APA

Dictionary
of
Statistics
and Research
Methods
 APA
Dictionary
of
Statistics
and Research
Methods
Sheldon Zedeck, PhD
Editor in Chief

American Psychological Association

Wa.shington, DC
Copyright 2014 by the American Psychological Association. All rights reserved. Except as
permitted under the United States Copyright Act of 1976, no part or this publication may be
reproduced or distributed in any form or by any means, including, but not limited to, the
process of scanning and digitization, or stored in a database or retrieval system, without the
prior written permission o f the publisher.
Published by
American Psychological Association
750 First Street, NE
Washington, DC 20002
wvm.apa.org
To order
APA Order Department
P.O. Box 92984
Washington, D C 20090-2984
Tel: (800) 374-2721; Direct: (202) 336-5510
Fax: (202) 336-5502; TDD/TTY: (202) 336-6123
Online: www.apa.org/pubs/books/
E-mail: order@apa.org
In the U.K., Europe, Africa, and the Middle East, copies may be ordered from
American Psychological Association
3 Henrietta Street
Covent Garden, London
WC2E 8LU England
AMERICAN PSYCHOLOGICAL ASSOCIATION STAFF
Gary R. VandenBos, PhD, Publisher
Julia Frank-McNeil, Senior Director, APA Books
Theodore J. Baroody, Director, Reference, APA Books
Patricia D. Mathis, Senior Reference Development Editor, APA Books
Typeset in Aylesbury, England, by Market House Books, Ltd.
Printer: United BOOK Press, Baltimore, M D
Cover Designer: Naylor Design, Washington, DC
Library of Congress Cataloging-in-Publication Data
APA dictionary of statistics and research methods / Sheldon Zedeck,
editor in chief; Lisa L. Harlow, Shelley A. Blozis, A. T. Panter,
associate editors,
pages cm
ISBN-13: 978-1-4338-1533-1
fSBN-10: 1-4338-1533-8
1. PsychologyResearchMethodology. 2. StatisticsMethodology. 1.
Zedeck, Sheldon.
BF76.5.A7263 2013
150.72'7dc23
2013017131

British Library Cataloguing-in-Publication Data

A CIP record is available from the British Library.

The citation for this publication is American Psychological Association. (2014). APA dictionary
of statistics and research methods. Washington, DC: Author.

Printed in the United States of America

First Edition

DOI: 10.1037/14336.000
1 dedicate this volume to the memory of my parents, Judith and Hyman; to
the support and love from my brother Morris and his wife Ellen; to my
mentors and colleagues who influenced my data analytic and research
perspectives, particularly Carol Vale, Robert M. Guion, Patricia Cain
Smith, Edwin E. Ghiselli, and Geoffrey Keppel; to my wife Marti, our
children, and their spousesCindy and Jason Singer, Jason and Stacey
Skura Zedeck, and Tracy; and to my grandchildren, Molly, Ella, Lilly,
Aidan, and Noah, all of whom have played a statistically and practically
significant role in my life.
Sheldon Zedeck

I dedicate this volume and boundless gratitude to previous statistical

mentors who got me started, particularly Peter Bentier, George Huba, and
Edward Stearns; and to my husband, Gary, and daughter, Rebecca, who
keep me going.
Lisa L. Harlow

I dedicate this volume to my mentor and the driving inspiration of my

work in quantitative methods, Robert Cudeck, and to the three most
precious people in my life; Richard, Faustino, and Maximiliano, who give
endless support and encouragement.
Shelley A. Blozis

1 dedicate this volume to my most inspirational quantitative

mentors, Jeffrey S. Tanaka and Lyle V. Jones, and to my loving
familyNechama, Yaakov, George, Danielle, Michaela, Jonathan, Joshua,
Giulia, Dara, Sarajane, and most of all my dad.
A. T. Panter
 Contents

Preface ix
Editorial Staff xv
About the Editorial Board xvi
Quick Guide to Format xvii

APA Dictionary of Statistics and 1

Research Methods

Appendixes
Abbreviations and Acronyms 423
Entry Illustrations 428
Overview of Research Design 431
Considerations
Symbols 432
 Preface
The APA Dictionary of Statistics and Research Methods builds on a strong core
of lexicographical reference works published in the past seven years by APA
Books. APA inaugurated this collection with the parent APA Dictionary of
Psychology (2006)the culmination of some ten years of research and
lexicographic activity. A resource of 25,000 entries, it won critical
endorsement from both the publishing and reference library communities.
Two derivative works followed almost immediately: the abridged APA
Concise Dictionary of Psychology (2008; available both in print and as an app
for iPhone, iPad, and Android), whose 10,000 entries were selected for a
more general readership, and the student version, the APA College Dictionary
of Psychology (2009), which features 5,000 entries that are essential for
psychology advanced placement high-school students and university-level
psychology majors alike.
A third derivative (and the first With a subdisciplinary focus) followed in
2012; the APA Dictionary of Clinical Psychology (available in print and via
Kindle). Its 11,000 entries directly address the needs of clinical
psychologists, whether they work in health and mental health clinics, in
independent and group practices, or as consultants to professionals in such
fields as medicine, law, social work, and consumer relations. The content
focuses on clinical training, clinical supervision, and the diagnosis,
treatment, and prevention of emotional and behavioral disorders, with
definitions describing the biological, developmental, social, and individual-
difference underpinnings of health and mental health.
A fourth derivative, again subdisciplinary in nature, was published in
2013: The APA Dictionary of Lifespan Developmental Psychology is specifically
tailored to offer scholars and students balanced coverage (in some 7,500
entries) in such core areas as developmental theory; genetics; and the
biosocial, cognitive, and psychosocial perspectives of development.
Entries span all the stages of life, from birth through childhood;
through adolescence; and through the early, middle, and late periods of
adulthood.
To some degree, editorial work on each of the derivatives noted above
resulted in changes to APA's overall collection of dictionary terms, whether
those changes are revisions or updates of entries that appeared in the parent
dictionary or whether they are entirely new entries added to offer a more
complete representation of the evolving lexicon of psychology.
The APA Dictionary of Statistics arid Research Methods, however, has been
something of a more consciously advanced enterprise in the development
of our series. Although partially derived from the original corpus of terms in
the parent dictionary, the work may more accurately be seen as APA's first
endeavor to create an entirely new reference in this subareaone that
stands on its own and one that reaches out beyond psychology in its
applicability to allied social, behavioral, and health sciences.

Evolution of the APA Dictionary of Statistics

and Research Methods
In March 2009, the APA publisher began a search for an editor in chief
(EIC) for a long-planned dictionary that would focus specifically on the two
vital and related areas of statistics and research methods. The goal was to
generate a focused specialty product that would use the content-specific
corpus from the APA Dictionary of Psychology as a base but more than double
the coverage to some 4,000 entries. Sheldon Zedeck (see About the Editorial
Board, p. xvi) was the publisher's first choice to take the task in hand.
Zedeck began the process by identifying potential associate editors (AEs),
both to help develop the project and to serve as peer reviewers and entry
drafters. Because the dictionary is intended to cover all of psychology, and
whereas the ElC's fields of specialization are industrial and organizational
psychology and psychometrics, Zedeck undertook the task of identifying
candidates who would bring diverse breadth and expertise to the project. To
do so, he conducted a survey of all of the then-current editors of APA
scholarly journals, asking them, in turn, to identify colleagues and journal
peer reviewers who were methods and data analysis experts and upon
whom the journal editors relied to review submitted articles that are
especially centered on methodology or data analysis issues or that require
special expertise from these perspectives. Specifically, the EIC requested that
the journal editors identify, from among their most reliable reviewers, those
whom they considered (a) most knowledgeable across different strategies
and approaches, (b) highly articulate in explaining (communicating)
shortcomings in strategy or analysis in plain English, (c) able to recognize
alternative methods and strategies that would be more appropriate than
those presented in article manuscripts, and (d) up-to-date in the latest
methods and data analytic strategies.
The result of this survey was the identification and recruitment of five
distinguished scholars: Daniel J. Bauer, Shelley A. Blozis, Lisa L. Harlow, Jay
Myung, and A. T. Panter. All hve editors were involved in the initial
determination of which entries would constitute the corpus of this work
(see Selection of Terms to Define, p. xi). Drs. Bauer and Myung were not
able to continue beyond this first stage due to other commitments. Drs.
Harlow, Blozis, and Panter, however, continued their roles as AEs
throughout the entire editorial process (see their brief biographical
descriptions, p. xvi) and were involved both in the review and editing of
entries inherited from the parent APA Dictionary of Psychology and in the
drafting of entirely new entries.
Selection of Terms to Define
The first task for the editorial team was to review more than 8,000
headwords (i.e., terms to define) collected for consideration by APA
Reference staff from the original APA Dictionary of Psychology, other
dictionaries, handbooks, articles in the APA journal Psychological Methods,
and similar scholarly sources. The EIC divided the complete list of these
potential headwords into three sections and assigned two members of the
editorial team per section to assess which headwords should be included in
the new dictionary. Each member independently ranked the headwords for
their importance using the following coding system;
1. a basic concept in the field that is essential to understanding the
topic area and must be included not only in the current specialty
dictionary but also in any revised edition of the larger parent
dictionary;
2. a more focused concept in the field that enhances understanding and
should be included in the current specialty dictionary but not
necessarily in the larger parent dictionary;
3. an outdated, overly detailed, or narrowly applicable concept that is
inappropriate for the desired overall length and intended level of
readership and thus should be omitted from the final product but
nonetheless retained in the "consideration pool" for possible future
use; or
4. a term that is not related to psychology in any way and should be
deleted altogether from the consideration pool.
The rank codes were further defined for the editorial team as follows:
Rank 1: Essential vocabulary. Identify as Rank 1 all essential terms
in current use that have a precise rneaning to psychological research and
the statistical analysis of psychological data. Thus, Rank 1 terms represent
basic concepts without which one could not understand these fields. A
term's use may be determined from its current prevalence in both scholarly
and general psychological literature, its supporting relevance to key terms
in the same literature, or both. Generally, Rank 1 terms are superordinate,
comprising vocabulary that is the basis from which other words derive or to
which they refer (relate back). \
Also identify as Rank 1 any current terms in the culture at large that
nonetheless significantly factor into psychological research and data
analysis, even if these terms are controversial, misunderstood, or misused.
For example, the term power is a general one with a commonly understood
meaning, but it is applied much more specifically within data collection
and analysis, both to refer to a mathematical notation indicating the
number of times a quantity is multiplied by itself and to refer to the
probability that the null hypothesis will be rejected when the alternative
hypothesis is true.
Rank 2: Useful vocabulary. Identify as Rank 2 all specific, more
focused terms in current use that have a precise meaning to psychological
research and the statistical analysis of psychological data. In other words,
Rank 2 terms are those that provide more detailed knowledge of these fields
but without which one still has a reasonably complete theoretical
understanding. For example, complex alternative forms of the basic analysis
of variance, such as the betweeti-subjects analysis of variance and the fixed-
effects analysis of variance, typically should be ranked a 2.
Rank 2 terms may also be those that are in the process of falling out of
usage and generally are now understood by another name (e.g., Brown-
Spearman formula, which is now typically referred to as the Spearman-Brown
prophecy formula), those that have major historical importance to the
development of the field or of its constructs (e.g., hypothetico-deductive
method, which often appears in student or other literature dealing with the
history and philosophy of psychology and related domains), or those that
are so new as not to have yet gained widespread usage in the field (e.g.,
experimenter biosocial effect).
Rank 3: Unnecessary vocabulary. Identify as Rank 3 any terms that
are too specialized or otherwise overly concentrated to be widely applicable
within the field. For example, the term between-subjects factorial analysis of
variance, a very specific combination of several other types of analysis of
variance techniques, should be ranked a 3, because the separate entries for
its component concepts (analysis of variance, between-subjects analysis of
variance, and factorial analysis of variance) on their own are sufficient to
enable a reader to understand the meaning of this more complicated,
highly detailed concept.
Also identify as Rank 3 all terms that are obsolete within or tangential to
psychological research and the statistical analysis of psychological data. In
other words, Rank 3 terms are those that are no longer in use in these fields
at all, rarely if ever appearing in current literature (e.g., aftertest), and those
belonging primarily to other fields (e.g., quality management, economics)
with minimal or no connection to psychology or its domains (e.g., integer
programming).
During the ranking process, the editors researched various sources
(particularly APA's PsycINFO database) to identify terms that were
frequently used in the past decade. The EIC also responded to questions and
comments from APA Reference and made general notes that would
subsequently be helpful as definitions were finalized.
As each member of the pair of reviewing editors completed his or her
batch of entries, the rankings were collated and the entire team was
provided with the collective results for the purpose of resolving ranking
discrepancies. If after further review the AEs still could not agree, the EIC
resolved the ranking discrepancies, generally through mediation among the
rankers.
In addition to ranking existing headwords, each editor was asked to
identify new terms or words not on the list provided by APA Reference.
Identification of such terms was achieved by relying on the editorial board
members' expertise, as well as by examining vocabulary from a variety of
professional sources, including scholarly journals, recent monographs,
current handbooks, desk references, and other academic publications. These
terms, too, were ranked. The entire selection process resulted in the retention of
somewhat fewer than 5,000 headwords.
Editorial Process
In autumn 2010, the editorial board began to review, edit, and write
definitions according to the following directives:
1. provide the what, why, and how for each term;
2. provide examples where possible;
3. use plain English instead of jargon; and
4. avoid the overuse of complicated formulas and equations.
A fifth directive was to consider whether an entry would benefit from a
graphical or tabular illustration. Although there are clearly many entries
that would benefit from graphical display, due to space considerations APA
Reference ultimately researched and selected approximately 110 entries for
illustration. j
In spring 2012, the board presented its final set of definitions to APA
Reference and Market House Books Ltd, the dictionary compiling
professionals in the United Kingdom with w'hom APA Reference has worked
on all its dictionaries to date. APA and Market House jointly began a final
editorial and preproduction phase, generally cleaning up the copy in terms
of redundant compounds, variant terms, and proper lexicographical format.
In the end, some 4,080 entries were finalized for inclusion.

Future Perspectives
It is commonplace among those in the field of lexicography to assert that a
dictionaryany dictionary, but especially a first editionis a work-in-
progress, the only limitations upon the future of which are the publishing
company's continued interest in keeping the work alive and current and its
ability to locate and task the appropriate people to provide solid content
and stylistic expertise for revised and new editions.
We therefore issue this first edition of the APA Dictionary of Statistics and
Research Methods with an eye to these potential limitations and with the
recognition that, although all parties have collectively done the best they
couldgiven the temporal imperatives of scholarly publishingthere is
always room for correction and improvement. We are, in fact, proud of our
work and send it out into the world confident that it is a very strong start,
that it offers greater and better coverage than the competition of which we
are aware, and that we fully intend to keep the work a living project in
future generations.
To assist us in this task, we invite you to contact APA Reference at
books@apa.org to note errors of omission, inaccuracies, infelicities of
phrasing, new vocabulary, and omitted senses. As always, we welcome your
thoughtful appraisal and suggestions.
 Acknowledgments
The development and production of a dictionary requires an extraordinary
amount of effort and collaboration with many participants. Our
experiences as the editor and as publisher of this dictionary benefitted
immensely from an incredible group of colleagues and staff who devoted
many hours in guidance, consultation, demonstrated patience, fortitude,
and energy. It is difficult to express our true appreciation to the many
participants in a short space, but, to start, we want to acknowledge the
associate editors: Shelley Blozis, Lisa Harlow, and Abigail Panter. This team
contributed invaluable wisdom to the generation of the plan for the
dictionary, spent endless hours reviewing potential entries, and spent even
more hours writing and revising definitions. We also want to acknowledge
the contribution of the consulting editors, Dan Bauer and Jay Myung, for
their efforts in reviewing entries for possible inclusion in the dictionary. On
the project genesis and development side, we owe gratitude and thanks to
Ted Baroody (Director, Reference, APA Books) and to Trish Mathis (Senior
Reference Development Editor, APA Books), who worked with the editorial
team from day one, helping to generate the structure for the dictionary,
reviewed drafts on process and procedure, and provided wise feedback on
any issue requested throughout the project; they performed their tasks with
grace and support that resulted in the final product being a personally
rewarding partnership. Finally, we thank our consummately professional
U.K. editorial and production team at Market House Books. We want to
thank all of these participants for their support, encouragement, and
wisdom.
Sheldon Zedeck, PhD
Editor in Chief

Gary R. VandenBos, PhD

Publisher
 Editorial Staff

Editor in Chief
Sheldon Zedeck, PhD

Associate Editors
Lisa L. Harlow, PhD
Shelley A. Blozis, PhD
A. T. Panter, PhD :

Consulting Editors
Daniel J. Bauer, PhD
Jay Myung, PhD

Senior Editor (American Psychological Association)

Patricia D. Mathis

Senior Editors (Market House Books, Ltd)

Jonathan Law
Elizabeth Martin

Assistant Editor (American Psychological Association)

Kristen L. Knight ^

Editor in Chief, APA Dictionary of Psychology

Gary R. VandenBos, PhD
 About the Editorial Board

Sheldon Zedeck, PhD, is a professor of psychology and of the Graduate School at

the University of California, Berkeley, and the 2010 recipient of The Berkeley Citation
for distinguished service to the campus. He is an industrial/organizational
psychologist who has published articles on the topics of moderator variables, selection
and validation, test fairness, high-stakes testing, statistical banding, performance
appraisal, assessment centers, stress, and work and family issues. He has coauthored
texts on data analysis and measurement theory. He is former editor of the Joumal of
Applied Psychology (2003-2008) and the editor in chief for the APA Handbook of
hidustrial and Organizational Psychology (2010). His research on law school admissions
received the 2011 Smashing Bias Research Award (with M. Shultz) from the Level
Playing Field Institute for the Promotion of Outstanding Research on Ensuring
Fairness From the Classroom to the Boardroom.

Lisa L, Harlow, PhD, is a professor of psychology at the University of Rhode Island,

whose focus is on increasing interest, retention, performance, and diversity in
quantitative science. Since obtaining her doctorate in 1985 from the University of
California, Los Angeles, she has authored more than 75 publications on multivariate
methods and applications. She is editor of Psychological Methods (2014-2019), current
editor of the Multivariate Applications Series, and former associate editor of Structural
Equation Modeling. She is also a past president of APA Division 5 (Evaluation,
Measurement, and Statistics) and of the Society of Multivariate Experimental
Psychology. Her honors include the Jacob Cohen Award for Distinguished
Conthbutions to Teaching and Mentoring; a Distinguished Fellowship at the Institute
for Advanced Study, University of Melbourne, Australia; and a Fulbright Scholar
Award at York University, Toronto, Canada.

Shelley A. Blozis, PhD, is an associate professor of psychology at the University of

California at Davis. She is a quantitative psychologist, obtaining her doctorate from
the University of Minnesota, who has published articles on mixed models for
longitudinal data and methods for missing data. She has served on the editorial board
of Psychological Methods since 2003.

A. T. Panter, PhD, is the Bowman and Gordon Gray Distinguished Professor of

Psychology at the University of North Carolina at Chapel Hill. She develops
instruments, research designs, and data-analytic strategies for applied research
questions in higher education, personality, and health. She is an American
Psychological Association fellow, a member of the Graduate Record Examinations
Advisory Board, and a former member of a Social Security Administration advisory
panel to revise occupational information systems used for disability determination.
She and three other colleagues conducted the Educational Diversity Project to
examine links among race and other factors and educational diversity in law students
around the U.S., and she has coedited books on research design and quantitative
methods, including the APA Handbook of Research Methods in Psychology (2012), the
Handbook of Ethics in Quantitative Methodology (2011), and The Sage Handbook of
Methods in Social Psychology (2004).
 Quick Guide to Format

Headword [accuracyl n. |l7|lack of error or BIAS in a Sense number

measure; the more accurate the mea-
sure, the closer the measurement is to
Sense number the TRUE SCORE for an individual. [271 a
measure of performance on a task, usu-
ally defined as the proportion of correct
responses. Iaccurate] adj. Derived word

Akaike's i n f o r m a t i o n c r i t e r i o n
AIC) a SUMMARY STATISTIC used in com- Abbreviation
Cross-reference paring the relative IGOODNESS OF FIT| of
two or models for a given set of data,
while taking into account the number
of parameters in each model. The model
with the lowest AIC is considered the
best among all models specified. |[Hirot-
sugu Akaike (1927-2009), Japanese stat-
Etymology istician]
autocorrelation In] the situation in Part-of-speech label
which values of a variable measure over
time are correlated with other values of
the same series separated from them by a
specific interval. This often occurs with
economic or demographic data. Auto-
correlations are generally assumed to be
linear relationships and may be presented
Hidden entries graphically in an autocorrelogram| (or
I correlogram)|or formulaically in an auto-
correlation function (ACF). Also called] se-
Alternative name rial correlation.!See TIME-SERIES ANALY-
SIS.

