate sampling, 360 selection of, 150, 150t
Proposals perspective of, 57-59
critical appraisal of research, 453-454 philosophical orientation of, 60-61
revision of, 647 philosophical origin of, 24
verbal presentation of, 646-647 philosophy and, 57-59
Prospective cohort study, 219, 309 purpose of, 87-92
Protected health information (PHI), 163 quantitative research, comparing to,
authorization to use, 182-183, 184f 23-25
de-identifying under Privacy Rule, questions, generation of, 141-142
170 rigor and, 58-59
HIPAA, DHHS, FDA, comparison of, sample size in, 371-373
188t topics, problems, and purpose of,
Providers of care, 295 90t-91t
Proximate outcomes, 296 types of, 26b
PSROs. See Professional standards review uniqueness of conducting, 25
organizations variables in, 156
PSS-10. See Perceived Stress Scale Qualitative research proposal, content of,
Publication bias, 484, 485f 640-644, 641t
Publications, critical appraisal of research, communication plans, 643-644
453 inquiry method, 642-643
Published research, 627-632 introduction, 640-642
in books, 632 knowledge base, 643-644
duplicate publications, 632 limitations, 643-644
journal article, 628-631 philosophical and conceptual
manuscript preparation, 629-630 foundation, 642
manuscript revision requests, Qualitative research report, outline for,
630-631 614-618
manuscript submission for review, analysis plan, 617
630 discussion, 618
peer review, 630 dissertations, 619
query letter development, 629 introduction, 615
selection of, 628-629 methods, 616, 616b
in online journals, 631-632 results, 617-618, 618b
self-plagiarism, 632 theses, 619
Pure research, 34 Qualitative research synthesis, 29, 489
Purposive sampling, 365-366 Qualitative study
critical appraisal process for, 462-465
Q approach, 464
Q-sort methodology, 434-435 data collection, 464
Qualitative meta-synthesis, 29 data management and analysis, 465
Qualitative research discussion, 465
approaches to, 59-68, 60t evaluation summary, 465
characteristics of, 24t findings, 465
data for, 25 guidelines, 463-465
definition of, 3, 23 human study participant protection,
focus of, 24-25 464-465
 Index 727

Qualitative study (Continued) Quantitative research reports, 590, Randomization, 356-357

literature review, 463-464 603-614 Randomized blocking design, 206,
logic form of findings, 465 outline for, 604t 246-247
philosophical foundation, 464 analysis plan, 608 Randomized controlled trial (RCT),
prerequisites for, 463 data collection process, 607 26-28, 232, 251-254
problem statement, 463 design, 605-606 cluster, 336
purpose, 463 discussion, 613-614 crossover, 336
research questions, 463 framework, 604-605, 605f efficacy studies and, 329-330
sampling, 464 instruments, 607t of nurses role in pain management,
emergency contraceptive pills and, introduction, 604-605 327f
366-367 measures, 607 parallel group, 335-336
Qualitative study abstract, outline for, methods, 605-608 Range, as measure of dispersion, 554
622b reference citations, 614 Rating scales, 422, 429-430
Quality health care result presentation, 609-613, 609t, Ratio level, of measurement, 387
evaluation of, 294 610f, 611t-613t RCT. See Randomized controlled trial
theory of, 294, 295f review of literature, 604 RD. See Risk difference
Quantitative data analysis, stages of, 542 sample and setting, 606-607 Reality testing, 3-4
Quantitative research variables, 607t Reasoning, 6-7
characteristics of, 24t Quantitative study abstract, outline for, deductive, 7
concepts relevant to, 34-38 622b dialectic, 6
control in, 36-38, 37t Quasi-experimental designs, 231-244, inductive, 7
critical appraisal process for, 454-462, 335-337 logistic, 6-7
455b development of, 231 operational, 6
evaluation, 462 focus and purpose statement of, 244t problematic, 6
step identification, 455-458 Quasi-experimental research, 26 types of, 6
strengths in, 458-462 topics, problems, and purpose of, 88t-89t Recall, bases of, 310
weaknesses in, 458-462 variables in, 156 Recording interview data, 424
definition of, 3, 23 Quasi-experimental studies, 37 Recruiting research participants, 374-379,
focus of, 24-25 decision tree for, 259f 378t
goals of, 74-75 intervention development in, 228 Reference
introduction to, 34 manipulation in, 197 citation errors, 112, 112b
literature review of, 98-100, 99t process, steps of, 50b-53b literature review, verifying for, 112
methods of, 26-27 purpose of, 50 systemically recording of, 107-108
nonprobability sampling methods in, research design for, 43 Reference management software, 107-108
362-364 Query letter development, 629 Referencing, 388
objective, formulation of, 138-139 Questionnaires, 425-429 Referred journal, 628
objective, questions, hypothesis, data analysis, 429 Refusal rate, 355
selection of, 150, 150t development of, 426-428 Registered nurse (RN), 2
philosophical origin of, 24 validity of, 428-429 Regression analysis, 570
process, steps of, 38-43, 39f Questions independent variable types used in,
purpose of, 87 focus of, 140 574-575
qualitative research, comparing to, generating for research, 81-82, 150t dummy, 574-575
23-25 for qualitative research, 141-142 Regulation
questions, generation of, 140-141 for quantitative research, 140-141 in research, 159-163, 164t
rigor in, 36 research, formulation of, 42 set by DHEW, 162
sample size in, 367-371 Quota sampling, 364 Relational statement, 41, 116-117
topics, problems, and purpose of, characteristics of, 120-124
88t-89t R development of, 132
types of, 26b, 49-53 Random assignment, 205t, 206 examination of, 120-124
uniqueness of conducting, 25 to groups, 232-233 Relationships
Quantitative research proposal Random error, 384 associative, 144
content of, 637-640, 637t Random heterogeneity, 199 asymmetrical, 122
framework, 638 Random sampling, 205-206, 205t, causal, 123, 145
introduction, 638 356-357 characteristics of, 124t
methods, 638-640 Random sampling methods. See concurrent, 123
procedures, 638-640 Probability sampling methods contingent, 123-124
relevant literature review, 638 Random variation, 354 curvilinear, 121
example of, 647 Random-effects model, 486 example of, 122f
728 Index