Plural form axis n. (p/.|axes)| a fixed reference line in

a coordinate systemj See also ABSCISSA;
bRDINATE. Cross-references
 Aa
A-B-A-B-A design a type of SINGLE-CASE The A-B-A-C-A design helps to establish
DESIGN having five consecutive phases: the effects on the DEPENDENT VARIABLE
a baseline condition in which no treat- of alternating treatments by evaluating
ment is present (Phase A), a treatment the relationship between the introduc-
condition in which a manipulation is in- tion and removal of one treatment and a
troduced (Phase B), a return to the no- subsequent introduction and removal of
treatment condition (Phase A), a reintro- a second treatment. See also A-B-A-B-A
duction of the treatment (Phase B), and DESIGN.
a subsequent return to the no-treatment
condition (Phase A). It is an extension of A-B-A design a type of SINGLE-CASE DE-
the A-B-A-B DESIGN that provides stron- SIGN having three phases: a baseline con-
ger causal evidence of a relationship be- dition in which no treatment is present
tween the treatment and a consequen- (Phase A), a treatment condition in which
tial change in the DEPENDENT VARIABLE a manipulation is introduced (Phase B),
by removing the treatment a second time. and a return to the no-treatment condi-
See also A-B-A DESIGN; A-B-A-C-A DESIGN. tion (Phase A). The design allows for
evaluation of the introduction of the
A-B-A-B design a type of SINGLE-CASE treatment by comparing the DEPENDENT
DESIGN having four phases: a baseline VARIABLE between the first two phases
condition in which no treatment is pres- (A-B sequence) as well as evaluation of
ent (Phase A), a treatment condition in the removal of the treatment by compar-
which a manipulation is introduced ing the dependent variable between the
(Phase B), a return to the no-treatment last two phases (B-A sequence). This
condition (Phase A), and a subsequent greatly reduces the possibility of a coin-
reintroduction of the treatment (Phase cidental treatment effect, which may
B). In evaluating the treatment twice, occur in the simpler A-B DESIGN. See also
the A-B-A-B design helps to establish A-B-A-B-A DESIGN; A-B-A-B DESIGN.
causality by providing evidence of a re- A-B design the simplest SINGLE-CASE
peated relationship between the intro- DESIGN, comprising a pretreatment or
duction and removal of the treatment baseline phase (Phase A) followed by a
and a consequential change in the DE- treatment phase (Phase B). Although it
PENDENT VARIABLE. See also A-B-A-B-A allows for evaluation of the treatment's
DESIGN; A-B-A DESIGN. effect by comparing the DEPENDENT
A-B-A-C-A design a type of SINGLE- VARIABLE during the two phases, the de-
CASE DESIGN having five consecutive
sign does not provide evidence of causal-
phases: a baseline condition in which no ity since it does not establish a repeated
treatment is present (Phase A), a treat- relationship between the introduction
ment condition in which a manipula- and removal of the treatment and a con-
tion is introduced (Phase B), a return to sequential change in the dependent
the no-treatment condition (Phase A), a variable (compare A-B-A DESIGN). See
new treatment condition in which a dif- also PRETEST-POSTTEST DESIGN.
ferent manipulation is introduced a b d u c t i o n n. the process of studying an
(Phase C), and a subsequent return to event or phenomenon in order to gener-
the no-treatment condition (Phase A). ate possible explanatory hypotheses. The

I 1
ability level

object is to determine which hypotheses high or too low. Absolute error is

should be tested, rather than which should computed as the average ABSOLUTE DIF-
be adopted as correct. abduct vb. FERENCE between the intended or ex-
pected value and the actual value. See
a b i l i t y level an index of achievement
also CONSTANT ERROR; RANDOM ERROR.
or performance that reports the absolute
or relative ability of the participant in re- absolute frequency see FREQUENCY.
lation to the trait or characteristic being
absolute measurement a measure-
assessed.
ment made directly and independently
a b i l i t y parameter (symbol: 9) in ITEM of comparison with other measure-
RESPONSE THEORY, a theoretical value ments. An absolute measurement of an
that represents a person's capability or individual's height, for instance, would
probable performance on a task. It is as- yield a single definitive value that need
sumed that test takers possess some de- not be assessed relative to another per-
gree of underlying ability, and that for son to be meaningful. Compare RELA-
each individual at each ABILITY LEVEL TIVE MEASUREMENT.
there is a specific ITEM CHARACTERISTIC
CURVE defining the probability of re- absolute r a t i n g scale a type of assess-
sponding correctly. ment instrument in which the targets
(e.g., people, objects) are not compared
a b i l i t y scale a set of values that repre- with other targets or a standard stimulus
sent increasingly higher levels of perfor- but are judged in absolute terms. For in-
mance on a test or set of tasks. Ability stance, a respondent may be presented
scales are ORDINAL measures, as the dis- with the options excellent, good, fair, and
tances between any two adjacent scale poor and asked to choose the one that
values are not equal and the value of zero best describes his or her current state of
is arbitrary. See also DIFFICULTY SCALE. health. Compare COMPARATIVE RATING
SCALE.
abscissa n. the horizontal coordinate in
a graph or data plot; that is, the x-ax\s. absolute true score 1. a value for a
See also ORDINATE. random variable that exists independ-
ently of any test used to measure the
absolute deviation the distance be-
variable. For example, an individual has
tween a data point and a measure of
an absolute true score for blood pressure
CENTRAL TENDENCY, such as the mean.
or cholesterol level regardless of the ac-
The absolute deviation does not indicate
curacy of any tests used to measure these
the direction of difference. For example,
scores. 2. an average score that is derived
if the mean is 10 a value of 18 and one of
from testing every member of a popula-
2 both show an absolute deviation from
tion (e.g., every child of a certain age in
the mean of 8; the signs associated with
the state), rather than one derived from
-1-8 and -8 are ignored.
a sample of that population.
absolute difference the distance be-
absolute value a number considered
tween two numeric values disregarding without regard to its algebraic sign (i.e.,
whether this is positive or negative. The whether it is positive or negative). For
absolute difference thus provides no in- example, assume that for each person in
formation about relative magnitude. For a weight management program there
example, the absolute difference be- was a number indicating the difference
tween 11 and 20 is 9, as is the absolute between the current week's weight and
difference between 13 and 4. the weight on the previous week. This
absolute error the degree to which an could reveal a negative number (e.g., -1)
observation is inaccurate without speci- if the person lost one pound; conversely
fication of whether it errs by being too it could be a positive value (e.g., -i-l) if
 action research

the person weighed one pound more which a random sample is chosen from a
than last week. If the absolute value of larger group of items and used to make a
the weight difference was taken, it would decision about the quality of the items
simply reveal a difference of 1 pound, in that group. Acceptance sampling is
without any indication of whether it was often used in commerce to test the qual-
plus or minus. Also called modulus. ity of merchandise in a batch: A random
absolute zero a value on a measure- sample from the batch is inspected and
ment scale that denotes the complete the results used to determine whether
absence of the measured characteristic. the batch as a whole meets desired stan-
A RATIO SCALE has an absolute zero, dards or whether it fails to meet stan-
whereas an INTERVAL SCALE does not. dards and should be rejected as
Also called true zero. ; defective. Also called lot acceptance
sampling.
absorbing state a condition from
which there is no possibility of transi- accessible adj. in a MARKOV CHAIN, de-
tioning to another condition. In a se- scribing a state / that there is a possibility
quence of events such as a MARKOV of reaching from another state i in some
CHAIN, for example, a state is absorbing number of steps. accessibility n.
if there is zero probability of leaving that accidental s a m p l i n g see CONVE-
state once it has been obtained. Com- NIENCE SAMPLING.
pare TRANSIENT STATE.
accelerated f a i l u r e time model in accrual rate the rate at which some-
SURVIVAL ANALYSIS, a model in which thing accumulates. For example, the ac-
the response variable is a known trans- crual rate of volunteers to participate in
formation of the time until the event of a research study may be documented
interest occurs. When this transforma- and used to help determine how best to
tion is the LOGARITHM, the model is re- recruit individuals for similar studies in
ferred to more specifically as a censored the future.
linear regression model. In aging re- accuracy n. 1. lack of error or BIAS in a
search, for example, an investigator measure; the more accurate the measure,
might use an accelerated failure time the closer the measurement is to the
model to evaluate whether a specific in- TRUE SCORE for an individual. 2. a mea-
tervention increases lifespan. sure of performance on a task, usually
accelerated longitudinal design defined as the proportion of correct re-
see COHORT-SEQUENTIAL DESIGN. sponses. accurate adj.
acceleration n. in mathematics and sta- ACE abbreviation for ALTERNATING CON-
tistics, the rate of change in the SLOPE,of DITIONAL EXPECTATION.
a function or the rate of change in one
ACES abbreviation for ACTIVE CONTROL
variable as a function of an increase in a
EQUIVALENCE STUDY.
second variable. Compare DECELERATION.
acceptance region the range of values ACF abbreviation for AUTOCORRELATION
for a test statistic that leads to accep- function.
tance of the NULL HYPOTHESIS, such that acquiescent response set see YEA-
the ALTERNATIVE HYPOTHESIS Is rejected SAYING.
as a valid explanation for observed data.
Compare CRITICAL REGION. a c t i o n research socially useful and
theoretically meaningful research devel-
acceptance-rejection method see
oped and carried out in response to a so-
REJECTION METHOD. ;
cial issue or problem, results of which
acceptance s a m p l i n g a process in are applied to improve the situation
active control equivalence study

(e.g., by changing existing or developing a c t u a r i a l adj. in medicine, describing

new public policies). an approach to diagnosis and treatment
that uses data about previous events to
active control equivalence study
estimate the likelihood of a particular
(ACES) a three-group experimental de-
outcome. An actuarial approach will rely
sign in which one group receives the
on statistically established relationships
treatment of interest, a second group re-
to reach conclusions or make decisions
ceives a comparable standard treatment,
about, say, which approach to use in
and a third CONTROL GROUP receives a
treating a particular disorder. Compare
PLACEBO. The two treatment groups are
CLINICAL.
compared to each other to assess their
equivalence and also are compared to acyclic d i g r a p h see DIRECTED ACYCLIC
the placebo group to evaluate the effi- GRAPH.
cacy of the treatments.
A D abbreviation for average deviation.
active c o n t r o l t r i a l a two-group ex- See MEAN ABSOLUTE DEVIATION.
perimental design in which one group
adaptation period a period of time
receives the treatment under study and
during which a research participant be-
the second group receives a comparable
comes accustomed to the materials, in-
standard treatment. Although efficacy
struments, or equipment to be used in a
generally is best evaluated by comparing
study or reaches a certain performance
a treatment to a PLACEBO, active control
level. Adaptation periods help reduce
trials are used in situations when with-
the influence of situational novelty on a
holding treatment from individuals by
participant's behavior.
assigning them to a placebo group is not
ethical. Compare PLACEBO CONTROLLED adaptive cluster s a m p l i n g an ADAP-
TRIAL. TIVE SAMPLING scheme that also involves
an element of CLUSTER SAMPLING. The
active deception intentionally mis- study population is divided into a num-
leading research participants by giving ber of equally sized clusters, one of which
them false information in order to get is randomly drawn and examined. If the
more valid results. For example, a re- units in the set meet the predetermined
searcher might present participants with criteria, then adjacent sets are chosen
incorrect correlations between gender and evaluated. If the adjacent sets also
and performance in different academic meet the criteria, they too are added to
subject areas before testing whether the sample and additional sets are se-
such hypothetical statistics affect subse- lected. The process continues until cases
quent performance on different achieve- are found that do not meet the criteria.
ment tests. Also called deception by
commission. Compare PASSIVE DECEP- adaptive method any procedure tai-
TION. See DECEPTION RESEARCH. lored to a given problem or situation.
Examples of adaptive methods include
activity l o g a diary kept by a researcher ADAPTIVE TESTING, in which items
or research participant of activities in change in response to an examinee's
various settings. An activity log may in- performance, and ALGORITHMS that ad-
clude information about the location of just to a given data problem to optimize
the participant by time period (e.g., at performance of the algorithm in that
home, at work, traveling) and whether particular instance.
the time is spent alone or with family,
friends, or work associates. An activity adaptive s a m p l i n g a method of sam-
log as a method of obtaining a record of pling data in which information from
events generally is superior to interviews past outcomes is used to reduce the
based on a participant's memory. chances of collecting future data that
 adi

correspond to poor outcomes. A group is also known as a LINEAR FUNCTION.

of individuals randomly chosen from Compare NONADDITIVE.
the population of interest is evaluated additive effect the constant effect of
and the information gathered is used to one PREDICTOR VARIABLE on a DEPEND-
concentrate additional selection efforts ENT VARIABLE across all levels of other
where they are most likely to be success- related predictors. Thus, the total effect
ful. For example, if a researcher is inter- of all predictors is equal to the sum of
ested in studying a certain animal of a their individual effects, with no INTER-
particular age, he or she couldfirstassess ACTION EFFECT.
a subset of animals from a given geo-
graphical area and then target further additive model a description of the re-
sampling efforts in the specific portions lationship between a response variable
of that area shown to possess the most and a set of predictor variables in which
animals meeting the criteria. the effect of each predictor is assumed to
be the same across all levels of the other
adaptive testing a testing technique predictors in the model. Thus, the com-
designed to adjust to the response char- bined effect of all predictors is determined
acteristics of individual examinees by by summing their individual effects.
presenting items of varying difficulty
based on the examinee's responses to additive scale a scale with all points
previous items. The process continues distributed equally so that a meaningful
until a stable estimate of the ABILITY result can be obtained by addition (e.g.,
LEVEL of the examinee can be deter- a metric ruler).
mined. See also ITEM RESPONSE THEORY. a d d i t i v i t y test a method used to evalu-
ate whether the effects of each of a set of
added-variable plot a method of dis- predictor or INDEPENDENT VARIABLES on
playing data in which the variance re- a DEPENDENT VARIABLE are constant
maining in a response or DEPENDENT across all levels of the other predictors
VARIABLE after it has been predicted by included in the model. A significant re-
one or more INDEPENDENT VARIABLES is sult from this test indicates that the rela-
plotted against the variance remaining tionship of interest is not an ADDITIVE
after a new independent variable is EFFECT but is characterized by one or
added. In other words, added-variable more INTERACTION EFFECTS between the
plots show the result of adding another independent variables.
variable to the model while taking into
account the effects of the other inde- adequate sample a SAMPLE that ade-
pendent variables already in the model. quately represents the larger population
Also called partial regression plot. from which it was drawn in terms of size,
being large enough to provide satisfac-
a d d i t i o n rule a maxim of probability tory PRECISION by minimizing the pos-
theory stating that the likelihood of ob- sibility of chance affecting the data
serving a set of distinct events is equal to obtained.
the sum of the probabilities of observing
ad hoc for a particular purpose or in
the individual events. Also called ad-
response to some particular event or oc-
dition law; or rule. Compare MULTI-
currence. For example, an ad hoc com-
PLICATION RULE.
mittee is convened on a short-term basis
to address a single problem, and an ad
additive adj. characterized or produced
by addition. For example, a FUNCTION is hoc hypothesis is an explanation of a
said to be additive when the quantities particular phenomenon, rather than a
defining it may be summed to obtain a general theory. [Latin, literally: "to this"]
meaningful result. An additive function adj symbol for ADJUSTED R^.
adjusted effect

adjusted effect the effect of a predictor STANDARD DEVIATION generally except

or INDEPENDENT VARIABLE On a response that /V- 1 is used as the denominator in
or DEPENDENT VARIABLE after the influ- the formula instead of N (where N = sam-
ence of one or more other predictors has ple size), thus providing an unbiased es-
been removed. For example, a researcher timate of the POPULATION STANDARD
might find that education level predicts DEVIATION.
income via the MEDIATOR of residential adjusted variance the square of the
area; an adjusted effect would be one ADJUSTED STANDARD DEVIATION. The
without the influence of the mediator adjusted variance is an unbiased esti-
included. mate of the POPULATION VARIANCE.
adjusted mean 1. in ANALYSIS OF VARI- a d j u s t i n g f o r baseline a method of
ANCE, the average score value (MEAN) accounting for preexisting differences
obtained after removing all differences between individuals or groups prior to
that can be accounted for by COVARI- examining the effect of a treatment or
ATES. 2. a mean value obtained after re- experimental condition. For example, a
moving any OUTLIERS. researcher may wish to determine
whether a certain training program im-
adjusted R the correlation between
proves a person's performance on a task.
scores on a response or DEPENDENT
He or she might adjust for baseline by
VARIABLE and the values predicted by a
administering the individual a practice
set of INDEPENDENT VARIABLES, after ac-
test to determine the initial performance
counting for the number of predictors
level before holding a series of training
and the number of observations in-
sessions and then administering a
volved in the calculation. See also MUL-
posttest to determine the new perfor-
TIPLE CORRELATION COEFFICIENT.
mance level. If the DIFFERENCE SCORE
adjusted R^ (symbol: adj R^; R\d|) the between the two tests is significantly dis-
square of the correlation between scores similar there is some evidence that per-
on a response or DEPENDENT VARIABLE formance is improved after the training
and the values predicted by a set of INDE- session.
PENDENT VARIABLES after accounting for admissible hypothesis a HYPOTHESIS
the number of predictors and the num- that is logically possible and open to em-
ber of observations involved in the cal- pirical testing given a specific set of con-
culation. It gives the proportion of the ditions.
variance in a response that is accounted
for by its relationship with the predic- a d o p t i o n study a research design that
tors and yields a better estimate of the investigates the relationships among ge-
population variance than the ADJUSTED netic and environmental factors in the
R upon which it is based. In an ANALYSIS development of personality, behavior,
OF VARIANCE or MULTIPLE REGRESSION or disorder by comparing the similarities
the adjusted R^ often is known as EPSI- of biological parent-child pairs with
LON SQUARED. Also called shrunken those of adoptive parent-child pairs.
R^. See also COEFFICIENT OF MULTIPLE A E q abbreviation for AGE EQUIVALENT.
DETERMINATION.
age c a l i b r a t i o n a process in which a
adjusted standard deviation for a test score is assigned a value so that the
set of scores, a measure of the DISPER- score may be interpreted relative to
SION of these scores about the mean other test takers in the same age range.
after taking into account the fact that
age cohort see COHORT.
this mean is an estimate of the true value
in a larger populafion of interest. It is age effect in research, any outcome as-
calculated in the same manner as the sociated with being a certain age. Such
 all-possible-subsets multiple correlation

effects may be difficult to separate from aggregation n. in statistics, a process of

COHORT EFFECTS and PERIOD EFFECTS. combining and summarizing a set of
scores into a smaller set of scores that
age equivalent (AEq) a measure of de- capture an aspect of the original set. See
velopment or performance expressed in AGGREGATE DATA. Compare DISAGGRE-
terms of the average chronological age at GATION. aggregate vb. aggrega-
which the observed score is obtained. tive adj.
For example, assume a student obtains a
score of 95 on a particular test, a value aggregation p r o b l e m the difficulty
typical of the average performance of of drawing conclusions or making infer-
students in the eighth grade. Thus, the ences about an individual on the basis of
age equivalent of 95 is 13, the age of data for a group or population. See ECO-
most eighth graders. Also called age- LOGICAL FALLACY.
equivalent score; test age. agreement coefficient see COEFFI-
CIENT OF AGREEMENT.
age-equivalent scale a system for ex-
pressing test scores in terms of the chro- A H abbreviation for ALTERNATIVE HY-
nological ages at which the scores are POTHESIS.
typically obtained.
AID abbreviation for AUTOMATIC INTER-
age-grade scaling a method of stan- ACTION DETECTOR.
dardizing a test by establishing norms Akaike's information criterion
based on a sample of children who are of (AIC) a SUMMARY STATISTIC used in
the t)'pical chronological age for their comparing the relative GOODNESS OF FIT
grade in school. of two or more models for a given set of
data, while taking into account the num-
agglomerative clustering a proce-
ber of parameters in each model. The
dure used to combine separate entities
model with the lowest AIC is considered
into homogeneous sets (clusters) by
the best among all models specified. [Hi-
forming pairs of similar entities and then
pairing these pairs unfil all are merged rotsugu Akaike (1927-2009), Japanese
into one large group. That is, in agglom- statistician]
erative clustering one focuses initially algorithm n. a well-defined procedure
upon each individual or object and or set of rules that is used to solve a par-
moves progressively "upward" to corn- ticular problem or conduct a series of
bine them into a single, agglomerative computations, usually in a limited num-
group or cluster. It is one of two types of ber of steps. algorithmic adj.
HIERARCHICAL CLUSTERING, the Other alienation coefficient see COEFFI-
being DIVISIVE CLUSTERING. Also called CIENT OF ALIENATION.
agglomerative hierarchical clus-
tering; bottom-up (hierarchical) allocation ratio a ratio of the number
clustering; hierarchical agglom- of individuals assigned to one treatment
erative clustering. condition to the number assigned to a
different condition that serves to maxi-
aggregate data scores or observations mize the ability of a statistical test to de-
that have been re-expressed by a SUM- tect a difference in a measured outcome
MARY STATISTIC Calculating the arith- between the conditions.
metic average of a set of test scores
obtained over time for each individual allocation rule see CLASSIFICATION RULE.
in a group and then using each person's all-possible-subsets m u l t i p l e cor-
single average score as representative of relation a measure of the degree of as-
their test performance would be an ex- sociation between an outcome variable
ample of aggregating data. and its corresponding predicted value
all-possible-subsets regression

based on a given set of predictors from SUREMENT ERRORS are approximately

an analysis in which all possible subsets the same across the two versions. Also,
from a finite pool of predictors have each version should yield similar score
been evaluated. See ALL-POSSIBLE-SUBSETS distributions (i.e., similar MEANS and
REGRESSION. STANDARD DEVIATIONS). Alternate forms
of a test can be used to measure its reliabil-
all-possible-subsets regression a me-
ity (see ALTERNATE-FORMS RELIABILITY).
thod for predicting an outcome variable
Also called altemative test form;
based on a series of equations formed by
comparable form; equivalent
all possible subsets of predictors from a
form; parallel form.
finite pool of predictors. The "best" sub-
set is identified using criteria established alternate-forms reliability a mea-
by the researcher, such as the value of sure of the consistency and freedom from
AKAIKE'S INFORMATION CRITERION or of error of a test, as indicated by a CORRE-
the COEFFICIENT OF MULTIPLE DETERMIN- LATION COEFFICIENT obtained from re-
ATION. Also called all-possible-subsets sponses to two or more ALTERNATE
multiple regression; setwise re- FORMS of the test. Also called compara-
gression. ble-forms reliability; equivalent-
forms reliability; parallel-forms
a l p h a (symbol: a) n. 1. the likelihood of reliability.
incorrectly rejecting a statement or hy-
pothesis concerning a characteristic of a a l t e r n a t i n g c o n d i t i o n a l expecta-
population. More specifically, it is the t i o n (ACE) an algorithm designed to
probability of incorrectly rejecting a true obtain optimal TRANSFORMATIONS in
NULL HYPOTHESIS (i.e., committing a TYPE data analysis. In MULTIPLE REGRESSION,
I ERROR) in research. Although the value for example, where a response or DE-
of alpha is chosen by the investigator PENDENT VARIABLE is predicted by sev-
based on what is deemed acceptable for eral INDEPENDENT VARIABLES and where
a particular study, alpha values of .05 it is assumed that the effects of each of
and .01 are commonly used. 2. a mea- the predictors are constant, ACE may be
sure of RELIABILITY for a set of responses used to find a transformation that maxi-
to a test or measure. See CRONBACH'S mizes the proportion of variance in one
ALPHA. variable that is explained by the others.
a l p h a coefficient see CRONBACH'S a l t e r n a t i n g treatments design a
ALPHA. type of study in which the experimental
condition or treatment assigned to the
a l p h a error see TYPE I ERROR.
participant changes from session to ses-
a l p h a level see SIGNIFICANCE LEVEL. sion or within sessions. For example, a
researcher comparing two methods for
alternate f o r m a set of test items that eliminating the disruptive classroom be-
are developed to be similar to another havior of a student might have the
set of test items, so that the two sets rep- teacher use one method throughout the
resent different versions of the same test. morning and the other method through-
Each item set is intended for the same out the afternoon and then evaluate the
purpose (i.e., measures the same con- student's behavior with each technique.
cept) and is administered in the same
manner. In order to demonstrate that alternative hypothesis (AH; symbol:
one test is an alternate form of the other, / f j . Ha) a statement that is contrasted
a researcher usually must show that with or contradicts the NULL HYPOTHE-
there is matching content (each test has SIS as an explanation for observed data.
the same number of each kind of item) Generally, it is a scientific prediction of
and that FACTOR LOADINGS and MEA- significant results in HYPOTHESIS TEST-
 analysis of unweighted means