Relationships (Continued) Research (Continued) Research (Continued)

deterministic, 123 ethical codes and regulations in, 159 natural setting for, 37
direction of, 120-121 ethical considerations for, 86 ORI role in, 188-189
necessity of, 123 ethnonursing, 65 outcomes, interpretation of, 48
negative linear, 121, 121f evidence, focus of, 12 partially controlled setting for, 37
positive linear, 120-121, 121f evidence-based practice and, 11-14, phenomenological, 27
probability, 123 159 practical, 35
sequential, 123, 123f experimental, variables in, 156 prediction of, 13-14
shape of, 121 explanation of, 13 priority, for clinical practice, 79-81
strength of, 121-122, 122f, 561t facility/equipment availability for, 86 pure, 34
substitutable, 123 feasibility of, 84-86 questions. See Questions
sufficient, 123-124 by Florence Nightingale, 17 reality testing using, 3-4
symmetrical, 122, 122f funding for recommendations for further, 599-600
Relative risk, 486-487 application requests, 670 regulations for, 164t
Relevant literature apprenticeship, 666 report, content of, 602-619
location of, 106-107 capacity, 665 settings, 22-23, 373-374
review of, 40-41 capital, 665 time commitment for, 84
Reliability, 45, 389-393 commitment level, 665 topics, 73, 81
alternate-forms, 390 contribution, 664-665 development of, 83-84
equivalence, 390-391 course attendance, 666 unethical treatment of subjects in,
interrater, 390 foundations, 669 159-160
as outcomes assessment instrument, government, 669-670, 669f utilization, 22, 493
316t grants types for, 667 variable, 151-152
parallel-forms, 390 grantsmanship, 666 definition of, 42-43, 151
split-half, 391 industry, 668-669 Research evidence
stability, 389-390 investigator-initiated proposals, conducting meta-analysis, 482-488
testing, 389 669-670 guidelines for synthesizing, 471-493
test-retest, 389 local, 667-668 aims of, 474, 474b
Reliable measure, 198 national nursing organizations, 668, clinical question formulation, 474
Repeated measures design, 583 668t comprehensive search conduction,
Replication nongovernment, 667-669 476
approximate, 77 private, 667-668 critical appraisal of studies,
concurrent, 77 program building, 663-665, 664f, 477-479
exact, 77 664t discussion section, 479-480
of studies, 77-78 proposal requests, 670 final report for publication, 480
systematic, 77 regional nursing research literature search criteria and
types of, 77 organizations, 667, 667t strategies identification,
Representative, 353-354 research committees, serving on, 474-476
Request for applications (RFA), 669 667 meta-analysis conduction, 479, 481t
Request for proposals (RFP), 669 sources for, 667-670 review results, 479
Research, 159 support of other people, 665-666 state purpose and objectives, 474
AIDS patients as subjects of, 169 workshop attendance, 666 study selection, 476
applied, 35-36 generating hypotheses for, 150t levels of, 30-31, 30f
benefits and risks of, 175-176, 175f highly controlled setting for, 37 methodologies for developing, 23-25
bias in, 197 historical, 27 synthesizing, process used to, 29t
compensation for participation, 178 historical development of, 17-23 Research findings
concepts, 151-152 historical events influencing, 18t audiences for communication of,
controlling outcome of, 13-14 hypothesis, 147-148 619-621
critical appraisal of. See Critical formulation of, 42 of healthcare professionals, 621
appraisal informed consent for, 166, 166t of nurses, 619-621
definition of, 1-2 legally and mentally incompetent of policy makers, 621
description of, 12-13 subjects, 165-169 strategies for, 619-621, 620t
designs levels participating in, 21 communication of, 602
algorithms for selecting, 256 methodology implications of, 599
definition of, 195 classification of, 25-28 presentation of, 621-627
selection of, 43-49 quantitative/qualitative research, 26b poster sessions, 625-627, 626f
types of, 215b, 335-338 misconduct in, 159, 187-190 verbal, 622-625
education of nurses participating in, 21 money commitment for, 84-85 publishing of, 627-632

Research in Nursing & Health, 20
Research objectives. See also Objective
formats for developing, 138
formulation of, 42, 138-140
in qualitative research, 139-140
in quantitative research, 138-139
Research participants, 351-352, 352f
recruiting of, 374-379, 378t
retaining of, 374-379
Research plan
evidence examination from, 590-591
implementing, 45-48
Research problem, 73-75
background for, 73
clarifying and refining, 82-83
clinical practice as source of, 75-76
definition of, 73
development of, 83-84
formulation of, 39-40, 81-83, 81f
literature, review of, 76-78
personal interest in, 82
researcher and peer interactions, 76
significance of, 73
sources of, 75-81
theories generating, 78-79
Research proposal, 635
content of, 636-644
condensed, 644
preproposal, 644
qualitative, 640-644
quantitative, 637-640, 637t
by students, 636-644
writing, 635-636
critical point identification, 636
depth of proposal, 636
esthetically appealing copy
development, 636
idea development logically, 635-636
Research purpose, 74-75
development of, 83-84
formulation of, 39-40, 81-83, 81f
generation of, 83
subject availability for, 85
Research Utilization to Facilitate EBP
Model, 22
Researcher
expertise of, 85
peer interaction with, 76
replication and, 78
study problem formulation, 81
Researcher-administered instruments, in
data collection, 507-508
Residual variable, 227
Responder analysis, 334-335
Responsiveness, as outcomes assessment
instrument, 316t
Retaining research participants, 374-379
Retention rate, 356
Index 729
Retrospective cohort study, 310-311 Sampling methods (Continued)
Retrospective descriptive studies, 338 systematic, 361-362
Retrospective study, 219 theoretical, 366-367
Review of relevant literature, 40-41 Saturation of data, 371
RFA. See Request for applications Scales, 429-434
RFP. See Request for proposals construction of, 442-445
Rigor, 36 using classic test theory, 442-444
qualitative research and, 58-59 using item response theory,
Risk difference (RD), 487 444-445
Risk ratio (RR), 486-487 using multidimensional scaling, 445
Risks, research, assessment of, 176 using unfolding theory, 445
Rival hypothesis, 199 Likert, 45, 45b, 388, 430-431
RN. See Registered nurse rating, 422, 429-430
Robustness, 581 semantic differential, 431-433, 433f
Role-modeling, 10 summated, 429
Roper, William, 22 translation of, 445-446
Rosenthal effect, 201 visual analog, 222-223, 433-434, 434f
Roy adaptation model, 125 Scannable forms, data collection and, 508,
RR. See Risk ratio 509f-512f
Scatter diagrams, 560
S Scatterplots, 560, 561f-562f
Sackett, David, 22 Science, 7-8
Sample size Scientific American, 17
in qualitative research, 371-373 Scientific method, 23
in quantitative research, 367-371 Scientific misconduct, 189-190
Samples, 351, 352f, 537-538, 537f Scientific rationale, for interventions,
attrition, 355-356 328-329
characteristics, 154 Scientific theory, 117
database as source of, 306-308, 306f, Screening test
307t accuracy of, 406
definition of, 44 sensitivity of, 406, 406t
description of, 544 specificity of, 406, 406t
methods of obtaining, 306-308 SDAT. See Senile dementia of Alzheimer
random, 356-357 type
representativeness, 353-354 Search engine, 106
volunteer, 357 Secondary analysis design, 254-255
Sampling, 37, 351 Secondary source, 101
disproportionate, 360 Selection
distributions, 538-541 interactions with, 200
error, 354-356, 354f, 557-558 process of, 199-200
frame, 357 of research subjects, 173-174
plan, 351, 357 treatment interaction and, 202
proportionate, 360 Selective sampling. See Purposive
theory, 351-357 sampling
Sampling criteria, 352-353 Self-care, HRQoL, 245-246
exclusion, 353 Self-determination
Sampling methods, 357 right to, 164-169
cluster, 360-361 violation of, 164-165
convenience, 363-364 Self-plagiarism, 632
network, 366 Self-report, physiological measures by,
nonprobability 412-413
in qualitative research, 358t, 364-367 Self-Report College Student Risk
in quantitative research, 362-364 Behavior Questionnaire, 427, 428f
probability, 357-362, 358t Semantic differential scale, 431-433,
purposive, 365-366 433f
quota, 364 Seminal study, 100
simple random, 358-359, 359t Senile dementia of Alzheimer type
stratified random, 359-360 (SDAT), 168
730 Index
Sensitivity
of physiological measures, 404
screening test of, 406, 406t
Sequencing, 123, 123f