ING; that is, an alternative hypothesis two or more groups on one or more IN-
posits meaningful differences or relation- DEPENDENT VARIABLES. In other words,
ships between the variables under inves- it is a statistical method of studying the
tigation; responses of different groups to a DE-
alternative hypothesis distribu- PENDENT VARIABLE that adjusts for the
t i o n a theoretical set of plausible values influence of a variable that is not being
of a characteristic under certain assurnp- investigated but nonetheless is related to
tions that is compared to its correspond- the dependent variable and thus may in-
ing NULL DISTRIBUTION in the process of fluence the study results. An analysis of
conducting a POWER ANALYSIS. covariance is appropriate in two types of
cases: (a) when experimental groups are
alternative test f o r m see ALTERNATE suspected to differ on a background-
FORM. correlated variable in addition to the dif-
A M L abbreviation for ASYMMETRIC MAX- ferences attributed to the experimental
IMUM LIKELIHOOD.
treatment (i.e., the analysis corrects for
chance differences between groups that
analogue observation a response re- arise when participants are assigned ran-
corded from a participant in an ANA- domly to the treatment groups) and (b)
LOGUE STUDY designed to induce a where adjustment on a covariate can in-
particular behavior in a controlled envi- crease the precision of the experiment
ronment, such as a laboratory or clinic. (i.e., reduce the ERROR TERM). For exam-
Compare NATURALISTIC OBSERVATION; ple, suppose a researcher analyzes
SELF-MONITORING OBSERVATION. whether there is a difference in learning
among three types of instructionin-
analogue sample a group of individu-
class lecture, online lecture, and text-
als selected for inclusion in a study who
book only. He or she divides a random
possess the symptoms of a particular dis-
selection of adult students into three
order but who have not sought treat-
groups, implements the different instmc-
ment for the disorder. For example, a
tion types, and administers the same test
researcher may use college students with
to all parficipants to determine how much
mild problems resembling those seen in
they learned. If the researcher knows
outpatient clinical settings and expose
each participant's educational back-
them to specific therapy techniques to
ground, he or she could use an analysis
study their responses.
of covariance to adjust the treatment ef-
analogue study a study intended to in- fect (test score) according to educational
duce a particular behavior under con- level, which would reduce the observed
trolled environmental conditions that variation between the three groups
resemble or approximate the real-worM caused by variation in education levels
situation of interest. Examples include rather than by the instruction itself.
the use of hypnosis, drugs, and sensory
deprivation in a laboratory setting to in- analysis of covariance structures a
duce brief periods of abnormal behavior method of examining the relationships
that simulate those of psychopathologi- among a set of variables with regard to
cal conditions. Also called analogue how each one varies according to varia-
design; analogue experiment; ana- tion in the others. It is similar to STRUC-
logue research. TURAL EQUATION MODELING but disting-
uished by its emphasis on simultaneous
analysis of covariance (ANCOVA) an variation (i.e., covariance). Also called
extension of the ANALYSIS OF VARIANCE covariance structure analysis.
that adjusts for the influence of a CO-
VARIATE in testing whether there is a sig- analysis o f unweighted means see
nificant difference between means of UNWEIGHTED MEANS ANALYSIS.
analysis of variance

analysis o f variance (ANOVA) a sta- anchor test a set of test items used as a
tistical method of studying the variation reference point in comparing ALTERNATE
in responses of two or more groups on a FORMS of a test. One alternate form is ad-
DEPENDENT VARIABLE. A N O V A S test for ministered to one group of participants,
significant differences among the mean another is administered to a different
response values of the groups and can be group, and the items comprising the an-
used to isolate both the joint INTERAC- chor test are administered to both groups.
TION EFFECTS and the separate MAIN EF- Scores on each alternate form are then
FECTS of INDEPENDENT VARIABLES. compared with scores on the anchor test.
analysis o f weighted means a A N C O V A acronym for ANALYSIS OF CO-
method of comparing samples of un- VARIANCE.
equal sizes in which the data for each A n d e r s e n - G i l l model an extension of
group are averaged in a manner that the cox REGRESSION ANALYSIS model
takes into account the differing number that is used in SURVIVAL ANALYSIS for data
of observations contributing to the cal- that show LEFT CENSORING, time-varying
culation. See WEIGHTED AVERAGE. COVARIATES, recurrent events, and dis-
analysis u n i t see UNIT OF ANALYSIS. continuous intervals of risk. [Per Kragh
Andersen (1952- ), Danish statistician;
a n a l y t i c approach any method based Richard D. Gill (1951- ), Brifish-born
on breaking down a complex process Dutch mathematician]
into its parts so as to better understand
A n d e r s o n - D a r l i n g test see DARLING
the whole. For example, a researcher
TEST. [Theodore W. Anderson (1918- ),
studying cognition in children might
U.S. mathematician; Donald A. Darling
identify such component skills as prob-
(1915- ), U.S. mathematician]
lem solving, reasoning, imagination,
and memory; examine each in turn; and Andrews plot a method for displaying
then detemine how they interact. Com- multidimensional data using only two
pare S Y N T H E T I C A P P R O A C H . dimensions in which one curve is gener-
ated for each row of data, as in the hypo-
a n a l y t i c i n d u c t i o n a QUALITATIVE RE- thetical illustration overleaL Here f(t) is
SEARCH strategy for developing and test- a function derived for an observation
ing a theory in which the researcher (e.g., occupation type) that is based on
tentatively defines a phenomenon, cre- multiple variables (e.g., education re-
ates a hypothesis to explain it, and ex- quired, starring salary, promotion oppor-
amines a single specific occurrence of tuniries). The different values of t indi-
the phenomenon in order to confirm or cate the similarity of the three occupa-
refute the hypothesis. If the hypothesis tions shown. [David F. Andrews, Cana-
is confirmed, additional cases are exam- dian statistician]
ined until a sufficient degree of certainty
about the correctness of the hypothesis a n d r u l e see MULTIPLICATION RULE.
is obtained and the study may be con- anecdotal method an investigational
cluded. If the hypothesis is not con- technique in which informal verbal re-
firmed, the phenomenon is redefined or ports of incidents casually observed are
the hypothesis revised so as to accom- accepted as useful information. The an-
modate the findings. ecdotal method is scientifically inade-
quate but can offer clues as to areas of
anchor n. a number or descriptive qual- investigation that warrant more system-
ity used as a point of reference for mak- atic, controlled research.
ing a subjective judgment. For instance,
a study participant may be asked to rate angular t r a n s f o r m a t i o n see ARC
his or her health on a scale that uses an SINE TRANSFORMATION.
upper anchor labeled perfect health. a n i m a l rights the belief that animals

10
 Anscombe residual

n n \ \ r
0.2 0.4 0.6 0.8 1.0

Andrews plot

should be treated with respect and be senting the most important DESCRIP-
free from exploitation and abuse by hu- TIVE STATISTICS obtained from an ANAL-
mans. Animal research is monitored in YSIS OF VARIANCE (ANOVA). It typically
many universities and other organiza- provides the DEGREES OF FREEDOM, the
fions by an animal care committee, whose SUM OF SQUARES, the MEAN SQUARE, the
purpose is to ensure the humane care F RATIO, and the SIGNIFICANCE LEVEL as-
and use of animals by assuring compli- sociated with each effect studied, as in
ance with federal and state regulations the example below.
and by supporting veterinary oversight
and continuing education and training. Ansari-Bradley test a NONPARA-
METRIC procedure for determining the
a n o n y m i t y n. a principle of research equivalence of two samples without re-
ethics stating that the identity of a study quiring that the variables of interest
participant should remain unknown. have a NORMAL DISTRIBUTION. Rather, it
Relatedly, CONFIDENTIALITY applies to assumes the middle values of the distri-
situations in which a participant's iden- butions are equal and evaluates whether
tity is known but should not be dis- the DISPERSIONS are equal, whether the
closed by the researcher. scores within each sample are independ-
A N O V A acronym for ANALYSIS OF VARI- ent, and whether the shapes of the dis-
ANCE.
tributions are identical.

ANOVA summary table a table pre- Anscombe residual a particular type

Source df SS MS F P
Lecture topic (L) 2 1,200 600 21.18 < .001
Presentation method (?) 1 2,200 2,200 77.66 <.001
L X P interaction 2 700 350 12.35 <.001
Error 60 1,700 28.33

ANOVA summary table

antecedent variable

of error or unexplained variance in a re- a data point may fall. For example,
sponse variable that does not require the in displaying ages from 1 to 100, a re-
values of the response variable to follow searcher may use intervals such as 1-10,
a NORMAL DISTRIBUTION. It is used with 11-20, and so forth. The values deter-
GENERALIZED LINEAR MODELS. [Francis J. mining the intervals (e.g., 1 and 10) are
Anscombe (1918-2001), British statisti- the apparent limits.
cian]
a p p l i e d research studies conducted to
antecedent variable any variable that solve real-world problems, as opposed to
precedes a response variable. For ex- studies that are carried out to develop a
ample, in REGRESSION ANALYSIS it is the theory or to extend basic knowledge. Ex-
INDEPENDENT VARIABLE, in a MEDI- amples include ACTION RESEARCH and
ATIONAL PROCESS it is the intervening EVALUATION RESEARCH. Compare BASIC
variable, and in PATH ANALYSIS it is the RESEARCH.
variable beginning a chain of causal links.
Compare CONSEQUENT VARIABLE. applied science research conducted to
serve a practical human purpose rather
antedependence n. a relationship than with the purpose of extending
among multiple values of a response knowledge for its own sake. Compare
variable observed over time such that BASIC SCIENCE.
each value depends upon one or more
previous values. See AUTOCORRELATION. a p p l i e d statistics the use of statistical
methods and procedures to understand
A - o p t i m a l design a research design in data in psychology, sociology, econom-
which the set of INDEPENDENT VARI- ics, and other disciplines. Compare THE-
ABLES causing the least average variance ORETICAL STATISTICS.
in a DEPENDENT VARIABLE is chosen
from a larger pool of possible variables apprehensive-subject role behavior
for predicting values of that variable. See displayed by a research participant who
also D-OPTIMAL DESIGN; E-GPTIMAL DE- dislikes being evaluated or is otherwise
SIGN.
anxious about the study but nonetheless
tries to do well and convey a positive
a posteriori denoting conclusions de- impression. Compare FAITHFUL-SUBJECT
rived from observations or other mani- ROLE; GOOD-SUBJECT ROLE; NEGATIVIS-
fest occurrences: reasoning from facts. TIC-SUBJECT ROLE.
Compare A PRIORI. [Latin, "from the lat-
ter"] approximate bootstrap c o n f i -
dence method a procedure in which
a posteriori comparison see POST many subsets of an observed sample
HOC COMPARISON. data set are used to produce a range of
apparatus n. any instrument or equip- plausible values for a population charac-
ment used in an experiment or other re- teristic, such as the MEAN or VARIANCE.
search. It is less computationally involved than
similar BOOTSTRAPPING methods.
apparent error rate the inaccuracy
that can be observed when applying a approximation n. the process of ob-
REGRESSION EQUATION hypothesizing taining a value that is at least close to the
associations among a set of variables to a desired or actual value. For example, one
set of actual data points. An estimate of might round a measurement to the near-
how well the model fits the data, the ap- est decimal place for ease of subsequent
parent error rate often underestimates calculations. The degree of inaccuracy
the true error rate. inherent to this process is known as ap-
proximation error.
apparent l i m i t the lower or upper
bound of a CLASS INTERVAL within which a p r i o r i denoting conclusions derived
 arithmetic mean

from premises or principles: deducing journals, historical documents, and

from prior assumptions. Compare A POS- other existing records or data available
TERIORI. [Latin, "prior to"] in storage in scientific research. Archival
a p r i o r i comparison any examina- research allows for unobtrusive observa-
tion in which two or more quantities are tion of human activity in natural set-
compared in accordance with plans es- tings and permits the study of
tablished prior to conducting the re- phenomena that otherwise cannot eas-
search study. For example, even before ily be investigated. A persistent draw-
data are collected, a researcher might hy- back, however, is that causal inferences
pothesize that two groups given per- are always more tentative than those
sonal instruction would show better provided by laboratory experiments.
mean performance on a task compared Also called archival method.
to those who receive only written in- arc sine transformation a means of
struction. Thus, he or she could decide changing proportional data to approxi-
in advance to compare the combined mate a NORMAL DISTRIBUTION: Percent-
personal instruction groups to the written ages that denote counts or frequencies
instmcfion group. The researcher might are converted into a new set of scores
also decide ahead of time to compare whose distribution assumes a beU-
one personal instruction group with the shaped curve. The process minimizes
other as an additional a priori compari- variances across the different groups
son. Also called a priori contrast; being compared and allows for the ap-
planned comparison (or contrast). plication of certain analytic techniques
Compare POST HOC COMPARISON. requiring that normality and other AS-
SUMPTIONS be met. Also called angular
aptitude-treatment interaction
transformation; arcsine transfor-
(ATI) see TRAIT-TREATMENT INTERACTION.
mation; inverse sine transforma-
a r b i t r a r y constant in an equation, an tion.
undetermined quantity whose value is
unchanged across units so as to ensure area s a m p l i n g a method of selecting
the solution conforms to certain require- individuals for research in which spe-
ments. cific neighborhoods, streets, homes, or
other geographic areas are designated in
a r b i t r a r y o r i g i n a value chosen as a advance as the source of participants.
reference point for other values. In some
calculations, it is convenient to express a area under the curve (AUC) in a
set of scores as DEVIATIONS about an ar- graphical display of a DISTRIBUTION, the
bitrarily chosen point of origin (such as region between the plotted function and
the MEAN of a distribution) to ease later the horizontal A:-axis. It is used in such
computational burdens. calculations as determining the proba-
bility of the occurrence of specific values
a r b i t r a r y weight a WEIGHT used to as- of a RANDOM VARIABLE.
sign different levels of importance to dif-
ferent entities that is not specific to a argument n. a sequence of propositions
procedure or data set. that provides logical reasons for accept-
ing a conclusion as valid or true. A single
archival data information about past one of these statements is referred to as a
events, behaviors, and other phenom- premise.
ena that are stored in a relatively perma-
nent form, particularly as used in A R I M A model acronym for AUTORE-
ARCHIVAL RESEARCH. Also called archi- GRESSIVE INTEGRATED MOVING-AVERAGE
val records. MODEL.
a r c h i v a l research the use of books, a r i t h m e t i c mean see MEAN.

13
arithmetic progression

arithmetic progression a sequence term is often preferred over SAMPLING

of numbers in which the difference be- BIAS in clinical contexts.
tween consecutive numbers is a fixed Aspin-Welch-Satterthwaite test
value. For example, the numbers 1, 5, 9,
(AWS test) a statistical method used to
13, and 17 form an arithmefic progres-
evaluate the NULL HYPOTHESIS regarding
sion. Also called arithmetic sequence;
the differences between two sample
arithmetic series.
means when the population variances
A R M A model abbreviation for AUTO- are not equal. It is a proposed solution to
REGRESSIVE MOVING-AVERAGE MODEL. the BEHRENS-FISHER PROBLEM a n d thus
may not be appropriate when data are
A R model abbreviation for AUTORE-
not normal. Also called Aspin-Welch
GRESSIVE MODEL.
test; Welch test. [Alice A. Aspin; B. L.
array n. any ordered arrangement of Welch, British statistician; F. E. Satter-
data, particularly a two-dimensional thwaite]
grouping of data into rows and columns. assessment n. in research, a systematic
The following listing of students' scores process of obtaining information from
on a test is an example of a simple array: parficipants and using it to make infer-
Student A 55 ences or judgments about them.
Student B 76 assessment-classification model a
Student C 81 system in which a profile of scores is
Srijdent D 82 used to evaluate and group individuals,
Student E 89 as opposed to using a score average or
Student F 90 some other summary measure. For ex-
Student G 90 ample, members of a human resources
Student H 90 department at an organization might
Student i 94 flrst determine the characteristics of job
Student J 98 candidates and analyze the require-
The concept may be extended to more ments of different types of jobs at the
than two dimensions. company, then classify or match specific
people with specific job types.
a r t i f a c t n. an experimental finding that
is not a reflection of the true state of the assessment instrument any test, in-
phenomenon of interest but rather is the terview, questionnaire, or other tool for
consequence of aflaweddesign or ana- the evaluation of ability, achievement,
lytic error. For example, characteristics interests, personality, psychopathology,
of the researcher (e.g., expectations, per- or the like.
sonality) or the participant (e.g., aware- assessment research the study of an
ness of the researcher's intent, concern organizational unit in order to under-
over being evaluated) are common stand its functioning and make recom-
sources of arfifacts. See also CONFOUND; mendations for improvements or
DEMAND CHARACTERISTICS. changes. See also EVALUATION RESEARCH.
ascertainment bias error in selecting association n. the degree of statistical
individuals or units for a sample, such dependence or relationship between two
that those units selected are not repre- or more phenomena. See also CORRELA-
sentative of the relevant population. For TION; STRENGTH OF ASSOCIATION. as-
example, a medical researcher who stud- sociative adj. associational adj.
ies a sample of patients that omits cer-
tain types of people who have the association analysis see CORRELA-
disorder of interest is likely to obtain re- TION ANALYSIS.
sults having an ascertainment bias. The association model a method used to
 asymptotic method

study the strength of relationship be- midpoint. 2. an infrequent synonym of

tween a predictor variable and a re- DIRECTIONAL TEST.
sponse variable where values for the
asymmetric m a x i m u m likelihood
latter take the form of ORDINAL DATA.
(AML) an alternative method of model
More specifically, it tests for STATISTICAL
fitting in REGRESSION A N A L Y S I S . It is a
INDEPENDENCE between the variables.
form of MAXIMUM LIKELIHOOD used
a s s u m p t i o n n. one or more conditions when there is a high degree of scatter or
that need to be met in order for a statisti- dispersion around the REGRESSION LINE.
cal procedure to be fully justified from a asymmetry n. SKEWNESS: the condition
theoretical perspective. For example, in which the values of a data set are not
ANALYSIS OF VARIANCE aSSUmeS HOMO- arranged equally around a center point.
GENEITY OF VARIANCE and independ- Compare SYMMETRY. asymmetrical
ence of observations, among other adj.
criteria. If the assumptions were to be vi-
olated to an extreme extent, the results a s y m p t o t e n. a straight line that defines
would be invalid. See also ROBUSTNESS. the limit of a curve, such that the curve
continues to approach but never reaches
a s s u m p t i o n o f e q u a l v a r i a n c e see the line. The concept of asymptote is
HOMOGENEITY OF VARIANCE. often invoked in relation to the learning
curve. Participants in learning studies
a s s u m p t i o n o f i n d e p e n d e n c e the AS-
often show a steady improvement in
S U M P T I O N that scores in one data set are
performance that then levels off as the
unrelated to scores in another data set.
curve approaches asymptote; any fur-
a s s u m p t i o n o f n o r m a l i t y the AS- ther improvements will be minimal, re-
S U M P T I O N that a set of scores follow a gardless of further practice or training.
particular function described by a bell-
a s y m p t o t i c a l l y u n b i a s e d describing
shaped curve (i.e., the N O R M A L DISTRI-
a statistical procedure that on average
BUTION).
yields an increasingly true population
asymmetrical confidence interval value as the number of observations
a range of plausible values for an un- upon which it is based increases. Thus,
known population characteristic the upper an asymptotically unbiased estimator
and lower bounds of which are not equi- on average yields a correct calculation of
distant from the center of the range. a population characteristic as the sample
Thus, each side ofthe CONFIDENCE INTER- size used approaches infinity.
VAL is disfinct from the other. Compare a s y m p t o t i c c u r v e a curved line pos-
SYMMETRICAL CONFIDENCE INTERVAL. sessing an upper or lower limit (the
asymmetrical d i s t r i b u t i o n any ASYMPTOTE) that Is approached but
nonnormal distribution: a set of ordered never reached.
scores in which the frequency of values a s y m p t o t i c f l i s t r i b u t i o n a THEORET-
is not equal above and below the center ICAL DISTRIBUTION that takes the form
or midpoint of the set. If one divides the of an ASYMPTOTIC CURVE when plotted.
distribution at its M E A N , the arrange- Asymptotic distributions are used in hy-
ment of scores differs across the two pothesis testing, for example, to gener-
halves. See SKEWNESS. Compare S Y M - ate theoretically derived competing
M E T R I C A L DISTRIBUTION. distributions such as the N U L L DISTRIBU-
T I O N and the A L T E R N A T I V E HYPOTHESIS
a s y m m e t r i c a l test 1. any of various
DISTRIBUTION.
statistical methods used to evaluate an
entity whose range of values is not as- a s y m p t o t i c m e t h o d any statistical
sumed to be uniformly arranged about a procedure for esfimating a population

15
ATI

quantity that becomes more accurate as with an experimental manipulation or

the number of observations involved in intervention. Attribute variables may be
the calculation increases. Asymptotic CATEGORICAL or CONTINUOUS.
methods use assumptionsincluding
EFFICIENCY, NORMALITY, and SYMME- a t t r i t i o n n. the loss of study partici-
TRYthat generally are not absolutely pants over time. Attrition may occur for
true for finite samples but that become a variety of reasons (e.g., the nature of
steadily more tme as sample sizes increase. the data being collected, participant re-
Also called large-sample method. location, aversive or costly data collec-
tion procedures) and can threaten the
ATI abbreviation for aptitude-treatment EXTERNAL VALIDITY and INTERNAL VA-
interaction. See TRAIT-TREATMENT INTER- LIDITY of research. It also creates the po-
ACTION. tential for BIASindividuals who drop
attention-placebo c o n t r o l group a out may have unique characteristics that
set of research participants who are given are relevant to the phenomenon of in-
an inert treatment (PLACEBO) with cog- terest such that the remaining sample is
nitive requirements resembling those of no longer representative of the popula-
the actual treatment under investiga- tionand may reduce the POWER of sta-
tion. Their responses function as a psy- tistical analyses.
chologically neutral condition against
AUC abbreviation for AREA UNDER THE
which to assess the effectiveness of the
CURVE.
treatment under study. For example, a
researcher investigating a new imagery- audit n. an evaluation of a service, inter-
based stroke therapy would need to en- vention, or outcome. In research it refers
sure all participants have the same ex- to an examination of the soundness of a
perience, regardless of the experimental study's findings.
group into which they are placed, in
order to attribute any differences in out- a u d i t o r y test any test intended to eval-
come to the therapy. By replicating the uate a person's hearing ability. For ex-
nonactive characteristics necessary for ample, one might measure the differ-
delivery of the therapy, the researcher ence in time between presentation of a
thus ensures the members of an atten- sound and a participant's response, such
tion-placebo control group are subject to as pressing a specific letter on a key-
the same time, attention, and other cog- board.
nitive demands as they would be had
they received the actual treatment. autocorrelation n. the situation in
which values of a variable measured over
attenuation n. an underestimation of time are correlated with other values of
the size of an effect or relationship due the same series separated from them by a
to poor measurement or RESTRICTION OF specific interval. This often occurs with
RANGE. See also CORRECTION FOR AT- economic or demographic data. Auto-
TENUATION. correlations are generally assumed to be
attribute testing the assessment of the linear relafionships and may be presented
graphically in an autocorrelogram (or
abilities or skills possessed by a research
correlogram) or formulaically in an auto-
participant prior to the participant's in-
correlation function (ACF). Also called
volvement in a research study.
serial correlation. See TIME-SERIES
attribute variable a preexisting char- ANALYSIS.
acteristic of a study participant, such as
ethnic background or age, that is not autocovariance n. in a STOCHASTIC
changed during the course of the re- PROCESS, the COVARIATION of a variable
search but is considered in conjunction with earlier or later versions of itself.
 axis

Covariance implies that the variability is predicted by a previously observed score,

linear as opposed to nonlinear. and the second allows a score at one
point in time to be predicted by an un-
automatic interaction detector observed score at a previous occasion.
(AID) a method of accounting for varia- ARMA models incorporating additional
tion in a single response variable using a space-time variation are called SPACE-
set of predictors. Values of the response TIME AUTOREGRESSIVE MOVING-AVERAGE
variable are sequentially broken down MODELS.
into smaller and relatively more homo-
geneous subsets according to one or available-case analysis see PAIRWISE
more of the predictor variables that best DELETION.
achieve this goal. See also CHI-SQUARE average n. see MEAN.
AUTOMATIC INTERACTION DETECTOR.
average absolute deviation see
autoregression n. a pattern of relation- MEAN ABSOLUTE DEVIATION.
ship between repeated measures of a average deviation (AD) see MEAN AB-
variable taken over time, such that the SOLUTE DEVIATION.
variable as observed at one point in time
is predicted by the variable observed at average error the typical degree to
one or more earlier points in time. See which a series of observations are inaccu-
TIME-SERIES ANALYSIS. rate with respect to an absolute criterion
(e.g., a standard weight or length) or a
autoregressive integrated moving- relative criterion (e.g., the mean of the
average model (ARIMA model) a observations within a given condition).
strategy for modeling and forecasting average-linkage clustering a multi-
TIME-SERIES DATA that are presumed to variate procedure for grouping indi-
have a steady underlying trend. It is an viduals or entities into clusters (homo-
extension of the AUTOREGRESSIVE MOV- geneous groups) on the basis of a set of
ING-AVERAGE MODEL that examines the relevant variables. Each individual is as-
differences between successive values in signed to his or her own cluster initially
the time series instead of the values them- and at each subsequent step the most
selves. Also called Box-Jenkins model. similar pair of clusters are merged into a
new group. The process continues until
autoregressive model (AR model) a
all individuals have been combined into
model used to represent TIME SERIES that
demonstrate AUTOREGRESSION, that is, one large cluster. The process may be
where each successive observation de- mapped in a DENDROGRAM. Average-link-
pends, at least in part, on one or more age clustering is one of several forms of
AGGLOMERATIVE CLUSTERING procedures.
previously observed values.
Also called group-average clustering.
autoregressive moving-average Compare COMPLETE-LINKAGE CLUSTER-
model (ARMA model) a strategy for ING; SINGLE-LINKAGE CLUSTERING.
modeling and forecasting TIME-SERIES averages l a w see LAW OF LARGE NUM-
DATA that involve AUTOCORRELATION. BERS.
Incorporating components of both an
AUTOREGRESSIVE MODEL and a MOVING- A W S test abbreviation for ASPIN-
AVERAGE MODEL, it relies on two patterns WELCH-SATTERTHWAITE TEST.
of relationship between observations of axis n. (pl. axes) a fixed reference line in
a variable taken over time: The first al- a coordinate system. See also ABSCISSA;
lows a score at one point in time to be ORDINATE.

17
 Bb
b a c k g r o u n d variable see SUBJECT neously summarizing the data for both
VARIABLE. and showing the degree of relationship
between them. It includes a filled square
back-to-back stem-and-leaf plot a or other shape indicating the MEDIAN of
variation of a STEM-AND-LEAF PLOT in the scores, a bag or shaded polygon en-
which the trailing digits (leaves) from circling the middle 50% of scores, a loop
two data sets are displayed on either side or differently shaded polygon encircling
of the same central column of initial val- the remaining data points, and a fence
ues (stem). The leaf values from one data or curve marking the edge of the loop
set are given on the right-hand side beyond which any OUTLIERS are dis-
while those of the second group are shown played. The DISPERSION of the data
on the left-hand side. For example, con- points is indicated by the size of the bag,
sider the following hypothetical sets of while the shape and orientation of the
test scores obtained by participants on bag indicate SKEWNESS and correlation,
two different occasions: respectively. A generic example is given
79 79 83 83 86 87 92 99 opposite.
and
Balaam's design an experimental ap-
80 81 82 83 87 87 90 99 proach in which participants are divided
The back-to-back stem-and-leaf plot for into four groups and experience differ-
these values is ent conditions or treatments (A and B)
|7|99 in one of four possible sequences: AA,
012377|8|3367 BB, AB, or BA. More specifically, in one
09 I 9 I 2 9 sequence participants receive the first
treatment over two periods, in a second
back-translation see TRANSLATION sequence participants receive the second
AND BACK-TRANSLATION. treatment over two periods, in a third
b a c k w a r d e l i m i n a t i o n a technique sequence participants receive the first
used in REGRESSION ANALYSIS in which treatment in the first period and the sec-
the goal is to forecast an outcome or re- ond treatment in the second period, and
sponse variable according to a subset of in the fourth sequence participants re-
predictor variables narrowed down from ceive the second treatment in the first
a large initial set. In background elimina- period and the first treatment in the
tion, all available predictors are included second period. A type of CROSSOVER DE-
originally and then examined one at a SIGN, Balaam's design is used in studies
time, with any predictors that do not in which a treatment CARRYOVER EFFECT
contribute in a statistically meaningful is anticipated. [L. N. Balaam, Australian
manner systematically dropped until a biometrician]
predetermined criterion is reached. Also
called backward deletion; back- balanced design any research design
ward selection; stepdown selec- in which the number of observations or
tion. See also STEPWISE REGRESSION. measurements obtained in each experi-
mental condition is equal. For example,
bagplot n. an extension of a BOX-AND- a health researcher interested in exercise
WHiSKER PLOT to two Variables, simulta- and depression would be using a bal-

ls
 balanced replication

0)
.Q
.2

"T" T "T"
20 40 50

Variable A

bagplot

anced design if he or she examined 50 balanced longitudinal data obser-

people who exercise less than 30 min- vations or measurements collected from
utes per day, 50 people who exercise be- the same individuals on multiple occa-
tween 30 and 60 minutes per day, and sions and within the same time frame
50 people who exercise more than 60 for each period. For example, a re-
minutes per day. searcher may survey all of the children
in a particular classroom at 9 a.m., 12
balanced Latin square a type of study p.m., and 3 p.m. on a particular day. If,
design in which multiple conditions or however, the researcher were to survey
treatments are administered to the same some of the children at 9 a.m., some at
participants over time. It is a form of 12 p.m., and some at 3 p.m., then the
LATIN SQUARE that must fulfill three cri- observations would be unbalanced lon-
teria: Each treatment must occur once gitudinal data.
with each participant, each treatment
must occur the same number of times balanced repeated replication
for each time period or trial, and each (BRR) a statistical procedure used to cal-
treatment must precede and follow every culate the variance associated with an
other treatment an equal number of estimated value for a population PARAM-
times. For example, consider the follow- ETER. In this method, the data set is di-
ing balanced Latin square for an experi- vided into two equal portions and the
mental design involving four treatments parameter calculated for one of those
(A, B, C, and D) and four people: halves. The original full data set is then
divided a second time and a new param-
Treatment Order eter calculation obtained, with the pro-
1 2 3 4 cess continuing a given number of times
Participant 1 A B D C to generate a set of estimates from which
Participant 2 B C A D a measure of the variance can be calcu-
Participant 3 C D B A lated for the entire data set. BRR is simi-
Participant 4 D A C B
lar to other resampling techniques, such
As shown, one person receives the treat- as BOOTSTRAPPING and the JACKKNIFE,
ments in the sequence A, B, D, and then but uses half of the sample at a time and
C; a second person receives them in the also takes into account complex sample
sequence B, C, A, and D; a third person characteristics.
receives them in the sequence C, D, B, balanced r e p l i c a t i o n 1. a strategy for
and A; and a fourth person receives conducting a study at two or more loca-
them in the sequence D, A, C, and B. tions in which the different conditions

19
balanced scale

or treatments are randomly assigned to them are used to display data for vari-
participants and administered the same ables defined by qualities or categories.
number of times at each location. 2. an For example, to show the political affili-
occasional synonym for BALANCED RE- ations of Americans, the different parties
PEATED REPLICATION. would be listed on the x-ax'is, and the
height of the bar rising above each party
balanced scale a test or survey in which,
would represent the number or propor-
for each possible response, there is a re-
tion of people in that category. Also
sponse that means the opposite. A rating
called bar chart; bar diagram. See
scale with the four alternatives very poor,
also HISTOGRAM.
poor, good, and very good is an example, as
is a set of survey questions in which half
of the questions characterize a particular
trait (e.g., perceived stress level) in one
direction (e.g., low) and the other half
characterize the trait in the opposite di-
rection (e.g., high).
balancing score a score used for MATCH-
ING participants in a treatment group
and a comparison group on some vari-
able, with the goal of forming groups Democrat Republican Independent No Affiliation

that have identical distributions of values Political Party

for that variable. A PROPENSITY SCORE is
an example. Bartlett's correction a modification
applied to improve the quality of the sta-
band chart a type of LINE GRAPH in tistic obtained from the LIKELIHOOD-
which different subsets of data for the RATio TEST, whlch is uscd to evaluate
same variable are displayed in different hypotheses about the relationships
shaded areas or strata. For example, a among a set of variables. Also called Bart-
band chart might present overall weekly lett's adjustment (factor). [Maurice
sales of a company's product as well as Stevenson Bartlett (1910-2002), British
the portion of such that were completed statistician]
online and the portion completed
through calls to the customer service de- Bartlett test f o r eigenvalues in
partment. Also called stratum chart. PRINCIPAL COMPONENTS ANALYSIS, 3 pro-
b a n d i n g a strategy for personnel selec- cedure used to test for HOMOGENEITY OF
tion, whereby scores on an employment VARIANCE across a set of EIGENVALUES.
test are broken into bands based on its [Maurice Stevenson Bartlett]
RELIABILITY, STANDARD ERROR OF MEA-
Bartlett test f o r equality o f vari-
SUREMENT, and STANDARD ERROR OF ES-
ance a procedure used to determine
TIMATE. Banding assumes that there is
whether variation across two or more
substantial unreliability in any single
groups is equal, this being a critical as-
observed score and that candidates who
sumpfion in ANALYSIS OF VARIANCE and
score within the same band are equally
many other parametric methods. The
qualified. Consequently, hiring deci-
Bartlett test is sensitive to scores that are
sions are made from within band ranges,
not from a NORMAL DISTRIBUTION. Also
in contrast to traditional approaches
called Bartlett-Box test; Bartlett test
that choose the highest scorer first, the
for homogeneity of variance; Bart-
next highest scorer second, and so forth.
lett's test. [Maurice Stevenson Bartlett;
bar g r a p h a graph in which bars of George E. P. Box (1919- ), Brifish statis-
varying height with spaces between tician]
 Bayes estimator

Bartlett test of sphericity a method basement effect see FLOOR EFFECT.

often used in FACTOR ANALYSIS or PRIN-
base rate the naturally occurring fre-
CIPAL COMPONENTS ANALYSIS to deter-
quency of a phenomenon in a popula-
mine whether variables in a population
tion. An example is the percentage of
are unrelated. Specifically, the test is used
students at a particular college with
to assess whether a population CORRELA-
major depressive disorder. This rate is
TION MATRIX has ones along the diagonal
often contrasted with the rate of the
and zeros for the remaining elements, a
phenomenon under the influence of
pattern that implies that all of the vari-
some changed condition in order to de-
ables in the set are uncorrelated with
termine the degree to which the change
each other. See also MAUCHLY'S SPHERIC-
influences the phenomenon.
ITY TEST. [Maurice Stevenson Bartlett]
base year a 12-month period used as a
baseline n. data or information ob-
reference in constructing an index. A
tained prior to or at the onset of a study
base year is chosen (e.g., 2012) and set to
(e.g., before introduction of an interven-
some arbitrary value (e.g., 100), and dif-
tion) that serves as a basis for compari-
ferences between that base year and other
son with data collected at a later point in
years are then calculated to determine
time so as to assess the effects of par-
the amount of change.
ticular manipulations or treatments. For
example, a memory researcher may basic research research conducted to
measure how many words from a list obtain greater understanding of a phe-
parficipants remember inifially and then nomenon, explore a theory, or advance
compare that figure to the number of knowledge, with no consideration of any
words they remember following the use direct practical application. Also called
of a new mnemonic technique (i.e., the pure research. Compare APPLIED RE-
experimental manipulation). SEARCH.

baseline assessment a process of ob- basic science scientific research or the-

taining information about a participant's ory that is concerned with knowledge of
status (e.g., ability level, psychological fundamental phenomena and the laws
well-being) before exposure to an inter- that govern them, regardless of the po-
vention, treatment, or other study ma- tential applications of such knowledge.
nipulation. The information so obtained Also called fimdamental science; pure
is compared to data gathered during the science. Compare APPLIED SCIENCE.
course of the research in order to iden-
tify any effects associated with the study b a t h t u b curve a type of SURVIVAL
conditions. CURVE that has a specific shape with
three major parts: an initial decreasing
baseline characteristic any of vari- trend, which is followed by a relatively
ous qualities of a participant that are as- stable and unchanging trend, which in
sessed prior to a study or intervention as turn is followed by an increasing trend.
they may potentially influence outcome
measures. For example, a researcher Bayes estimator a rule for estimating
evaluating the effect of a new stress- the value of a population characteristic
reduction technique on health status that is based on both empirically ob-
may examine such baseline characteris- tained data and one's prior expectations
tics as sex, age, and ethnicity. regarding the parameter of interest. For
example, to estimate the mean student
baseline h a z a r d f u n c t i o n in SUR- score on a particular test using this
VIVAL ANALYSIS, the SURVIVAL FUNC- method, one would not only analyze the
TION obtained for an at-risk individual scores of a sample of students but also
when all COVARIATES are set at zero. take into account whatever information

21
Bayes factor

is already available about values on that Bayesian information criterion

variable in the population. [Thomas Bayes (BIC) in BAYESIAN statistics, a summary
(1702-1761), British mathematician and value used in comparing the relative fit
theologian] of one model to another for a given
set of data. Although adding to the com-
Bayes f a c t o r in BAYESIAN theory, a plexity of a model often will improve
quantification of the extent to which fit to a set of data, the Bayesian infor-
sample data favor one model over an- mation criterion adds a penalty for each
other: It is the probability of obtaining a such addition, such that relative model
given set of observed data assuming that fit also is judged in terms of model parsi-
the NULL HYPOTHESIS is tiue, divided by mony (simplicity). [Thomas Bayes]
the probability of obtaining the same
data assuming that a contradictory AL- Bayesian model averaging (BMA) a
TERNATIVE HYPOTHESIS is true. Small BAYESIAN approach to data analysis in
values are taken as evidence against the which two or more models are consid-
null hypothesis, as they indicate that ob- ered to account for patterns in observed
tained observations are relatively more data and the level of certainty (i.e.,
likely under the alternative hypothesis. PRIOR DISTRIBUTION) associated with
Also called Bayes ratio. [Thomas Bayes] each model is examined and combined
to yield a single POSTERIOR DISTRIBUTION
Bayesian adj. denoting an approach to for the set of models as a whole. The
statistical inference and probability that technique can be applied to any set of
enables previously known (a priori) in- plausible models of a certain class (e.g.,
formation about a population character- structural equation models, regression
istic of interest to be incorporated into models). [Thomas Bayes]
the analysis. In Bayesian methods, esti-
mated quantities thus are based in part Bayesian n e t w o r k a graphical repre-
on empirical data (i.e., what was actually sentation of a set of RANDOM VARIABLES
observed) and in part on collective or in- and their probabilistic relationships that
dividual knowledge about what to ex- takes into account uncertainty in the
pect in the population (as captured in a proposed system. For example, a Bayes-
PRIOR DISTRIBUTION). See also POSTE- ian network could display the likeliness
RIOR DISTRIBUTION. [Thomas Bayes] of being diagnosed with a particular ill-
ness given the presence of various symp-
Bayesian confidence i n t e r v a l see toms. [Thomas Bayes]
CREDIBLE INTERVAL. [Thomas Bayes]
Bayes theorem a formula for calculat-
Bayesian controversy the long- ing the probability that an event will
running debate about the relative merits occur that allows for the acquisition of
of the two major perspectives in statis- new information regarding that event.
tics and probability, namely, the BAYES- For example, consider the probability
IAN approach and the FREQUENTIST that an individual will have a stroke
approach. [Thomas Bayes] within the next year. Using Bayes theo-
rem, one could take an estimate of this
Bayesian inference a method of draw- probability based on general population
ing conclusions about a characteristic of data for that individual's age group (i.e.,
a population using both sample data the PRIOR PROBABILITY) and revise it to
and previously known information account for the results of that person's
about that characteristic. It relies upon stress tests and other cardiological mark-
BAYES THEOREM to derive POSTERIOR DIS- ers, creating what is known as the POS-
TRIBUTIONS from obtained observations TERIOR PROBABILITY. [Thomas Bayes]
and PRIOR DISTRIBUTIONS. Compare FRE-
QUENTIST INFERENCE. [Thomas Bayes] B C a (BC) abbreviation for BIAS-COR-

22
 Bentler-Bonett index

RECTED ACCELERATED PERCENTILE IN- or more observers. Behavior observation

TERVAL. may be carried out live or through video
B coefficient see REGRESSION COEFFI-
media and often involves the use of rat-
CIENT.
ing scales, checklists, or charts.

before-after design see PRETEST-

behavior r a t i n g use of a scale to assess
POSTTEST DESIGN.
the degree to which an individual dis-
plays one or more behaviors in a given
behavioral p r o f i l e an overall repre- situation.
sentation of the behavioral characteris-
tics of a participant or group of par- behavior s a m p l i n g a data collection
ticipants in a test or experiment, often method in which the behaviors of one
used to study patterns and trends. or more individuals are observed and re-
corded during a designated period of
behavioral science any of various dis- time. Observations can be made in ei-
ciplines devoted to the scientific study ther a natural or a research setting, and
of human and animal actions and reac- they may be conducted over multiple
tions using systematic observation and time periods with or without the aware-
experimentafion. Psychology, sociology, ness of the individual being observed.
and anthropology are examples.
Behrens-Fisher p r o b l e m the prob-
behavior checklist a list of actions, re- lem of comparing the means of two sam-
sponses, or other behaviors that are to be ples drawn from different populations
recorded each time they are observed, as when the variances of these populations
by an experimental investigator, study are not assumed to be equal. In other
participant, or clinician. words, the problem involves the diffi-
behavior coding a method of record- culty of making statistical inferences in
ing observations in which defined labels the absence of equal population vari-
are used by specially trained individuals ances. The Behrens-Fisher problem may
to denote specific qualities and charac- be addressed in many ways, such as by
teristics of behaviors as they are wit- using a T TEST to estimate a PROBABILITY
nessed. For example, a developmental LEVEL (p value) or by using CONFIDENCE
researcher might use the following cod- INTERVALS to estimate the range of un-
ing scheme for infant vocalizations: certainty around the mean difference.
(1) vowels; (2) syllables (i.e., consonant- [W. U. Behrens; Sir Ronald Aylmer
vowel transitions); (3) babbling (a se- Fisher (1890-1962), British statistician]
quence of repeated syllables); and (4)
other (e.g., cry, laugh, vegetative sounds). bell curve the characteristic curve ob-
Behavior coding systems typically are tained by plotting the values of a NOR-
specific to a given study and the behav- MAL DISTRIBUTION. The shape resembles
iors under investigation. a cross-sectional representation of a bell
(i.e., a large rounded central peak taper-
behavior d i a r y a detailed log of a per- ing off on either side). Also called bell-
son's activities or responses as recorded shaped distribution.
by a trained observer or the individual
Bentler-Bonett index in STRUCTURAL
him- or herself, usually on a regular basis
EQUATION MODELING, a value that re-
over a specified interval of time. For ex-
flects the degree to which a hypothesized
ample, a behavior diary might be used to
pattern of relationships among variables
log a study participant's activities over a
is consistent with the observed pattern
24-hour period.
in a specific data set, relative to an alter-
behavior observation the recording native pattern that assumes that the
or evaluation of the ongoing behavior of variables are entirely unrelated to one
one or more research participants by one another. It differs from the Bentier com-