Sequential explanatory strategy, 209-210,
209f
Sequential exploratory strategy, 210, 210f
Sequential relationship, 123
Sequential transformative strategy,
210-211, 211f
Serendipitous results, 595
Serendipity, 530
Serial, 100
Settings, 373
for control, 37
for experimental study, 87
highly controlled, 373-374
natural, 373
partially controlled, 373
research, 37, 373-374
for treatment, 202
SF-36. See Short Form (36) Health survey
Sham interventions, 330-331
Shape, of relationships, 121
Shapiro-Wilk’s W test, 541
Short Form (36) Health survey (SF-36),
frequency distributions of, 556, 556f
Sigma Thata Tau, 20
Directory of Nurse Researchers,
440-441
Simple hypothesis, 146-147
Simple interrupted time-series design,
241f, 242-243
Simple linear regression, 570-573
Simple random sampling, 358-359, 359t
Single-blinded, 331
Situated freedom, 61
Situation, nursing
questions, generation of, 81-82
study problem formulation, 81
Skewed, 540
distribution, 540
negatively, 540
positively, 540
Skewness, 540, 540f
Slope of line, 570
Small area analysis, geographical analysis
as, 311
SMD. See Standardized mean difference
Smoking, lung damage from, 7-8
Smoking status
frequency table of, 552t
histogram of, 552f
SMR. See Standardized mortality ratio
Snowball sampling. See Network
sampling
SORT. See Standardized Reporting of
Trials
Sources
appraisal and analysis of, 108-109
comprehension of, 108
reading and critiquing, 108
selection of, 490
skimming, 108
sorting by relevance, 109-110
synthesis of, 110, 110f
Southern Regional Education Board
(SREB), 19
Spearman’s rank-order correlation
coefficient, 564
Specific proposition, 124
Specificity, screening test of, 406, 406t
Spielberger State-Trait Anxiety Inventory,
383, 385
Split-half reliability, 391
Spurious correlations, 565
SREB. See Southern Regional Education
Board
Stability reliability, 389-390
Stage-based interventions, 325
Standard deviation, as measure of
dispersion, 555
Standard error of mean, 557
Standardized mean difference (SMD),
484, 484f, 486
Standardized mortality ratio (SMR), 310
Standardized Reporting of Trials (SORT),
251
Standards for Privacy of Individually
Identifiable Health Information, 163
Standards of care, 296-297
Statement analysis, 132
Statement hierarchy, 124
Statistical analyses, measurement levels
for, 387
Statistical conclusion validity, 198-199
Statistical control, 205t, 207
Statistical hypothesis, 147
Statistical power, 536-537
Statistical procedures, choice of
appropriate, 546-548, 547f
Statistical significance, clinical
importance vs., 537
Statistical tests, violated assumptions of,
198
Statistical theory, concepts of, 534-542
classical hypothesis testing, 535-536
clinical importance, 537
confidence intervals, 541-542
descriptive statistics, 538
inference, 537
inferential statistics, 542
normal curve, 538, 539f
populations, 537-538
probability theory, 535
samples, 537-538, 537f
Statistical theory, concepts of (Continued)
sampling distributions, 538-541
statistic types, 538
statistical significance, 537
Type I error, 535-537, 536t
Type II error, 536-537, 536t
variation, 541
Statistics, 537
data deviation exploration using,
554-558
confidence intervals, 558
degrees of freedom, 558
measures of dispersion, 554-558
normal curve, 555-557
sampling error, 557-558
data summarizing with, 550-554
central tendency measures, 552-554
frequency distributions, 550-552
descriptive, 538
inferential, 538, 542
nonparametric, 542
parametric, 542
types of, 538
Stetler Model of Research Utilization to
Facilitate Evidence-Based Practice,
474, 493-496
phases of, 493
application, 495-496
comparative evaluation, 495
decision making, 495
evaluation, 496
preparation, 493-495
translation, 495-496
validation, 495
Stratification, 205t, 207
Stratified random sampling, 359-360
Strength, 121-122
Structural analysis, validity from, 400
Structural equation modeling, 227-228
Structure
evaluation of, 298-300
studies examining aspects of, 299t
Structured interviews, 422-424, 425b
Structured observations, 421-422
Studies
appraisal of, 491
approval for, 644-647
political factors in, 646
preparation for review committees,
645-646
process of, 645
social factors in, 646
benefits and risks of, 175-176, 175f
classification of, opinions about, 231
confirmatory, 370
design, selection of, 256
development/implementation of, 37
ethical considerations for, 86