23
Bernoulli distribution

parative FIT INDEX (CFI) in being normed. take on only one of two possible values
Also called normed f i t index (NFl). indicating either success or failure.
[Peter M. Bentier, U.S. quantitative psy- Usually a value of 0 is used to denote
chologist; Douglas G. Bonett, U.S. psy- a failure and a value of 1 is used to de-
chometrician] note a success. See BERNOULLI PRO-
CESS; BERNOULLI DISTRIBUTION. [JacqUCS
B e r n o u l l i d i s t r i b u t i o n a theoretical
distribution of the number of trials re- Bernoulli]
quired before the first success is obtained Bessel's correction an adjustment
in a BERNOULLI PROCESS. Such a distri- used to reduce bias when calculating the
bution is defined by two values: 0 and 1. VARIANCE and S T A N D A R D DEVIATION in
Usually a value of 0 is used to denote a a sample. Specifically, it is the use of n -
failure (i.e., the item of interest does not 1 instead of n in the denominator of the
occur) and a value of 1 is used to denote computational formulas used to com-
a success (i.e., the item of interest does pute these values, where n is the number
occur). On this basis, the likeliness of a of observations in the sample. [Friedrich
success is denoted as p and the likeliness Wilhelm Bessel (1784-1846), German-
of a failure is denoted asq = l - p . For ex- born Russian astronomer and mathema-
ample, a single toss of a coin has a Ber- tician]
noulli distribution with p = 0.5 (where 0
= heads and 1 = tails). A Bernoulli distri- best estimator the statistic obtained
bution is a special case of a BINOMIAL from a sample data set that over re-
DISTRIBUTION. [Jacques Bernoulli (1654- peated sampling exhibits the least
1705), Swiss mathematician and scien- amount of variance relative to other sta-
tist] tistics in estimating the corresponding
value in the larger population.
B e r n o u l l i process a sequence of unre-
lated events whose outcome values can best f i t the theoretical pattern that best
only be equal to 0 or 1 (denoting either accounts for the relafionships among
failure or success) with each outcome variables in a data set. For example, a RE-
having the same probability. For exam- GRESSION EQUATION having the best fit
ple, the results from tossing a coin a to sample data is the one that minimizes
given number of times when the condi- differences between observed and pre-
tions for each toss of the coin are iden- dicted values. On a SCATTERPLOT, a line
tical could be described as a Bernoulli of best fit provides a visual depiction of
process. In other words, a Bernoulli pro- this pattern, allowing extrapolation to
cess is a sequence of BERNOULLI TRIALS. values not part of the original data set. It
[Jacques Bernoulli] is important to note that the best-fitting
pattern is not necessarily the one that
Bernoulli's theorem see LAW OF LARGE generated the observed data, as other
NUMBERS. [Jacques Bernoulli] patterns not considered may provide a
B e r n o u l l i t r i a l a single experiment in superior fit.
which the only two possible outcomes
best linear unbiased estimator
are success or failure. Usually a value of 0
(BLUE) the LINEAR EQUATION that most
is used to denote a failure (i.e., the item
accurately characterizes the relationship
of interest does not occur) and a value of
between an outcome variable and a set
1 is used to denote a success (i.e., the
of predictor variables when applied to a
item does occur). A sequence of Ber-
sample data set. Thus, the quantity ob-
noulli trials is known as a BERNOULLI
tained from a BLUE on average equals
PROCESS. See also BERNOULLI DISTRIBU-
the quantity in the larger population
TION. Oacques Bernoulli]
from which it is derived and over re-
B e r n o u l l i variable a variable that can peated sampling has the least amount of
 between-groups variance

variation relative to other procedures. the number of groups to be compared

See also GAUSS-MARKOV THEOREM. minus one. For example, the between-
beta (symbol: P) n. 1. the likeliness of groups DEGREES OF FREEDOM in a study
committing a TYPE II ERROR in research.
examining the effectiveness of four dif-
2. the effect of a predictor variable on an ferent therapies for depression would be
outcome variable, such as in REGRESSION 3 (i.e., one less than the number of ther-
ANALYSIS or STRUCTURAL EQUATION apy groups being examined). This quan-
tity is used to calculate the BETWEEN-
MODELING.
GROUPS MEAN SQUARE. Also called be-
beta coefficient the multiplicative tween-conditions degrees of free-
constant in a REGRESSION ANALYSIS that dom; between-subjects degrees of
indicates the change in an outcome vari- freedom; between-treatments de-
able associated with the change in a pre- grees of freedom. Compare WITHIN-
dictor variable after each has been GROUPS DEGREES OF FREEDOM.
standardized to have a distribution with
a mean of 0 and a STANDARD DEVIATION between-groups design see BETWEEN-
of .1. A beta coefficient controls for the SUBJECTS DESIGN.
effect of other predictors included in the between-groups mean square in an
analysis and thus allows for direct com- ANALYSIS OF VARIANCE, an Index of the
parison of the individual variables de- variance among the different levels of
spite differences in measurement scales the INDEPENDENT VARIABLE being ana-
(e.g., salary measured in dollars and edu- lyzed: It is calculated by dividing the BE-
cational degree measured in years of TWEEN-GROUPS SUM OF SQUARES by the
schooling). Also called beta weight; BETWEEN-GROUPS DEGREES OF FREEDOM.
standardized regression coeffi- The between-groups mean square forms
cient. the numerator of the F RATIO. Also called
beta d i s t r i b u t i o n a family of PROBA- between-conditions mean square;
BILITY DISTRIBUTIONS whose values between-groups variance; between-
range from 0 to 1 and whose shape is de- subjects mean square; between-sub-
fined by two quantities (parameters) re- jects variance; between-treatments
ferred to as a and (i. A beta distribution mean square; between-treatments
provides the likelihood of obtaining each variance; mean square between.
of the possible values within the 0, 1 Compare WITHIN-GROUPS MEAN SQUARE.
range.
between-groups sum of squares in
beta error see TYPE II ERROR. an ANALYSIS OF VARIANCE, a quantity
beta level the likeliness of failing to re- that is used to determine the BETWEEN-
ject the NULL HYPOTHESIS when it is in GROUPS MEAN SQUARE. It is calculated by
fact false; that is, the probability of mak- adding together the squared differences
ing a TYPE II ERROR. The beta level for between each group average (MEAN) in a
a given statistical procedure is related set and the overall average (GRAND
to the POWER of that procedure (P level = MEAN), with each squared difference
1 - power). weighted by the number of members in
the respective group. Also called be-
beta weight see BETA COEFFICIENT. tween-conditions sum of squares;
between-groups analysis of vari- between-subjects sum of squares;
ance see BETWEEN-SUBJECTS ANALYSIS between-treatments sum of squares.
OF VARIANCE. Compare WITHIN-GROUPS S U M OF
SQUARES.
between-groups degrees o f free-
dom in an ANALYSIS OF VARIANCE for between-groups variance see BE-
data from a BETWEEN-SUBJECTS DESIGN, TWEEN-GROUPS MEAN SQUARE.

25
between-subjects analysis of variance

between-subjects analysis of vari- SIS OF VARIANCE, an INDEPENDENT VARI-

ance an examination of the variance in ABLE with multiple levels, each of which
a set of data obtained from a study in is assigned to or experienced by a dis-
which a different group of participants is tinct group of participants. In a study
randomly assigned to each condition. In examining weight loss, for example,
this procedure, the researcher determines the different amounts of daily exercise
how much of the variance in the ob- under investigation would be a between-
tained data is due to the influence of the subjects factor if each was undertaken
variables under investigation and how by a different set of people. Also called
much is due to random differences among between-subjects variable.
the group members. For example, a re-
between-subjects mean square see
searcher studying how amount of daily
BETWEEN-GROUPS MEAN SQUARE.
walking (e.g., none, 30 minutes, 60 min-
utes, 90 minutes) affects quality of sleep between-subjects sum of squares
might divide participants into different see BETWEEN-GROUPS SUM OF SQUARES.
groups that each walk for one of the
between-subjects variable see BE-
specified lengths of time and then evalu-
TWEEN-SUBJECTS FACTOR.
ate the results using a between-subjects
analysis of variance. Also called be- between-subjects variance see BE-
tween-groups analysis of variance; TWEEN-GROUPS MEAN SQUARE.
independent-groups analysis of
Bhattacharya distance (Bhattach-
variance; independent-measures
aryya distance) a measure of the degree
analysis of variance; independent-
of similarity between two PROBABILITY
samples analysis of variance. Com-
DISTRIBUTIONS, A Bhattacharya distance
pare WITHIN-SUBJECTS ANALYSIS OF of 1 indicates that the two distributions
VARIANCE. are completely different, whereas a Bhat-
between-subjects degrees o f free- tacharya distance of 0 indicates that the
d o m see BETWEEN-GROUPS DEGREES OF two distribufions are probabalisfically the
FREEDOM. same. [A. Bhattacharya, Indian staHsttcian]
between-subjects design a study in bias n. 1. systematic error arising during
which individuals are assigned to only SAMPLING, data collection, or data anal-
one treatment or experimental condi- ysis. See BIASED ESTIMATOR; BIASED SAM-
tion and each person provides only one PLING. 2. any deviation of a measured or
score for data analysis. For example, in a calculated quantity from its actual (true)
between-subjects design investigating value, such that the measurement or cal-
the efficacy of three different drugs for culation is unrepresentative of the item
treating depression, one group of de- of interest. 3. any tendency or prefer-
pressed individuals would receive one of ence, such as a RESPONSE BIAS. biased
the drugs, a different group would re- adj.
ceive another one of the drugs, and yet
bias-corrected accelerated percen-
another group would receive the re-
tile i n t e r v a l (BCa; BC) a method of
maining drug. Thus, the researcher is
obtaining a CONFIDENCE INTERVAL for a
comparing the effect each medication
quantity of interest in which BOOT-
has on a different set of people. Also
STRAPPING procedures are used to en-
called between-groups design; inde-
hance EFFICIENCY and ROBUSTNESS. This
pendent-groups design; indepen-
may be necessary where sample sizes are
dent-measures design; independent-
small. For example, in determining a BCa
samples design. Compare WITHIN-
for a POPULATION MEAN, a researcher
SUBJECTS DESIGN.
would draw a subset of values from the
between-subjects factor in an ANALY- larger set of observed scores, determine
 binomial confidence interval

the confidence interval for that subset, BIC abbreviation for BAYESIAN INFORMA-
return the values to the larger pool, and TION CRITERION.
then repeat the process numerous times. b i m o d a l d i s t r i b u t i o n a set of scores
He or she would then calculate a single with two peaks or MODES around which
overall confidence interval from the set values tend to cluster, such that the
of intervals obtained. frequencies at first increase and then de-
biased estimator a value obtained crease around each peak, as in the fol-
from sample data that consistently un- lowing hypothetical depiction.
der- or overestimates the true quantity
in the larger population of interest. In
other words, a biased estimator is one
whose value on average differs from the
value of the PARAMETER it purports to
represent. Also called biased statistic.
Compare UNBIASED ESTIMATOR.
biased s a m p l i n g selecting individuals
or other study units from a population
in such a manner that the resulting sam- As shown, when graphing the heights
ple is not representative of the popula- of a sample of adolescents one would ob-
tion. Compare UNBIASED SAMPLING. See tain a bimodal distribution if most peo-
also SAMPLING BIAS. ple were either 5'7" or 5'9" tall. See also
MULTIMODAL DISTRIBUTION; UNIMODAL
biased statistic see BIASED ESTIMATOR.
DISTRIBUTION.
bias-variance t r a d e o f f the situation b i n a r y t r i a l a single experiment in
in which increasing the complexity of a which there are only two possible out-
model to better account for the observed comes. For example, consider a behav-
relationships among a set of variables ioral researcher observing aggression in
reduces the precision of any estimated a group of schoolchildren during recess.
values subsequently derived from that If he or she makes a record every 1 min-
model. That is, the model becomes so
ute, then each such period may be con-
specific to the particular sample from sidered a binary trial, with aggression
which it was derived that it no longer is present or aggression absent as its only two
an accurate representation of the larger possible outcomes.
population of interest. For example, a
REGRESSION EQUATION for predicting a b i n a r y variable see DICHOTOMOUS
person's income may include two vari- VARIABLE.
ables (education level and career field) b i n o m i a l n. an algebraic expression
that only partially account for variance consisting of two terms, such as (x + y)^
in salary but nonetheless are applicable or 2x + 4y. See BINOMIAL EXPANSION; BI-
to a broad group of individuals beyond NOMIAL THEOREM. See also POLYNOMIAL.
those studied. Adding further variables
to the equation (e.g., area of residence, b i n o m i a l confidence i n t e r v a l a
parents' education levels, prior experi- CONFIDENCE INTERVAL for estimates of
ence, age, sex, ethnicity) may ensure the proportion of a population that ex-
that all variance is fully accounted for, hibits a particular characteristic or out-
but only in the limited set of individuals come, as determined from a sample of
studied. Researchers therefore tend to that population. A BINOMIAL DISTRIBU-
seek a balance, opting for either greater TION is assumed. For example, one could
intricacy or greater certainty as circum- sample the pass rate for students in sev-
stances warrant. eral different states on a nationally ad-

27
binomial distribution

ministered mathematics exam and then the categorical proportions in the ob-
use the information to compute a bino- tained data differ significantly from
mial confidence interval indicating a their proportions in the population
range of values for the proportion of all from which they are believed to derive.
U.S. students who will pass the exam. See also BINOMIAL PROBABILITY.
b i n o m i a l d i s t r i b u t i o n the PROBABIL- b i n o m i a l theorem the mathematical
ITY DISTRIBUTION for each possible se- rule that allows one to reformulate an al-
quence of outcomes on a variable that gebraic expression of two termsfor ex-
has only two possible outcomes, with ample, (x + y)^ or (x + yf'as a longer
the likeliness of obtaining each outcome series of sums. See BINOMIAL EXPANSION.
remaining constant. For example, a bi-
b i n o m i a l variable a variable for
nomial distribution of the results of try-
which there are only two outcomes, typ-
ing to predict the outcomes of 10 coin
ically designated 0 and 1, with each
tosses would display the probability of
possible outcome having a specified
observing each possible set of results, in-
likeliness of being observed. Flipping a
cluding one success and nine failures,
coin is an example, as it may land on ei-
two successes and eight failures, and so
ther heads or tails. See also BINOMIAL
forth. This distribution is often denoted
DISTRIBUTION.
by b(,9).
b i o g r a p h i c a l method the systematic
binomial expansion the EXPANSION
use of personal historiesgathered
of an algebraic expression consisting of
through such means as interviews, focus
two terms into a larger series of sums.
groups, observations, and individual re-
For example, a possible binomial expan-
flections and other narrativesin psy-
sion of (x + y)^ is (x + y)(x + y) or x^ + xy +
chological research and analysis. This
yx + / .
method emphasizes the placement of
b i n o m i a l p r o b a b i l i t y within a series the individual within the context of so-
of independent observations, the proba- cial connections, historical events, and
bility of observing a particular combina- life experiences.
tion of outcomes for a variable that can
biostatistics n. the branch of statistics
assume only one of two values (e.g., 0 or concerned with collecfing and analyzing
1; pass or fail) where there is a fixed data that pertain to biological processes
likeliness of each value occurring. For or health characteristics, especially in
example, consider a group of 10 individ- medicine and epidemiology. Also called
uals selected from a population of per- biometrics; biometry. biostatis-
sons aged 40-45, with each selection tical adj. biostatistician n.
unrelated to any other selection. A bino-
mial probability could be calculated to b i p l o t n. a multivariate graphic that si-
determine the likeliness that, for exam- multaneously displays information
ple, seven of those individuals are mar- about variables and participants. Data
ried while three are not, given known about two or more variables usually are
information on the probability of per- denoted by arrows or vectors, and spe-
sons of that age being married. cific measurements for individual cases
usually are denoted by dots. The exam-
b i n o m i a l sign test see SIGN TEST. ple (opposite) shows the relationship
b i n o m i a l test a statistical procedure to among five task variables: the Boston
determine whether an observed data Naming Test, various letter and category
pattern for a variable that can have only fluency tests, an adapted Simon task, the
one of two values (often represented as 0 Stoop Color-Word Interference Test,
and 1) matches a theoretical or expected and the Sustained Attention to Response
pattern. In other words, it tests whether Task. The dimensions indicate the per-

28
 bivariate normality

Simon \ Stroop

E
5

"I 1 i 1 1 1 1

Dimension 1 (33.1%)

biplot

centage of variance in aspects of lexical bivariate adj. characterized by two vari-

access and execufive control explained by ables or attributes. For example, a set of
each. Biplots often are used in PRINCIPAL height and weight measurements for
COMPONENTS ANALYSIS. each participant in a study would be
bivariate data. Compare MULTIVARIATE;
b i p o l a r f a c t o r in FACTOR ANALYSIS, a UNIVARIATE.
latent variable whose values (FACTOR
LOADINGS) range from positive to nega- bivariate distribution a distribution
tive with a neutral point at a relatively showing each possible combination
central position within the range. of values for two RANDOM VARIABLES ac-
cording to their probability of occur-
b i p o l a r r a t i n g scale a set of response rence. For example, a bivariate distrib-
options whose lower- and upper-end ution may show the probability of ob-
terms are opposites. For example, a bipo- taining specific pairs of heights and
lar rating scale for motivation might in- weights among college students. Also
clude response options that range from called bivariate probability distri-
very strong to very weak. Also called bipo- bution.
lar scale. Compare UNIPOLAR RATING
SCALE. See also SEMANTIC DIFFERENTIAL. bivariate frequency d i s t r i b u t i o n a
distribution showing each possible com-
b i r t h cohort see COHORT. bination of two CATEGORICAL VARI-
ABLES according to their observed fre-
birth-cohort study a LONGITUDINAL
quency. For example, a researcher may
DESIGN examining groups of individuals
use a bivariate frequency distribution to
who were born around the same time
(e.g., in the same year). display how many male and female stu-
dents at a university are majoring in par-
biserial correlation coefficient ticularfieldsof study.
(symbol: r^, r^is) a measure of the bivariate method any approach to
strength of the association between a data analysis that relies on the simulta-
variable whose values may span a given neous examination of two variables.
range (e.g., grade point average) and a
CORRELATION ANALYSIS is an example of
variable with only two discrete values
a bivariate method.
(e.g., 0 or 1; pass or fail). It is distinct
from the POINT BISERIAL CORRELATION bivariate n o r m a l i t y a joint quality
COEFFICIENT in assuming that there is a of two variables, whereby at each level
NORMAL DISTRIBUTION underlying most of the first variable values of the second
characteristics. Also called biserial r. follow a NORMAL DISTRIBUTION, and

29
bivariate outlier

vice versa. Bivariate normality is an ular aspects of the study from which the
important assumption in certain statisti- data derive.
cal procedures, such as CORRELATION
ANALYSIS. b l i n d review an evaluation of a manu-
script to assess its suitability for publica-
bivariate o u t l i e r a data point whose tion or of a grant proposal to assess its
combination of values on two variables suitability for funding by a person who
is extreme or unusual relative to the does not know the identity of the author
other combinations in a data set. For ex- or proposer.
ample, a student who obtains perfect
scores of 100 points on two different block n. 1. a group or subset of study par-
math tests would be a bivariate outlier if ticipants who share a certain characteris-
most other students had combination tic and are treated as a unit in an
scores between 75 and 90 points. experimental design. 2. a set of variables
entered as a single entity into a REGRES-
bivariate probability distribution SION ANALYSIS or similar statistical pro-
see BIVARIATE DISTRIBUTION. cedure.
black box a system in which both the
input and output are observable but the block design a type of research study in
processes that occur between them are which participants are divided into rela-
unknown or not observable. For exam- tively homogeneous subsets (blocks)
ple, in the relationship between leader- from which they are assigned to the ex-
ship ability (input) and on-the-job perimental or treatment conditions. For
performance as a project manager (out- example, in a simple block design to
put), the role of the organizational cli- evaluate the efficacy of several antide-
mate could be considered a black box if pressants, participants with similar pre-
it is not understood. test depression scores might be grouped
into homogeneous blocks and then as-
b l i n d adj. denoting a research procedure signed to receive different medications.
in which information about particular The purpose of a block design is to en-
aspects or protocols of a study is un- sure that a characteristic of the study
known to participants, investigators, or participants that is related to the target
both. A single blind (or single masked) outcome (i.e., a COVARIATE) is distrib-
procedure is one in which the study par- uted equally across treatment condi-
ticipants do not know the experimental tions. See COMPLETE BLOCK DESIGN; IN-
conditions or groups to which they have COMPLETE BLOCK DESIGN; RANDOMIZED
been assigned; a double blind (or double BLOCK DESIGN.
masked) procedure is one in which nei-
ther the study participants nor the re- b l o c k d i a g r a m a graphical representa-
searchers who interact with them know tion of a system in which blocks repre-
which people have been assigned to sent key parts and connecting lines
which groups; and a triple blind (or tri- indicate relationships between parts. For
ple masked) procedure is one in which example, a block diagram could be used
the study participants, researchers, and to display the decision rules one might
individuals evaluating the data are all follow when conducting statistical tests
unaware of the specific study conditions to compare data from groups of study
to which people were assigned. Blinding participants.
is undertaken to prevent conscious or
unconscious BIAS by eliminating knowl- b l o c k i n g n. the process of grouping re-
edge that could skew results. search participants into relatively ho-
mogeneous subsets on the basis of a
b l i n d analysis an evaluation of data in particular characteristic. Such blocking
which the analyst is not aware of partic- helps adjust for preexisting patterns of
30
 box-and-whisker plot

variation between experimental units. Emilio Bonferroni (1892-1960), Italian

See BLOCK DESIGN. mathematician]
b l o c k i n g f a c t o r in a BLOCK DESIGN, an B o n f e r r o n i t test see DUNN-BONFER-
attribute or variable used as a basis for RONi PROCEDURE. [Carl Emilio Bonfer-
subdividing study participants or other roni]
sampling units into homogeneous sub- i
sets. Boole's i n e q u a l i t y a rule in probabil-
ity theory stating that the probability of
b l o c k r a n d o m i z a t i o n a method for one event from a fixed set of events oc-
assigning study participants to experi- curring is at most equal to the sum of the
mental conditions in which individuals probabilities of each individual event.
are arbitrarily divided into subsets or [George Boole (1815-1864), Brifish math-
blocks and then some random process is ematician and logician]
used to place individuals from those bootstrapping n. a statistical tech-
blocks into the different conditions. For nique to estimate the variance of a PA-
example, a researcher might divide par- RAMETER when standard assumptions
ticipants into blocks of 10 and then ran- about the shape of the data set are not
domly assign half of the people in each met. For example, bootstrapping may be
to the CONTROL GROUP and half to the used to estimate the variance of a set of
EXPERIMENTAL GROUP. Block random-
scores that do not follow a NORMAL DIS-
ization is distinct from BLOCKING in that TRIBUTION. In this procedure, a subset of
the block does not have any significance values is taken from the data set, a quan-
other than as an assignment unit. tity (e.g., the mean) calculated, and the
b l o c k s a m p l i n g a technique, mainly values reinserted into the data; this se-
used as part of a multistage procedure, quence is repeated a given number of
for selecting units for study. A popula- times. From the resulting set of calcu-
tion is divided into groups (blocks) that lated values (e.g., the set of means), the
each have approximately the same num- sunimary value of interest is calculated
ber of targets (e.g., adults to be inter- (e.g., the standard deviation of the
viewed), a random subset of those blocks mean). See also JACKKNIFE.
is chosen, and a random subset of targets bot1:om-up clustering (bottom-up
within each selected block is identified. hierarchical clustering) see AGGLOM-
This type of sampling helps to ensure
ERATIVE CLUSTERING.
that characteristics of the initial popula-
.
tion are well represented in the final
box-and-whisker plot a graphical
sample.
display of the central value, variance,
BLUE acronym for BEST LINEAR UNBI- and extreme values in a data set. A rect-
ASED ESTIMATOR. angle (box) is drawn along the vertical y-
axis of the plot, which shows the range
B M A abbreviation for BAYESIAN MODEL of data values. The length of the box in-
AVERAGING. dicates the middle 50% of scores and its
two ends indicate the upper and lower
B o n f e r r o n i i n e q u a l i t y a general rule
HINGES. Lines (whiskers) extending out-
of probability used when conducting
ward from the box denote variation in
MULTIPLE COMPARISONS on the Same
the upper and lower 25% of scores, while
data set. It states that the probability of
a separate line within the box indicates
at least one comparison resulting in a
the score that falls in the very middle of
TYPE I ERROR (i.e., incorrectly rejecting
the !set (i.e., the MEDIAN). Stars or other
the NULL HYPOTHESIS) Cannot exceed
single points indicate extreme scores.
the sum of the individual probabilities
Consider the following example (over-
from each separate comparison. [Carl
leaf).
Box-Cox transformation

showing a UNIFORM DISTRIBUTION are

changed to conform to a NORMAL DIS-
ID 500
TRIBUTION. The Box-Miiller transfor-
mation commonly is used in the com-
O puter generation of data. [George E. P.
o Box; Mervin E. Muller, computer scien-
in
o
tist]
o
Box-Pierce test a method for evaluat-
ing the assiimption that scores in a TIME
SERIES are unrelated. More specifically, it
uses an AUTOREGRESSIVE INTEGRATED
MOVING-AVERAGE MODEL to determine
901h %ile
75tti %lle
whether there is any AUTOCORRELATION
50tti %ile
25tli %lle
among the RESIDUALS of the observa-
10th %lle tions. See also LJUNG-BOX TEST. [George
E. P. Box; David A. Pierce]
box-score method a method of sum-
A box-and-whisker plot is useful in EX- marizing results from multiple studies of
PLORATORY DATA ANALYSIS for indicafing the same general research question that
whether a distribution is skewed and considers both statistically meaningful
whether the data set includes any OUT- and nonmeaningful results but ignores
LIERS; it can also be used to compare the magnitude of the research findings
data sets when large numbers of obser- (e.g., the size of a treatment effect).
vations are involved. Also called box- META-ANALYSIS generally is preferred
and-whisker diagram; box-and- over the box-score method.
whisker display; boxplot.
Box's test a statistical procedure to eval-
Box-Cox t r a n s f o r m a t i o n a TRANS- uate the similarity of two or more
FORMATION that enables the relation- SQUARE MATRICES showing the extent to
ship between one or more predictor which variance in each variable from a
variables and an outcome variable to be set is related to variance in all other vari-
described by a summative formula and ables in the set. In other words, it is a
thus to be plotted by a straight line when method to test the hypothesis that
graphed. Based on MAXIMUM LIKELIHOOD COVARIANCE MATRICES are equal. Ty-
estimation, the technique transforms pically used in MULTIVARIATE ANALYSES
the outcome variable to obtain LINEAR- OF VARIANCE and similar to the BART-
ITY and approximate NORMALITY in a LETT TEST FOR EQUALITY OF VARIANCE,
data set. Compare BOX-TIDWELL TRANS- Box's test is not appropriate for NONLIN-
FORMATION. [George E. P. Box (1919- ) EAR data and has limited applicability to
and David Roxbee Cox (1924- ), British data that do not follow a NORMAL DIS-
statisticians] TRIBUTION. Also called Box's M test;
Box's test for equality of covari-
Box-Jenkins model see AUTOREGRES-
ance matrices. [George E. P. Box]
SIVE INTEGRATED MOVING-AVERAGE
MODEL. [George E. P. Box; Gwilym M . B o x - T i d w e l l t r a n s f o r m a t i o n a
Jenkins (1933-1982), Brifish stafistician] TRANSFORMATION used to modify a set
of PREDICTOR VARIABLES SO that the re-
B o x - L j u n g test see LJUNG-BOX TEST.
lationship between those predictors and
Box-MiiUer transformation a TRANS- the outcome variable resembles a sttaight
FORMATION in which scores obtained line. It is similar to the B O X - C O X TRANS-
through a process that generates values FORMATION for correcting nonlinearity
32
 burn-in

but instead is applied to the independent ance, is a helpful preliminary step in de-
variables (predictors) involved. [George termining whether the required AS-
E. P. Box; Paul W. Tidwell] SUMPTIONS have been met for certain
statistical procedures, such as ANALYSIS
b r a c k e t i n g n. in estimation problems, OF VARIANCE. [Morton B. Brown, Cana-
the specification of a range or interval diah-born U.S. statistician; Alan B.
known to contain a target value. For ex- Forsythe, U.S. statistician]
ample, the probability of an event may i
be expressed asp < .05 or as .01 <p < .05. B r o w n - S p e a r m a n f o r m u l a see
SPEARMAN-BROWN PROPHECY FORMULA.
Bradley-Terry model a model used to I
1
create a scale of preferences from data B R R abbreviation for BALANCED RE-
obtained from PAIRWISE COMPARISONS PEATED REPLICATION.
of items. It is widely used to estimate the
probability of certain items being pre- b r u s h i n g n. an interactive data-analysis
ferred over others. For example, the technique in which a computer user
Bradley-Terry model might be applied manipulates several multidimensional
in determining whether a depressed in- SCATTERPLOTS in real time, using a
dividual will opt to receive one treat- mouse or other device to continuously
ment (A) rather than another treatment change display features and immedi-
(B). [Ralph A. Bradley (1923-2001), Ca- ately view the results.
nadian-born U.S. statistician; Milton E. bubble plot a graphical representation
Terry, U.S. statistician] of data that is similar to a SCATTERPLOT
b r e a k d o w n p o i n t the smallest num- butj includes an additional third vari-
ber or proportion of extreme observa- able whose values are represented by dif-
tions (OUTLIERS) that can be present in a ferently sized circles. For example, a
data set before an ESTIMATOR (e.g., the researcher might use a bubble plot to
sample mean) WiU assume arbittarily large show the relationship between income,
values. years of education, and age within a
particular profession. The first two vari-
Breslow-Day test a test for the homo- ables would be arrayed along the hori-
geneity of ODDS RATIOS across different zontal A;-axis and vertical y-axis, re-
levels of a variable (e.g., participant in- spectively, with circles placed in the
comes) within a data set. [Norman E. graph to indicate points of intersecting
Breslow (1941- ), U.S. biostatistician; values; the radius of each circle would
Nicholas E. Day (1939- ), British bio- correspond to values of age, with larger
statistician and epidemiologist] circles indicating greater age (see exam-
ple overleaf).
Breslow test a method used in SUR-
VIVAL ANALYSIS to compare the out- b u m p h u n t i n g an examination of the
comes of different groups of individuals. distribution of values in a data set to
More specifically, it evaluates the equal- identify any SPIKES or MODES; in graphic
ity of two or more SURVIVAL FUNCTIONS terms, it is a search for any significant
depicting mortality or failure across "bumps" on an otherwise smooth curve.
time. Also called Gehan's generalized Bump hunting is useful for finding por-
Wilcoxon test. [Norman E. Breslow] tions of data that are different enough to
potentially represent distinct subgroups
Brown-Forsythe test a statistical of cases.
method for determining whether there
is HOMOGENEITY OF VARIANCE across two b u r n - i n n. in the MARKOV CHAIN MONTE
or more populations. The test, which uses CARLO METHOD, an initial series of sam-
the absolute values of the deviations of pling runs whose values are discarded
scores from the MEDIAN to calculate vari- and not incorporated into subsequent

33
Burt matrix

20,000 40,000 60,000 80,000 100,000 120,000

Income (dollars)

bubble plot

calculations and analysis. Researchers of CATEGORICAL VARIABLES. Used in

may run a chain for several hundred (or
several thousand) steps during a burn-
in, as these initial points are not likely to
represent the target distribution.
B u r t m a t r i x a matrix showing all two-
way CROSS-TABULATIONS between a set
m u l t i p l e CORRESPONDENCE ANALYSIS, a
Burt matrix is created by arranging
the data in rows and columns, obtaining
the TRANSPOSE of this matrix, and mul-
tiplying the latter by the former. Also
called Burt table. [Ronald S. Burt
(1949- ), U.S. sociologist]

34
 Cc
C symbol for COEFFICIENT OF CONTIN- tends MULTIPLE REGRESSION, which re-
GENCY. lates several INDEPENDENT VARIABLES
with only one DEPENDENT VARIABLE. For
c a l i b r a t i o n n. the process of assigning
example, a researcher might use a ca-
values to a measuring device (instru-
nonical correlation analysis to investi-
ment, test, or scale) relative to a refer-
gate whether several measures of
ence standard. For example, it would be
attitude toward learning (e.g., confi-
useful to compare the scores on a new
dence, anxiety, interest, enjoyment) are
test of intelligence with those from an
related to several measures of achieve-
older, well-accepted test to ensure that
ment (e.g., grade point average, the
the new test scores provide comparable
number of math classes taken, the num-
ratings or values. To do so, a researcher
ber of memberships in honor societies).
might select a group of people (the cali-
bration sample), administer each indi- canonical correlation coefficient
vidual both the old and new tests, and (symbol: R^) an index of the magnitude
then assess the results. of the linear relationship between a lin-
ear combination of one set of variables
caliper m a t c h i n g see MATCHING.
and a linear combination of a different
canonical adj. in data analysis, describ- set of variables, with each linear combi-
ing procedures that utilize linear combi- nation made in such a way as to maxi-
nations of measured variables rather mize the strength of the relationship
than the actual individual variables. between the two variable sets. It ranges
in value from -1 to +1.
canonical analysis any of a class of
statistical procedures that assess the de- canonical discriminant analysis
gree of relationship between sets of vari- see DISCRIMINANT ANALYSIS.
ables via interpretation of a limited
number of linear combinations of spe- canonical discriminant function
cific values of those variables. The over- see DISCRIMINANT FUNCTION.
all objective of such procedures is to canonical l o a d i n g see DISCRIMINANT
reduce the DIMENSIONALITY of the data LOADING.
under investigation. Examples include
CANONICAL CORRELATION ANALYSIS, canonical variate the result of a
DISCRIMINANT ANALYSIS, MULTIPLE RE- weighted, linear combination of a set of
GRESSION, and MULTIVARIATE ANALYSIS variables, often used as a dependent or
OF VARIANCE, among others. independent variable in a CANONICAL
CORRELATION ANALYSIS.
canonical correlation analysis a
method of data analysis that provides a c a p i t a l i z a t i o n o n chance drawing a
measure of the strength of the relation- conclusion from data wholly or partly
ship between a linear combination of biased in a particular direction by
predictor variables and a linear combi- chance. A common example of capital-
nation of outcome measures. Essen- ization on chance is the presentation of
tially, it extends a basic CORRELATION all the significant results in a study with-
ANALYSIS, which relates only one vari- out considering the number of results
able with another variable, and it ex- examined.

35
capture-tag-recapture sampling

capture-tag-recapture s a m p l i n g a are shown to branch off from higher

method of SAMPLING that is used to esti- level subsets of data.
mate population size. For example, in
Cartesian coordinate system a sys-
order to estimate the number offishin a
tem for locating a point in an n-dimen-
lake, a random sample of fish (e.g., 100)
sional space on a graph by indicating the
would be drawn and tagged, then re-
distance of the point from a common or-
turned to the lake. The lake would be
igin along each of n axes. The most com-
resampled, and the results (i.e., the frac-
mon application of the Cartesian system
tion of tagged fish in the new sample)
is one based on two dimensions, repre-
would be used to estimate the total num-
sented by two perpendicular lines that
ber of fish in the lake. Also called cap-
intersect at a specific spot called the ori-
ture-recapture sampling; mark-
gin. In such a two-dimensional Carte-
and-recapture sampling.
sian coordinate system, the horizontal
line is often referred to as the x-AXis and
c a r d i n a l adj. describing a number that
the vertical line as the Y-AXIS. For exam-
indicates a count of a set of items, such
ple, the coordinate (3, 4) represents an x
as two dogs or three circles, but does not
value of 3 and a y value of 4 on the
indicate order within the set. A child's
graph; thus, from the origin one would
understanding of the concept of cardi-
count three units over on the x-axis and
nal numbers is an important area of
four units up on the y-axis to locate the
study in developmental psychology.
coordinate point, as shown in the fol-
Compare NOMINAL; ORDINAL.
lowing illustration.
carryover effect the effect on the cur-
rent performance of a research partici-
pant of the experimental conditions
that preceded the current conditions;
where such an effect is significant, it
may be difficult to determine the spe-
cific influence of the variable under
-A -3 -2
study. For example, in a CROSSOVER DE- H 1 1
SIGN in which a particular drug is ad-
ministered to animals in the first
experimental condition, a carryover ef-
fect would be evident if the drug contin-
ued to exert an influence on the
animals' performance during a subse-
quent experimental condition. Also
called holdover effect. See WASH-OUT
PERIOD.
[Rene Descartes (1596-1650), French
CART analysis classification and re- philosopher, mathematician, and scien-
gression tree analysis: a method of classi- tist]
fying data into successively smaller and
more homogeneous subgroups accord- Cartesian product a set of all possible
ing to a set of PREDICTOR VARIABLES. pairings of units in one set with units in
CART analysis is similar to but distinct another set. As an example, let A be a set
from REGRESSION ANALYSIS in its homo- of three colors (e.g., blue, green, yellow),
geneity of case subsets. The results of a and let B be a set of two shapes (e.g.,
CART analysis may be displayed in a dia- square, circle). The Cartesian product of
gram resembling a treesubsets of data, A X B is the set of all possible pairings:
based on values of the outcome variable. blue square, blue circle, green square,

36
 Cauchy-Schwarz inequality

green circle, yellow square, and yellow consists of counts or observations in

circle. That is, the Cartesian product of A specific categories rather than measure-
X B is 3 X 2 = 6. [Rene Descartes] ments. Categorical data that have a
meaningful order are referred to more
case n. a unit or observation to be ana- speciflcally as ORDINAL DATA, whereas
lyzed, such as a nonhuman animal, a categorical data without a meaningful
person, a group, an institution, an ob- order are known as NOMINAL DATA.
ject, or any other entity from which a re-
searcher gathers data. categorical data analysis any of sev-
eral statistical procedures used to model
case-control study a type of study in
variables that indicate counts or obser-
which a group of individuals diagnosed
vations in specific categories, often as a
with a disease or other condition is com-
function of one or more predictor vari-
pared to a group of individuals without
ables. For example, counts of survi-
the disease diagnosis, specifically with
val status (e.g., life vs. death) may be
regard to the proportion of people in
compared by treatment condition (e.g.,
each group who were exposed to a cer-
treatment vs. no treatment). Techniques
tain risk factor. Also called case-control
commonly used in categorical data
design; case-referent study.
analysis include the CHI-SQUARE TEST,
case study an in-depth investigation of LOGISTIC REGRESSION, and PROBIT
a single individual, family, event, or ANALYSIS.
other entity. Multiple types of data (psy-
categorical scale a sequence of num-
chological, physiological, biographical,
bers that identify items as belonging to
environmental) are assembled, for ex-
mutually exclusive categories. For exam-
ample, to understand an individual's
ple, a categorical scale for the political
background, relationships, and behav-
party affiliation of a group of Americans
ior. Although case studies allow for in-
might use 1 to denote Republican, 2 to
tensive analysis of an issue, they are
denote Democrat, and 3 to denote Inde-
limited in the extent to which their find-
pendent. When the number sequence
ings may be generalized.
has a meaningful order a categorical scale
casewise deletion see LISTWISE DELE- is more precisely called an ORDINAL SCALE;
TION. when it is devoid of such meaningful
order it is known as a NOMINAL SCALE.
catch t r i a l a trial within an experiment
in which a stimulus is not present but categorical variable a variable that is
the participants' responses nonetheless defined by a set of two or more catego-
are recorded. For example, in an experi- ries. Examples include a person's sex,
ment in which participants identify au- marital status, or rankings of particular
ditory signals, catch trials are those in stimuli (such as the relative loudness of
which no signal is given. The use of a different sounds). Specialized methods of
catch trial may help to estimate the level CATEGORICAL DATA ANALYSIS are avail-
at which a participant is guessing when able for assessing information obtained
no stimulus is present. from the measurement of categorical
variables.
categorical adj. referring to a character-
istic that is used to classify units, usually Cattell's scree test see SCREE PLOT.
individuals or experimental conditions, [Raymond Cattell (1905-1998), Brifish-
for the purposes of study. For example, born U.S. psychologist]
participants may be classified as males or
females in a particular study; in this case, Cauchy-Schwarz inequality a state-
sex is categorical. ment asserting that the ABSOLUTE VALUE
of the product of two numbers can equal
categorical data information that but not exceed the product of their indi-

37
causal analysis

vidual absolute values. For example,
consider the numbers -2 and +3, whose
product is -6, the absolute value of which
is +6; similarly, the pair of +2 and -3 also
yields a product of -6, which also has an
absolute value of +6. In either case, the
absolute values of the individual num-
bers are 2 and 3, which yields the same
product of 6. The Cauchy-Schwarz in-
equality is fundamental in many areas of
statistical theory, including probability
theory. [Augustin-Louis Cauchy (1789-
1857), French mathematician; Hermann
Amandus Schwarz (1843-1921), German
mathematician]
causal analysis an attempt to draw de-
pendable inferences about cause-and-
effect relationships from research data.
Encompassing a variety of methods (e.g.,
PATH ANALYSIS, STRUCTURAL EQUATION
MODELING), such analyses differ in the
degree to which they are statistically
complex and the degree to which causal
inferences from them are, in fact, justi-
fied.
causal d i a g r a m see PATH DIAGRAM.
evaluative tool, strict conditions con-
cerning the measurement of the vari-
ables must be met.
causal variable see INDEPENDENT
VARIABLE.
causation n. the empirical relation be-
tween two events, states, or variables
such that change in one (the cause)
brings about change in the other (the ef-
fect). causal adj.
CDF abbreviation for CUMULATIVE DIS-
TRIBUTION FUNCTION.
c e i l i n g effect a situation in which the
majority of values obtained for a vari-
able approach the upper limit of the
scale used in its measurement. For exam-
ple, a test whose items are too easy for
those taking it would show a ceiling ef-
fect because most people would achieve
or be close to the highest possible score.
In other words, the test scores would ex-
hibit SKEWNESS and have little VARI-
ANCE, thus prohibiting meaningful
analysis of the results. Compare FLOOR
EFFECT.

causal hypothesis a HYPOTHESIS about cell n. a combination of two or more

how change in one variable leads to a characteristics represented by the inter-
subsequent change in another variable. section of a row and a column in a statis-
A causal hypothesis may be used, for in- tical table. A tabular display resulting
stance, to formally state how the appli- from a study of handedness in men and
cation of a treatment will subsequently women, for instance, might consist of
affect behavior or performance. four cells: left-handed females, left-
handed males, right-handed females,
causal indicator in STRUCTURAL EQUA- and right-handed males. See also EMPTY
TION MODELING, a measure used to rep-
CELL.
resent a factor that is assumed to be the
cause of a dependent LATENT VARIABLE. cell frequency a count of observations
Compare EFFECT INDICATOR. that have a specific combination of two
or more characteristics as displayed in a
causal inference the reasoned process
statistical table. For example, results
of concluding that change in one vari-
from a survey in which respondents are
able produced change in another vari-
asked to report their transportation
able. See CAUSAL ANALYSIS. mode and sex may be described by a
causal m o d e l i n g any procedure used two-way CONTINGENCY TABLE, in which
to test for cause-and-effect relationships cell frequencies provide the number of
(as opposed to mere correlation) between people who fall into the different possi-
multiple variables, STRUCTURAL EQUA- ble category combinations (e.g., females
TION MODELING is the best known ex- who bike, males who bike, females who
ample. For a causal model to be a useful drive, males who drive, females who