Studies (Continued)
as exempt from review, 183-185, 185b
exploratory, 370
facility/equipment availability for, 86
feasibility of, 84-86
money commitment for, 84-85
researcher expertise for, 85
scope of, 371-372
statistical procedures choice for,
546-548, 547f
by students, critical appraisal of, 452
subject availability for, 85
subject knowledge of, 38
time commitment for, 84, 122
types of, 369-370
variables, measurement of, 38
Study design, 195, 372-373
development and implementation of,
questions to direct, 207-208
Study outcomes. See Outcomes
Study protocol, 645
Study section flow chart, 478f
Study validity, 197-202
threats to, 345b, 347
Study variables, 138
Subject attrition, 200
Subject terms, 104
Subjects, 351-352, 352f
AIDS patients as, 169
animals as, 190-191
as autonomous agent, 164
availability for research, 85
of care, 294-295
children as, 166-167
coercion of, 164
comprehension of, 178-180
confidentiality of, 171-173
controlling equivalence of, 203
deception of, 165
discomfort, protection from, 174-175
fair treatment of, 174
harm, protection from, 174-175
informed consent from, 176-183
legally and mentally incompetent,
165-169
National Commission for Protection of,
162-163
neonate as, 165-166
permanent damage to, 175
prisoners as, 169
selection of, 173-174
terminally ill as, 169
unethical treatment of, 159-160
Substitutable relationship, 123
Substituted judgement standard, 168
Sufficient relationship, 123
Summary, 112
Summated scales, 429
Index 731
Support systems, access to, 530-531 Therapeutic nursing intervention, 228-231
academic committees, 530 Therapeutic research, 160
colleague support, 531 ThermoSENSOR, 416-417, 416f-417f
data safety and monitoring board, 531 Threats
institutional support, 531 designs to reduce, 203
Survey, 224 to interrupted time-series designs, 241
Symbolic interaction theory, 63 to statistical conclusion validity, 198
Symbolic meanings, 63 Time factors, data collection and, 515-516
Symmetrical curve, 539-540, 540f Time-dimensional design, 218-223
Symmetrical relationship, 122 Time-lag bias, 484
Symmetry, 539-540 TM. See Telemonitoring; Treatment
of relationships, 122, 122f mechanism trial
Systematic bias, 354-356 Tolerability, 251
Systematic error, 384, 384f Topic, nature of, 372
Systematic replication, 77 Tradition, 9
Systematic reviews, 28-29 Transitional Care Model, 650f
biases for, 484-485 Transitional care nurse practitioner
critically appraising checklist for, 473t (TCNP), intervention protocol for,
guidelines for, 472-482 655-656
mixed-methods, 29-30, 492-493 Translation, Stetler Model of Research
basic structure for, 493 Utilization to Facilitate Evidence-
published, critical appraisal of, 480-482 Based Practice phase, 495-496
Systematic sampling, 361-362 Translational research, introduction to,
Systematic variation, 354-356 503
Treatment
T compensatory equalization and rivalry
Target population, 351-352, 352f of, 200
Taxonomy, nursing intervention, 325 control of, 203-204
TCNP. See Transitional care nurse different, interaction of, 202
practitioner diffusion or irritation of, 200
Telemonitoring (TM) history and, 202
applied research using, 35-36 removed, pretest and posttest designs
effects of, 51f with, 201, 239f
intervention study using, 37-38 reversed, pretest and posttest designs
program, for BP in African Americans, with, 201, 240f
35, 37-38, 50, 51f selection and, 202
quasi-experimental study using, 50-51 setting and, 202
Tentative theory, 132 testing and, 202
Terminally ill, research on, 169 tolerability of, 251
Testable hypothesis, 149 two-treatment design, 237-238,