38
 central limit theorem

take public transit, males who take pub- portion of the scale. A response mea-
lic transit). sured by a meter, for instance, may not
be recorded because it is too small to be
cell mean a mathematical average of
detected or is so large that it exceeds the
the numeric values associated with a meter's capability to record. In such a
specific combination of two or more situation, the minimum value is as-
characteristics as displayed in a statisti- signed to an undetectable observation,
cal table. For example, results from a and the maximum value is assigned to
study of response time under specific an excessive observafion.
combinations of several factors may rely
on a comparison of cell means to GROUP censored regression see TOBIT ANALY-
DIFFERENCES in rcsponse times between SIS.
experimental conditions.
censoring n. the situafion in which some
cell-means model an approach in observafions are missingfroma set of data
which a set of observed responses on a (see CENSORED DATA). Censoring is com-
DEPENDENT VARIABLE are analyzed at mon in studies of survival time, in which
each level or condition of an INDEPEN- the research often ends before the event
DENT VARIABLE in order to identify dif- of interest occurs for all study units. See
ferences in their average values. Each also LEFT CENSORING; RIGHT CENSORING.
such average is obtained by adding a
random error term to the typical re- census n. the complete count of an en-
sponse value for the larger population tire population. A census differs from
from which the observations derive, as most experimental studies, which use a
estimated from the sample data. For ex- SAMPLE from a population in hopes of
ample, a researcher examining the influ- generalizing from that observed subset
ence of sleep on exam performance to the larger group.
might use the cell-means model to deter- centile a shortened name for PERCENTILE.
mine whether there are any differences
in the average exam score obtained by centile reference chart a table used
participants who got 6 hours of sleep the to compare a measurement on an indi-
previous night, those who got 8 hours, vidual to values in a population. The
PERCENTILE value from the population
and those who got 10 hours.
may be used to judge whether the indi-
censored data a set of data in which vidual's score is atypical in that context.
some values are unknown because they Thus, a centile reference chart may be
are not observed or because they fall used in assessing a child's growth rela-
below the minimum or above the maxi- tive to her weight and height (see chart
mum value that can be measured by the overleaf from the U.S. Centers for Dis-
scale used. For example, in a study of ease Control).
survival rates of a group of people with a
central d i s t r i b u t i o n see NONCEN-
particular disease, censored data would
TRAL DISTRIBUTION.
be obtained if the deaths of some partici-
pants occurred after the study ended. See central l i m i t theorem (CLT) the sta-
also DOUBLY CENSORED DATA. tistical principle that the sum of inde-
pendent values from any distribution
censored linear regression model
will approach a NORMAL DISTRIBUTION
see ACCELERATED FAILURE TIME MODEL.
as the number of values in the distribu-
censored observation a score that is tion increases. In other words, the larger
missing from a data set, either because the sample size, the more closely the
the event of interest has not occurred by SAMPLING DISTRIBUTION approximates a
the end of the study period or because normal distribution. The central limit
the response falls into an unmeasurable theorem is used to justify certain data

39
central moment

Weight-for-stature percentiles: Girls

33 34 35 36 37 38 39 40 41 42 43 44 45 46 47

centile reference chart

analysis methods when the appropriate- point of a set of scores. The central ten-
ness of a method relies on an assump- dency of a sample data set, for instance,
tion of normality. may be estimated by a number of differ-
ent statisfics (e.g., MEAN, MEDIAN,
central moment a MOMENT that de- MODE). See also MEASURE OF CENTRAL
scribes the shape of a set of scores with TENDENCY.
regard to its deviation about the mean.
Four common central moments that de- central-tendency bias see END-
scribe a RANDOM VARIABLE are the MEAN AVERSION BIAS.
(the first central moment), the VARI-
ANCE (the second central moment), the central-tendency measure see MEA-
SKEWNESS (the third central moment), SURE OF CENTRAL TENDENCY.
and the KURTOSIS (the fourth central
moment). Also called moment about centroid n. 1. in geometry, the center
the mean. point of an object or area in multidi-
mensional space. It may be determined
central tendency the middle or center by calculating the intersection of all
40
 change-point problem

straight lines or planes that divide the c h a i n graph a graph in which the di-
area into two equal parts. 2. the MEAN of rectional and nondirectional relation-
multivariate data. It may be obtained by ships among variables in a data set are
calculating the average of the VECTORS indicated by arrows and lines, respec-
for the set of variables. See also CEN- tively. Consider the following generic
TROID METHOD. example.
centroid f a c t o r a single variable iden-
tified through its CENTROID value as
summarizing responses on multiple
other variables. Centroid factors are ob-
tained by the CENTROID METHOD, a
data-reduction technique sometimes
used in FACTOR ANALYSIS.
chain-of-events data data regarding a
centroid method 1. a technique used series of events that have occurred to a
in FACTOR ANALYSIS to reduce multiple group of individuals in the same order
correlated variables to a smaller set of ex- but not necessarily at the same times.
planatory variables. It is similar toand
indeed was a precursor ofPRINCIPAL c h a i n p a t h model a rare synonym of
COMPONENTS ANALYSIS but involves MARKOV C H A I N .
simpler calculations. Generally, the cen- chance difference a difference be-
troid method is now used only in situa- tween two samples that arises from the
tions when computers are not available. nature of sampling itself, rather than
2. an approach to AGGLOMERATIVE CLUS- from any intervention by a researcher.
TERING in which CENTROIDS are calcu- For example, scores obtained from two
lated to determine similarity among sep- groups of research participants might
arate entities and combine them into vary because of personal factors affect-
higher order groups. Each individual is ing some individuals (e.g., financial wor-
assigned to the subset closest in centroid ries) and not because of the conditions
value, with the process continuing until of the study.
the maximum grouping of cases has been
obtained. chance v a r i a t i o n see RANDOM VARIA-
TION.
CERES p l o t combined conditional ex-
pectations and residuals plot: a graph change i n R^ a change in the COEFFI-
that is similar to a PARTIAL RESIDUAL CIENT OF MULTIPLE DETERMINATION (R^)
PLOT in displaying relationships among from one model to another. In proce-
the predictors of a DEPENDENT VARIABLE dures such as HIERARCHICAL REGRES-
but distinct in being appropriate for use SION and STEPWISE REGRESSION, in
with nonlinear entities. which predictor variables are individu-
ally added to or removed from a set, a
CF abbreviation for CUMULATIVE FRE- significant change in R^ indicates that
QUENCY. the predictor of interest has a meaning-
CFA abbreviation for CONFIRMATORY ful role in the explanation of the re-
FACTOR ANALYSIS. sponse variable.
CFI abbreviation for Bentier comparative change-point p r o b l e m in the obser-
FIT INDEX. vafion of TIME-SERIES data, the difficulty
in judging when exactly a meaningful
CGF abbreviation for CUMULANT GENER- change in the pattern of data occurs,
ATING FUNCTION.
given a certain degree of indeterminacy
CHAID abbreviation for CHI-SQUARE AU- or randomness in the behavior of the
TOMATIC INTERACTION DETECTOR. variable (see STOCHASTIC PROCESS).

41
change score

change score see DIFFERENCE SCORE. pletely unknown except for mean and
changing-criterion design an exper- variance. [Pafnuty Chebyshev (or
Tchebyshev; 1821-1894), Russian math-
imental approach in which an initial
ematician]
baseline phase is followed by implemen-
tation of a treatment delivered in a series checklist n. a list of items that are to be
of phases. Each phase has its own crite- observed, recorded, corrected, or other-
rion rate for the target behavior, and wise considered in some manner. See
once responding becomes stable an in- also BEHAVIOR CHECKLIST. .
cremental shift in the criterion occurs to C h e r n o f f faces a representation of data
implement another phase. For example, in the form of stylized faces, designed to
a researcher studying the effectiveness of take advantage of the ability of observers
money in reducing caffeine consump- to discern subtle changes in facial ex-
tion may use a changing-criterion de- pressions.
sign with four treatment phases, each
gradually decreasing the amount below
which participants must maintain their
caffeine intake. The changing-criterion
design is distinguished from other ap-
proaches (e.g., the A-B-A DESIGN) in that In the example above, features such as
it does not require the withdrawal of an head and eye eccentricity, eyebrow slant,
intervention, an extended baseline, or and mouth size represent different values
treatment implementation across multi- of the variables of interest. [Herman
ple behaviors or treatments. Chernoff, 20th-century U.S. statistician]

characteristic n. a PARAMETER that de- chi-square (symbol: x^) n. a statistic that

scribes a population distribution, such is the sum of the squared differences be-
as its mean or STANDARD DEVIATION. tween the observed scores in a data set
and the EXPECTED VALUE. That is, to ob-
characteristic f u n c t i o n a formula tain the chi-square one finds the differ-
that defines the PROBABILITY DISTRIBU- ence between each observed score and
TION of values on a random variable. the expected score, squares that differ-
characteristic root see EIGENVALUE. ence, and divides by the expected score;
finally one adds the resulting values for
characteristic value see EIGENVALUE.
each score in the set. The smaller the chi-
characteristic vector see EIGEN- square, the more likely it is that the
VECTOR. model from which the expected score is
chart n. a graphic or tabular display of obtained provides a legitimate represen-
data. See also GRAPH; TABLE. tation of the phenomenon being mea-
sured. See CHI-SQUARE DISTRIBUTION;
Chebyshev's i n e q u a l i t y a theorem CHI-SQUARE TEST.
stating that in any sample or distribu-
tion the great majority of values are chi-square automatic interaction
close to the mean; the probability that a detector (CHAID) an EXPLORATORY
random variable will take on a value DATA ANALYSIS technique in which the
within k STANDARD DEVIATIONS of the relationships between a response vari-
mean is at least 1 - l/k^. In simpler able and a set of potential predictor vari-
terms, this implies that at least 75% of ables are examined to identify inter-
the measurements of any distribution actions or patterns. In this process, val-
will fall in the range from -2 to +2 stan- ues of the response variable are split into
dard deviations from the mean. The the- successively smaller subsets according to
orem has great utility because it can be values of the predictor variables, such
applied to distributions that are com- that the resulting subsets are relatively

homogeneous with regard to the re-
sponse variable. The response and the
predictor variables may be CATEGORICAL,
continuous, or any combination thereof.
c h i - s q u a r e d i s t a n c e a measure of the
difference between two or more groups
with regard to their respective average
values on a set of variables. It is similar to
EUCLIDEAN DISTANCE in purpose but
used with C A T E G O R I C A L D A T A .
c h i - s q u a r e test f o r h o m o g e n e i t y a
chi-square d i s t r i b u t i o n (^ flistri-
b u t i o n ) a distribution of the sums of in-
dependent squared differences between
the observed scores in a data set and the
expected score for the set. If a random
sample is repeatedly drawn from a nor-
mal population and measured on some
variable and the obtained scores trans-
formed via STANDARDIZATION, multi-
plied by themselves, and then added,
the result will be a chi-square distribu-
tion with DEGREES OF FREEDOM equal tO c h i - s q u a r e test f o r i n d e p e n d e n c e a
the size of the samples drawn. See CHI-
SQUARE TEST.
c h i - s q u a r e g o o d n e s s - o f - f i t test a
statistical method of assessing how well
a mathematical model or theoretical ex-
pectation fits a set of observed data. The
test is used to evaluate hypotheses about
the proportions of individuals within
different categories of a given N O M I N A L
variable, such as whether there are dif-
ferent percentages of males and females
enrolled in various college majors. The
smaller the statistic obtained from this
test, the smaller the difference between
the proposed model and the obtained
data and thus the better the model fit.
c h i - s q u a r e p r o b a b i l i t y p l o t a graph
of the frequency with which particular
CHI-SQUARE values are obtained in a
c h i - s q u a r e test f o r t r e n d a statistical
data set as compared to the frequency
expected according to theory. See PROB-
ABILITY PLOT.
c h i - s q u a r e test any of various proce-
dures that use a CHI-SQUARE DISTRIBU-
TION to evaluate whether there is a
relationship between variables or the de-
gree to which a theory fits a set of ob-
43
chi-square test for trend
served data. Examples include the C H I -
SQUARE GOODNESS-OF-FIT TEST, the CHI-
SQUARE TEST FOR HOMOGENEITY, the
CHI-SQUARE TEST FOR INDEPENDENCE,
the CHI-SQUARE TEST FOR TREND, and
the CHI-SQUARE TEST FOR VARIANCE.
When unqualified, however, this term
usually refers to the test for independ-
ence. Also called chi-square proce-
dure.
statistical method to evaluate an as-
sumption that some feature of interest is
equal between two or more groups (see
HOMOGENEITY). The test is used to eval-
uate whether the frequencies within two
or more categories of a given N O M I N A L
variable are equivalent in different pop-
ulations, such as whether the same
number of fifth graders and sixth graders
watch particular cartoons.
statistical method to evaluate whether
there is a relationship between two vari-
ables whose values are categories. For ex-
ample, it may be used to test whether sex
(male vs. female) is independent of
that is, not related to or associated
withhaving a household pet (yes vs.
no). It compares observed data to the
data that would be expected in each cell
of a CONTINGENCY TABLE if the two vari-
ables were entirely independent. This
discrepancy between observed and ex-
pected counts is then used to compute
the CHI-SQUARE statistic. Overall, larger
values of the statistic relative to the
number of participants in the study (de-
grees of freedom) are more likely to be
statistically significant. Also called chi-
square test of association.
method to evaluate change over time in
a CATEGORICAL VARIABLE Or to deter-
mine whether two or more trends in cat-
egorical data differ from each other. For
example, it may be used to test whether
there is a difference in incidences of can-
cer severity (low, medium, or high) as
people age.
chi-square test for variance

chi-square test f o r variance a statis- ear equations. [Andre-Louis Cholesky

tical method to evaluate a hypothesis (1875-1918), French mathematician]
about the DISPERSION in a set of scores. C h o w test a statistical procedure used to
For example, it may be used to evaluate evaluate the equality of the coefficients
whether the STANDARD DEVIATION in re- in two separate REGRESSION EQUATIONS
cent test scores for a classroom of stu- developed to calculate a response variable
dents is the same as that for the school as from one or more predictor variables.
a whole. [Gregory Chow (1929- ), Chinese-born
chi-square test of association see U.S. economist]
CHI-SQUARE TEST FOR INDEPENDENCE. chronograph n. an instrument that
choice reaction a response that is de- records time sequences graphically.
pendent on the detection of a stimulus chronographic adj.
or the differentiation of its different lev-
chronometer n. a precise clock that
els. For example, study participants
runs continuously and is designed to
might be measured as to how long it
maintain its accuracy under all condi-
takes them to push a certain computer
tions of temperature, pressure, and the
key after seeing a picture appear on the
like.
monitor.
CI abbreviation for CONFIDENCE INTER-
Cholesky factorization the re-expres-
VAL.
sion of a SQUARE MATRIX as the product
of two related matrices. More specifi- circumplex n. a circular depiction of
cally, it is the re-expression of a POSI- the similarities among multiple vari-
TIVE-DEFINITE symmetric matrix as the ables. For example, a researcher studying
product of a lower TRIANGULAR MATRIX emotions might focus on certain core af-
and its TRANSPOSE. The process by which fects and create a circumplex representa-
this is accomplished is called Cholesky tion of them, with variables having
decomposition and is one of several nu- opposite values or characteristics (i.e.,
merically stable methods of solving lin- tranquility-frenzy; sadness-enthusiasm;

Xll
Activation
Unpleasant Deactivation (0-g-. aroused, activated) Pleasant Activation
(e.g-. frenzied, jittery) (e.g.. energetic, excited)

Activated Displeasure Activated Pleasure

(e.g-, distressed, upset) (e.g-, entijusiastic, elated)

III
Pleasure
(e.g-, un/iappy, dissatisried) (e-g.. satisfied, pleased)

Vlll
Deactivated Displeasure Deactivated Pleasure
(e.g., sad, gloomy) (e.g-, serene, peaceful)

Unpleasant Deactivation Pleasant Deactivation

(e.g-, sluggish, tired) y| (e.g., placid, tranquil)
Deactivation
(e-g., quiet, still)

circumplex
 classification table

sluggishness-excitement) displayed at ITY. It is often contrasted with BAYESIAN

opposite points on the circumplex. methods.
c i t a t i o n analysis a form of research classical test theory (CTT) the theory
that examines the history, frequency, that an observed score (e.g., a test result)
and distribution of citations of particu- that is held to represent an underlying
lar writers or particular books, articles, or attribute (e.g., intelligence) may be di-
other sources. It now mostly involves vided into two quantitiesthe true
the automated search of online or elec- value of the underlying attribute and the
tronic databases. error inherent to the process of obtain-
ing the observed score. CTT may be rep-
city-block distance an index of the
resented mathematically as Xp = Tp + e^,
degree of similarity between sets of mea-
where Xp is the observed score for re-
surements on different variables: It is
spondent p, Tp is the respondent's TRUE
calculated as the sum of the ABSOLUTE
SCORE for the construct or characteristic
DIFFERENCES between two values of one
being measured, and Cp is RANDOM ERROR
variable and the corresponding values of
that dilutes the expression of respon-
a second variable. City-block distance
dent p's true score. The theory serves as
conceptualizes entities as being arranged
the basis for models of test RELIABILITY
in a rectangular grid, such that calcula-
and assumes that individuals possess sta-
tions of distance resemble the approach
ble characteristics or traits that persist
used by a person when traversing a
through time. See also GENERALIZABILITY
citywalking horizontally down some
THEORY. Compare ITEM RESPONSE THE-
streets and vertically up others until ar-
ORY.
riving at his or her destination. It gener-
ally is insensitive td extreme data points classification n. the process of group-
(OUTLIERS) and contrasts with the linear ing individuals or units into categories
EUCLIDEAN DISTANCE approach. Also based on one or more characteristics. It
called city-block metric; Manhat- usually involves a qualitative descrip-
tan distance. tion of the unit, in which the categories
represent types or kinds and not fre-
class n. a group, category, or division. quency, amount, or degree.
classical inference see FREQUENTIST
classification a n d regression tree
INFERENCE.
analysis see CART ANALYSIS.
classical p r o b a b i l i t y an approach to
classification rule a mathematical
the understanding of PROBABILITY based
formula for placing units into groups
on the assumptions that any random
(e.g., within DISCRIMINANT ANALYSIS) so
process has a given set of possible out-
as to maximize the similarities among
comes and that each possible outcome is
group members with regard to a set of
equally likely to occur. An example
characteristics. Also called allocation
often used is rolling a die, in which there
rule.
are six possible outcomes and each out-
come is assumed to be equally likely. classification table a table showing
the number or percentage of cases cor-
classical scaling traditional MULTIDI-
rectly categorized by a given model, rule,
MENSIONAL SCALING, in which dissimi-
or algorithm derived for that purpose.
larities among stimuli are identified via
Columns represent the model predic-
distances between data points. See also
tions and rows represent the actual
NONMETRIC SCALING. classes to which the items in a sample
classical statistics a traditional ap- data set belong. For example, a re-
proach to statistics based on FREQUENT- searcher might use a classification table
IST INFERENCE and CLASSICAL PROBABIL- similar to the hypothetical overleaf to

45
classification tree

Predicted attendance category

Actual attendance n 2-year college 4-year college No college

2-year college 21 16 (76.2%) 2 (9.5%) 3 (14.3%)

4-year college 45 2 (4.4%) 34 (75.6%) 9 (20.0%)

No college 18 4 (22.2%) 7 (38,9%) 7 (38.9%)

classification table

display the results of a DISCRIMINANT gain versus no gain in psychotherapy for

ANALYSIS of college attendance for a a variety of disorders. Ellipses are nodes,
group of high school students so as to arrows show branches, and squares illus-
evaluate the general accuracy of the trate prediction end points. Numbers
model used. Also called classification within the squares give the number of
matrix; confusion matrix. correct predictions divided by the total
number of observations classified at that
classification tree a diagram display-
end point. See also CART ANALYSIS.
ing a set of rules for creating successively
smaller and increasingly similar sub- class i n t e r v a l a range of scores or nu-
groups or categories of items from a merical values that constitute one seg-
larger original group. Each node of the ment or class of a variable of interest. For
classification tree divides into two example, individual weights can be
branches with additional nodes, with placed into class intervals such as 100-
the process continuing until there are 120 lb, 121-140 lb, 141-160 lb, and so
too few observations in a node to divide forth. Class intervals often are used in
them further or until a node is consid- FREQUENCY DISTRIBUTIONS and HISTO-
ered homogeneous. For example, con- GRAMS to present a large data set in a
sider the classification tree (below) for simpler manner that.is more easily inter-
discriminating children achieving early preted.