Testing effect, 199 238f-239f
Test-retest reliability, 389 Treatment fidelity, 332
Textbooks, 100 Treatment mechanism (TM) trial, 251
Theoretical limitations, 598 Treatment partitioning
Theoretical literature, 100 cohort designs with, 222
discussion of, 111 cross-sectional study with, 222f
Theoretical sampling, 366-367 longitudinal design with, 223f
Theories, 7, 41, 117 Trend design
definition of, 117 example of, 221f
descriptive, 327 purpose of, 235f
intervention, 327 Trial and error, 9-10
Orem’s model of self-care, 126t Triple-blinded, 331
parent, 129 True negative, 406
practice, 125 True positive, 406
prescriptive, 327 True score, 383
research problems and, 78-79 Truncation, 225
scientific, 117 t-tests, 371, 580-584
Theory-based interventions, 327-328 for paired samples, 583-584
732 Index
Tuskegee Syphilis Study, 160-161
permanent damage to subjects, 175
Twenty Thousand Nurses Tell Their Story,
19
Two-tailed test of significance, 545
Type I error, 535-537, 536t
Type II error, 536-537, 536t
Type III error, 338
U in experimental research, 156
Unexpected results, 595
Unfolding theory, scale construction
using, 445
Ungrouped frequency distribution,
550-551
Urn randomization, 232
V operationalizing, 155-156
Validation, Stetler Model of Research
Utilization to Facilitate Evidence-
Based Practice phase, 495
Validity, 45, 393-402, 393f
construct, 200-202
content, 394-397
from contrasting groups, 401
convergent, 400-401
from discriminant analysis, 401
divergent, 401
from event prediction, 401-402
external, 202
face, 394-397
from factor analysis, 398-400
internal, 199-200
of questionnaires, 428-429
statistical conclusion, 198-199
of structural analysis, 400
study, 197-202
successive verification of, 402
types of, 198
Variables
of associative hypothesis, 144-145
of causal hypothesis, 145-147
concept as, 138, 151-152
confounding, 152, 342
definition of, 42-43, 138-139, 150-151
demographic, 154, 154t
dependent, 145, 151
environmental, 153-154
extraneous, 152-154, 199, 204-207,
342
identifying and defining, 150-155
independent, 145, 151
mediator, 154-155
moderator, 154-155
number of, 355
in qualitative research, 156
in quasi-experimental research, 156
study, 138
types of, 151-155
Variance, as measure of dispersion,
554-555, 555t
Variance analysis, 315-317
Variation, in statistical theory, 541
VAS. See Visual analog scale
Verbal presentation, of research findings,
622-625
oral research presentation development,
624-625
question response, 625
receiving acceptance as presenter,
622-624
research report delivery, 625
Visual analog scale (VAS), 222-223,
433-434, 434f
Voluntary consent, 180
Volunteer samples, 357
W
Walter Reed Army Institute of Research,
19
Washout period, 249
Ways of Coping-Cancer Version
(WOC-CA) scale, 211-212
Wennberg’s design, 337
Western Interstate Commission on Higher
Education (WICHE), 19, 21
Western Journal of Nursing Research,
20
WHO. See World Health Organization
WICHE. See Western Interstate
Commission on Higher Education
Wilcoxon signed-rank test, 584
Willowbrook Study, 161-162
WOC-CA. See Ways of Coping-Cancer
Version scale
WOCF. See Worst observation carried
forward
Wong-Baker FACES Pain Rating Scale,
382, 430f
World Health Organization (WHO), 80
World Wide Web. See Internet
Worldviews on Evidence-Based Nursing,
22
Worst observation carried forward
(WOCF), 334
X
X2. See Chi-square test
Y
y-intercept, 570
Z
z value, distribution of, 557f
Zelen’s randomized consent design,
337
This page intentionally left blank
This page intentionally left blank
This page intentionally left blank
Strongest or Best Research Evidence

Systematic Review of experimental studies (well designed randomized

controlled trials [RCTs])

Meta-analyses of experimental (RCT) and quasi-experimental studies

Integrative Reviews of experimental (RCT) and quasi-experimental

Single Experimental study (RCT)

Single Quasi-experimental study

Meta-analysis of correlational studies

Integrative Reviews of correlational and descriptive studies

Qualitative Research Metasynthesis and Metasummaries

Single Correlational study

Single Qualitative or Descriptive study

Opinions of respected authorities based upon clinical evidence, reports

of expert committees

Weakest Research Evidence

Levels of research evidence

Methods of Statistical Data Analyses for Quantitative Nursing Research: Quick-Access Chart
Independent groups,
two variables
Dependent groups,
two variables
Dependent groups,
three or more variables
Bivariate Data Analyses
Two Samples Three or More Samples
Nominal data Chi-square, p. 587 Chi-square, p. 587
Odds ratio, p. 575
Logistic regression, p. 576
Ordinal data Spearman rank-order correlation
coefficient, p. 564
Kendall’s tau, p. 564
Mann-Whitney U, p. 588
Interval or ratio data t-Test for independent samples, p. 580 Analysis of variance, p. 584
Pearson’s product-moment correlation Multiple regression, p. 573
coefficient, p. 561
One-way analysis of variance, p. 584
Simple linear regression, p. 570
Bland and Altman plots, p. 565
Cox proportional hazard regression, p. 577
Interval or ratio data t-Test for related samples, p. 583
Multivariate Data Analyses
Interval or ratio data Multiple regression, p. 573
Factor analysis and principal components
analysis, p. 566

TABLE 2-3  Processes Used to Synthesize Research Evidence

Synthesis
Process Purpose of Synthesis
Systematic Use of specific, systematic methods to identify,
review select, critically appraise, and synthesize
research evidence to address a particular
problem in practice (Craig & Smyth, 2012;
Higgins & Green, 2008).
Meta-analysis Synthesis or pooling of the results from several
previous studies using statistical analysis to
determine the effect of an intervention or the
strength of relationships (Higgins & Green,
2008).
Meta-synthesis Systematic compiling and integration of
qualitative studies to expand understanding
and develop a unique interpretation of the
studies’ findings in a selected area (Barnett-
Page & Thomas, 2009; Finfgeld-Connett,
2010; Sandelowski & Barroso, 2007).
Mixed methods Synthesis of the findings from independent
systematic studies conducted with a variety of methods
review (both quantitative and qualitative) to
determine the current knowledge in an area
(Higgins & Green, 2008).
Analysis for
Types of Research Included in the Achieving
Synthesis (Sampling Frame) Synthesis
Usually includes quantitative studies with Narrative and
similar methodology, such as statistical
randomized controlled trials (RCTs), and
can also include meta-analyses focused
on an area of the practice problem.
Includes quantitative studies with similar Statistical
methodology, such as quasi-experimental
and experimental studies focused on the
effect of an intervention or correlational
studies focused on relationships.
Uses original qualitative studies and Narrative
summaries of qualitative studies to
produce the synthesis.
Synthesis of a variety of quantitative, Narrative
qualitative, and mixed methods studies.