No assistance Gets assistance

Predict
Medication early gain
consult
38/64 = 59.4%
No I I , , \ I Yes
p=.043 I

Predict no Predict early

early gain gain

18/34 = 52.9% 22/30 = 73.3%

classification tree

46
 clustered data

class l i m i t the uppermost and lower- cases involving specific procedures for
most values between which lie a range of diagnosis or treatment.
values constituting a segment or class of
a variable: the upper and lower bound- closed i n t e r v a l a range of values that
aries of a CLASS INTERVAL. includes both its endpoints. For exam-
ple, the closed interval [0,3] consists of
class size the width of a range of values all values not smaller than 0 and not
constituting a segment or class of a vari- larger than 3. Compare OPEN INTERVAL.
able: the width of a CLASS INTERVAL.
closed question see FIXED-ALTERNA-
c l i n i c a l adj. in medical diagnosis or TIVE QUESTION.
treatment, describing a decision based
on a clinician's personal opinion and CLT abbreviation for CENTRAL LIMIT THE-
OREM.
judgment, as opposed to one based on
statistical evidence. Compare ACTUAR- cluster analysis a method of
IAL. multivariate data analysis in which indi-
c l i n i c a l sample a sample made up of viduals or units are placed into distinct
individuals who have been given a for- subgroups based on their strong similar-
mal diagnosis of a disorder. ity with regard to specific attributes. For
example, one might use cluster analysis
c l i n i c a l significance the extent to to form groups of individual children on
which a study result is judged to be the basis of their levels of anxiety, ag-
meaningful in relation to the diagnosis gression, delinquency, and cognitive
or treatment of disorders. An example of difficulties so as to identify useful
a clinically significant result would be an typologies that could increase under-
outcome indicating that a new interven- standing of co-occurring mental disor-
tion strategy is effective in reducing ders and lead to more appropriate
symptoms of depression. See also PRAC- treatments for specific individuals.
TICAL SIGNIFICANCE; STATISTICAL SIG- There are several different forms of clus-
NIFICANCE. ter analysisincluding HIERARCHICAL
CLUSTERING, K-MEANS CLUSTERING, and
c l i n i c a l test a test or measurement
LATENT CLASS ANALYSISand each is ap-
made in a clinical or research context for
the purpose of diagnosis or treatment of
propriate for use with different types of
a disorder. data. Results of a cluster analysis often
are presented in a DENDROGRAM. Also
c l i n i c a l t r i a l a research study designed called clustering.
to compare a new treatment or drug
with an existing standard of care or cluster centroid the central or typical
other control condition (see CONTROL value in a set of values relating to multi-
GROUP). Trials generally are designed to
ple variables. For example, in a data set
answer scientific questions and to find that contains patient ages and sex, the
better ways to treat individuals who cluster centroid would be an individual
have a specific disease or disorder. Also of the average age and the most com-
called clinical study; therapeutic mon sex for that group. See also CEN-
trial. See also RANDOMIZED CLINICAL TROID METHOD.
TRIAL. clustered data a set of observations or
c l i n i c a l u t i l i t y research see EFFEC- scores that can be grouped into multiple
TIVENESS RESEARCH.
subsets (clusters), such that the items in
each subset are similar to one another
c l i n i c a l v a l i d a t i o n the process of ac- with respect to certain attributes and the
quiring evidence to support the accu- distinctions between subsets help ex-
racy of a theory by studying multiple plain the overall variation among the
clustering

values as a whole. See CLUSTER ANALY- have been rounded to the nearest
SIS. integer or survival times that are missing
clustering n. see CLUSTER ANALYSIS.
observations for some units.

cluster r a n d o m i z a t i o n a process in Cochran C test a statistical procedure

which preexisting groups are assigned for determining whether HOMOGENEITY
randomly to specific experimental condi- OF VARIANCE exists amoug data ob-
tions, as occurs in GROUP EXPERIMENTS,
tained from two or more different
GROUP-RANDOMizED TRIALS, and Other groups. It is a less popular alternative to
designs in which collective entities form the LEVENE TEST FOR EQUALITY OF VARI-
the UNIT OF ANALYSIS. When such a pro- ANCE and the BARTLETT TEST FOR EQUAL-
cess is applied to single individuals it is ITY OF VARIANCE. [William Gemmell
called RANDOM ASSIGNMENT. Cochran (1909-1980), Scotfish-born U.S.
statistician]
cluster-randomized t r i a l see GROUP-
RANDOMIZED TRIAL.
Cochran-Mantel-Haenszel test see
MANTEL-HAENSZEL TEST.
cluster s a m p l i n g a tiered method of
obtaining units for a study. A popula- Cochran Q test a NONPARAMETRIC sta-
tion is first subdivided into smaller tistical procedure applied when each ex-
groups or clusters (often administrative perimental unit (e.g., an individual
or geographical), and a random sample participant) is observed under multiple
of these clusters is drawn. The process is conditions and the data can have only
then repeated for each sampled cluster one of two outcomes (e.g., yes or no,
until the required level is reached. An pass or fail). The Q test is used to evalu-
example would be sampling voters in a ate the equality of the outcomes under
large jurisdiction (e.g., a state) by ran- the different conditions. [William Gem-
domly choosing subgroups (e.g., coun- mell Cochran]
ties) and then further subgroups (e.g., codebook n. a source of information re-
towns and cities) until individual partic- lating to a set of variables used in a
ipants are obtained for a study. See also study, such as the names given to vari-
ADAPTIVE CLUSTER SAMPLING. ables, how they were defined, and their
cluster seeds a procedure for dividing possible response options.
data into subgroups (clusters). A seed c o d i n g n. the replacement of data values
point is identified for each cluster and all with labels or other markers (i.e., a code)
objects or individuals within a pre- so as to conceal the identities of partici-
scribed distance of that point are in- pants or facilitate subsequent analysis.
cluded in that cluster. The number of coefficient n. a quantity or value that
seed points can be established at the be- serves as a measure of some property. For
ginning of the process and the clustering example, the CORRELATION COEFFICIENT
repeated several times with different is a measure of linear relatedness and a
starting seed points. REGRESSION COEFFICIENT is a measure of
cluster-specific model a representa- the relationship between a predictor and
tion of a variable or the relationships a response variable.
among a set of variables that is specific coefficient a l p h a see CRONBACH'S
to a subset (cluster) of the data obtained ALPHA.
for analysis.
coefficient o f agreement a numeri-
Co abbreviafion for COMPARISON STIMU- cal index that reflects the degree of
LUS. agreement among a set of raters, judges,
coarse data data that are imprecise in or instruments as to which of several cat-
some way, such as reaction times that egories a case belongs. Coefficients of
48
 coefficient of multiple determination

agreement, such as COHEN'S KAPPA, are Obtained by multiplying the value of

often corrected for chance agreement. the CORRELATION COEFFICIENT (r) by it-
Also called agreement coefficient. self, the coefficient of determination
ranges in value from 0 to 1. Low values
coefficient of alienation (symbol: k)
indicate the outcome is relatively unre-
a numerical index that reflects the
lated to the predictor, whereas values
amount of unexplained variance be-
closer to 1 indicate that the two vari-
tween two variables. It is a measure of
ables are highly related. For example, if
the lack of relationship between the two
r = .30, then the squared correlation
variables. Also called alienation coef-
coefficient is .30^ = .09 and interpreted
ficient.
to mean 9% of the variance between
coefficient of association any of var- the two variables is common or overlap-
ious quantities that index the relation- ping. Also called determination co-
ship between two variables. For example, efficient; squared correlation co-
the Pearson PRODUCT-MOMENT CORRE- efficient. See also COEFFICIENT OF MUL-
LATION COEFFICIENT might be used to TIPLE DETERMINATION.
measure a relationship involving linear coefficient o f dispersion a measure
associafions, whereas the SPEARMAN COR- of the spread in a set of scores. It is calcu-
RELATION COEFFICIENT might be used
lated as a ratio of the average distance
for one involving nonlinear associations. that a score in the set lies from the score
coefficient o f concentration see at the middle of the set, relative to that
GINI COEFFICIENT. middle score. In other words, it is a ratio
of the mean deviation about the MEDIAN
coefficient of concordance (sym- relative to the median.
bol: W) a numerical index that reflects
the degree to which the rankings of k coefficient o f equivalence a mea-
conditions or objects by m raters are in sure of RELIABILITY for parallel test
agreement. Its value ranges from 0 (no forms: It quantifies the degree of associa-
agreement) to 1 (perfect agreement). tion between two tests that assess the
Also called Kendall's coefficient of same material using different items. The
concordance; Kendall's W. two tests are given to the same group of
respondents and the correlation be-
coefficient of contingency (symbol: tween scores on each version is calcu-
C) a measure of the degree of association lated.
between two variables whose values are
unordered categories. Its value ranges coefficient of m u l t i p l e determi-
ftom 0 to 1, with 0 indicating that the n a t i o n (symbol: 7?^) a numerical index
two NOMINAL VARIABLES are indepen- that reflects the degree to which varia-
dent of each other. Examples include tion in a response or outcome variable
CRAMER'S V and YULE'S Q. Also called (e.g., workers' incomes) is accounted for
contingency coefficient. by its relationship with two or more pre-
dictor variables (e.g., age, gender, years
coefficient o f determination (sym- of education). More specifically, it is a
bol: r^) a numerical index that reflects measure of the percentage of variance in
the proportion of variation in an out- a DEPENDENT VARIABLE that is ac-
come or response variable that is ac- counted for by its relationship with a
counted for by its relationship with a weighted linear combination of a set of
predictor variable. More specifically, it is INDEPENDENT VARIABLES. Obtained by
a measure of the percentage of variance multiplying the value of the MULTIPLE
in a DEPENDENT VARIABLE that is ac- CORRELATION COEFFICIENT (R) hy itself,
counted for by its linear relationship the coefficient of multiple determina-
with a single INDEPENDENT VARIABLE. tion ranges in value from 0 to 1. Low

49
coefficient of reliability

values indicate that the outcome is rela- Cohen's k a p p a (symbol: K) a numerical

tively unrelated to the predictors, index that reflects the degree of agree-
whereas values closer to 1 indicate that ment between two raters or rating sys-
the outcome and the predictors are tems classifying data into mutually
highly related. For example, if R = .40, exclusive categories, corrected for the
then the coefficient of multiple determi- level of agreement expected by chance
nation is .40^ = .16 and interpreted to alone. Values range from 0 (no agree-
mean 16% of the variance in outcome is ment) to 1 (perfect agreement), with
explainable by the set of predictors. Also kappas below .40 generally considered
called multiple correlation coeffi- poor, .40 to .75 considered fair to good,
cient squared; squared multiple and more than .75 considered excellent.
correlation coefficient. See also CO- In accounting for chance, Cohen's
EFFICIENT OF DETERMINATION. kappa avoids overestimating the true
level of agreement as might occur
coefficient o f r e l i a b i l i t y see RELI- through simply determining the num-
ABILITY COEFFICIENT.
ber of times that two raters agree relative
coefficient o f r e p r o d u c i b i l i t y a to the total number of ratings. [Jacob
measure of the extent to which a set of Cohen]
observed responses on a GUTTMAN
cohort n. a group of individuals who
SCALE match the pattern of responses
share a similar characteristic or experi-
expected according to a theory. Its value
ence. The term usually refers to an age
ranges from 0 to 1, with larger values in-
cohort or birth cohort, that is, a group of
dicating greater GOODNESS OF FIT.
individuals who are born in the same
coefficient o f stability see STABILITY year and thus of similar age.
COEFFICIENT.
cohort analysis a statistical procedure
coefficient of variation a measure of to assess the effects attributed to mem-
variability in a set or distribution of val-
bers of a group sharing a particular char-
ues. It is determined by dividing the dis-
acteristic, experience, or event. In other
tribution's STANDARD DEVIATION by its words, it is the evaluation of data ob-
MEAN. Also called variation coeffi- tained from studies involving COHORTS,
cient. as may be the case in LONGITUDINAL DE-
SIGNS, for example.
Cohen's d a measure of EFFECT SIZE
based on the standardized difference be- cohort effect any outcome associated
tween two means: It indicates the num- with being a member of a group possess-
ber of STANDARD DEVIATION units by ing a common characteristic (such as in-
which the means of two data sets differ. dividuals who are born in the same year)
For example, a mentoring intervention and therefore influenced by common
associated with a Cohen's d of +0.25 in- historical events and practices. Cohort
dicates an increase of 0.25 standard devi- effects may be difficult to separate from
ation units for the average child who AGE EFFECTS and PERIOD EFFECTS in re-
received mentoring relative to the aver- search.
age child who did not receive mentor-
ing. The metric is used to represent effect cohort s a m p l i n g a method of sam-
sizes in META-ANALYSIS as well as in the pling data in which one or more groups
determination of POWER, with values of sharing a similar characteristic, such as
0.20, 0.50, and 0.80 representing small, year of birth, are identified and ob-
medium, and large effect sizes, respec- served.
tively. See also GLASS'S D. [Jacob Cohen cohort-sequential design an experi-
(1923-1998), U.S. psychologist and stat- mental design in which multiple mea-
istician] sures are taken over a period of time
 combination

from two or more groups of different get taller). Collinearity leads to diffi-
ages (COHORTS). If, for instance, individ- culties in interpreting the unique influ-
uals ranging in age from 5 to 10 years are ences of the independent variables and
sampled and then the members of each requires the use of PARTIALING proce-
age group are studied for a 5-year period, dures to distinguish their separate ef-
the resulting data would span 15 years of fects. See also MULTICOLLINEARITY.
development. Such studies essentially
are a combination of a LONGITUDINAL c o l u m n m a r g i n a l a summary of the
DESIGN and a CROSS-SECTIONAL DESIGN. values across each vertical set of cells in a
Also called accelerated longitudinal table. For a table containing frequency
design. counts it is a sum of the number of
counts in each column, whereas for
cohort study see LONGITUDINAL DE- other types of tables it is the average
SIGN. value of all data observations within a
given column. Compare ROW MAR-
cold-deck i m p u t a t i o n one of several
GINAL.
methods of inserting values for missing
data (see IMPUTATION) in which missing c o l u m n sum of squares 1. in an
observations are replaced by values from ANALYSIS OF VARIANCE involving two in-
a source unrelated to the data set under dependent variables, the amount of vari-
consideration. Suppose, for example, ance among individuals that is
that a patient questionnaire was admin- associated with either variable, as de-
istered in a hospital and that five people rived from the values given in the rele-
failed to respond to an item. Substi- vant column of the data table. It is
tuting responses from a similar item on a obtained by determining the average of
survey conducted previously would be all observations in the column, calculat-
an example of cold-deck imputation. ing how much each score deviates from
Compare HOT-DECK IMPUTATION. that average, multiplying the resulting
collaborative evaluation see PAR- value by itself, and adding it to the simi-
TICIPATORY EVALUATION.
larly obtained values for all other indi-
viduals in the column. The calculated
collapsing n. the process of combining quantities for each column are then
multiple response options or categories used to compute the BETWEEN-GROUPS
to form a smaller number of responses or SUM OF SQUARES, which in turn is used
categories. For example, a researcher to compute an F RATIO. See also TOTAL
could collect demographic data for stu- SUM OF SQUARES. Compare ROW SUM OF
dents at a particular college by each year SQUARES. 2. in EXPLORATORY FACTOR
of matriculation (i.e., freshman, sopho- ANALYSIS, the variance accounted for by
more, junior, and senior) and then col- each factor across variables. It is deter-
lapse the information from four to two mined by raising the FACTOR LOADINGS
categories (e.g., lower division and for each column in a factor loading ma-
upper division). trix to the second power and then total-
collinearity n. in REGRESSION ANALY- ing the values.
SIS, the situation in which two INDE- c o l u m n vector a data matrix with a
PENDENT VARIABLES are so highly associ-
single column of values. In other words,
ated that one can be closely or perfectly
it has the dimensions r x l , where r refers
predicted by the other. For example,
to the number of rows and 1 denotes the
collinearity likely is present if a re-
single column. Compare ROW VECTOR.
searcher examines how height and age
contribute to children's weight, since c o m b i n a t i o n n. the selection of r ob-
the two predictors are highly interre- jects from among n objects without re-
lated (i.e., as children grow older they gard to the order in which the objects

51
commensurate variable

are selected. The number of combina- search that considers a community as

tions of n objects taken r at a time is the unit of study, such that whole com-
often denoted as C^. A combination is munities are idenfified and undergo an
similar to a PERMUTATION but distin- intervention or treatment. For example,
guished by the irrelevance of order. an investigator interested in the contri-
bution of microbial water contaminants
commensurate variable a variable
to risks of illness might compare a neigh-
that has a similar scale and degree of
borhood for which a water utility re-
variation in responses when compared
cently changed filtration procedures to
to another variable. For example, the
one for which the utility still employs
monthly expenditures and the monthly
the standard filtration procedures.
contributions to savings for households
in a particular geographic area are com- c o m p a r a b i l i t y research a study car-
mensurate variables. ried out to evaluate the EQUIVALENCE of
c o m m o n a l i t y analysis in MULTIPLE different test versions or testing meth-
REGRESSION analysis, a technique by ods. A common example involves com-
which the known variance in an out- paring participant performance on a
come or response variable is separated paper-and-pencil exam versus a com-
out into the parts that can be uniquely puter-administered exam.
attributed to each individual predictor comparable-forms r e l i a b i l i t y see
variable and the parts that are common ALTERNATE-FORMS RELIABILITY.
to any two or more variables.
comparable groups two or more rep-
c o m m o n f a c t o r see SPECIFIC FACTOR. resentative SAMPLES drawn from the
c o m m o n metric a unit or scale of mea- same POPULATION for the purpose of ob-
surement that is applied to data from servation or experiment.
different sources. In a META-ANALYSIS,
comparable f o r m see ALTERNATE
for instance, the results from multiple
studies may need to be placed on a com- FORM.
mon metric so that they may be mean- comparative method an experimen-
ingfully compared. tal research method of analyzing and
c o m m u n a l i t y n. in FACTOR ANALYSIS, comparing the behavior of different spe-
the proportion of variance in one vari- cies of animals, different cultures of hu-
able that is accounted for by an underly- mans, and different age groups of
ing element common to all of the humans and other animals so as to iden-
variables in a set. It is given by the tify similarities and dissimilarities and
COMMUNALITY COEFFICIENT. Compare obtain an understanding of a phenome-
UNIQUENESS. non of interest. For example, macaques
are often used as a model for under-
c o m m u n a l i t y coefficient (symbol: standing human mother-infant rela-
h^) an index of the COMMUNALITY in a tionships. However, male macaques
variable. It is the sum of squared FACTOR rarely become interested in or involved
LOADINGS of that variable over all of the with care of the young, so comparative
common factors (underlying dimen- study of other species is needed to un-
sions) in the analysis. The communality derstand when and how male care of
coefficient is scaled so that, if the factors young develops. Also called compara-
completely account for all the variance tive analysis.
in the variable, the coefficient is 1.
comparative r a t i n g scale a type of
c o m m u n i t y c o n t r o l see NEIGHBOR- rating instrument in which the items are
HOOD CONTROL. not evaluated independently but rather
c o m m u n i t y i n t e r v e n t i o n study re- through comparison with other stimuli.

52
 complete factorial design

For instance, a respondent may be asked blocks (relatively homogeneous subsets)

to rank a set of given options in relation according to some characteristic (e.g.,
to one another, assigning the value 1 to age) that is not a focus of interest and are
the option that is most important to then assigned to the different treat-
him or her, 2 to the second most impor- ments or conditions in such a manner
tant, and so on. Compare ABSOLUTE that each treatment appears once in
RATING SCALE. each block. Thus, the number of partici-
pants in each block must equal the num-
comparative t r i a l see CONTROLLED
ber of experimental conditions. For
TRIAL.
example, the following arrangement of
comparison n. any appraisal of two or four treatments (A, B, C, D) and 16 indi-
more groups in order to identify differ- viduals (from four age groups) is a com-
ences, such as that between the mean of plete block design:
a variable in one population and its
mean in another. Comparisons may be Treatment
planned in advance, as in A PRIORI COM- Block #
1 (children) A B C D
PARISONS, or decided upon after data
2 (adolescents) B C D A
analysis has already begun, as in POST 3 (young adults) C D A B
HOC COMPARISONS. Also called con- 4 (older adults) D A B C
trast.
comparison group see CONTROL
GROUP. By ensuring that the "nuisance" char-
acteristic (here, age) is equally repre-
comparison stimulus (Co) in psycho- sented across all conditions, complete
physical testing, one of a set of stimuli to block designs reduce or eliminate its
be compared with a standard stimulus. contribution to experimental error.
comparison-wise error rate see Compare INCOMPLETE BLOCK DESIGN.
TESTWISE ERROR RATE. See also BLOCK DESIGN; RANDOMIZED
BLOCK DESIGN.
competing risks in SURVIVAL ANALY-
SIS, the set of multiple outcomes that complete-case analysis see LISTWISE
indicate a target event (failure) has oc- DELETION.
curred. For instance, a school district complete counterbalancing a pro-
might identify such competing risks for cess of arranging a series of experimental
student dropout as lack of interest, need- conditions or treatments in such a way
ing to work to support the family, and that every possible sequence of condi-
moving away. Consequently, adminis- tions is given at least once during the
trators might develop a competing-risks study. For instance, the following ar-
model, or statistical representation of rangement of sequences of three treat-
the relationships among these concepts, ments (A, B, C), each assigned to a
in order to understand when dropout is different subgroup of participants, dem-
likely to occur and what specific combi- onstrates complete counterbalancing: A-
nation of factors will be involved for dif- B-C, A-C-B, B-C-A, B-A-C, C-A-B, and C-
ferent students. B-A. Compare INCOMPLETE COUNTER-
complementary events a situation in BALANCING.
which there are only two possible out- complete f a c t o r i a l design a research
comes and observation of one outcome
study involving two or more INDEPEND-
necessarily indicates that the other did
ENT VARIABLES in whlch every possible
not occur.
combination of the levels of each vari-
complete block design a study in able is represented. For instance, in a
which participants are first divided into study of two drug treatments, one (A)

53
complete-linkage clustering
having two dosages and the other (B)
having three dosages, a complete facto-
rial design would pair the dosages ad-
ministered to different individuals or
groups of participants as follows: A j with
BJ, A l with B2, A l with B3, A2 with Bj, A2
with B2, and A2 with B3. Compare FRAC-
TIONAL FACTORIAL DESIGN.
c o m p l e t e - l i n k a g e c l u s t e r i n g in HI-
ERARCHICAL CLUSTERING, a method in
which the distance between two clusters
of items (e.g., people, objects) is com-
puted as the greatest distance between
any two objects in the different clusters.
Also called f a r t h e s t n e i g h b o r . Com- c o m p o n e n t i a l a n a l y s i s 1. any analy-
pare AVERAGE-LINKAGE CLUSTERING;
SINGLE-LINKAGE CLUSTERING.
c o m p l e t e n u l l h y p o t h e s i s a state-
ment regarding the outcome of an ex-
periment that totally (rather than
partially) contradicts the ALTERNATIVE
HYPOTHESIS.
c o m p l e x c o m p a r i s o n an evaluation
that involves comparing some combina- c o m p o n e n t - p l u s - r e s i d u a l p l o t see
tion of two or more groups against one
or more other groups. For example, a re-
c o m p o s i t e h y p o t h e s i s a statistical hy-
searcher investigating the influence of a
new teaching style on test scores might
examine whether the average score
across two classrooms of students differs
from that of a third classroom. Compare
SIMPLE COMPARISON.
c o m p o n e n t b a r g r a p h a BAR G R A P H
in which stacked columns are used to
represent two or more aspects of a data
Freshmen Sophomores
Matriculation
component bar
54
set involving categorical variables. For
example, a researcher who observed the
frequencies of males and females by ma-
triculation year at a particular college
might present the results in a component
bar graph (see below). The categories of
matriculation (freshmen, sophomores,
juniors, seniors) would be shown along
the A:-axis, and the number of males and
females would be given by differently
colored or patterned bars rising above
each category. Also called sectional
bar graph; segmented bar graph;
stacked bar graph.
sis of data in which a process or system is
separated into a series of subprocesses or
components. 2. a set of information-
processing and mathematical tech-
niques that enables an investigator to
decompose an individual's performance
on a cognitive task into the underlying
elementary cognitive processes.
PARTIAL RESIDUAL PLOT.
pothesis that is not specific about all rel-
evant features of a population or that
does not give a single value for a charac-
teristic of a population but allows for a
range of acceptable values. For example,
a statement that the average age of em-
ployees in academia exceeds 50 is a com-
posite hypothesis, as there are a variety
of ages above that number that the aver-
Juniors Seniors
Year
graph
 compound symmetry

age could assume to validate the state- c o m p o u n d bar g r a p h a BAR GRAPH

ment. Alternatively, a composite hy- used to display two or more related sets
pothesis could give a specific value for of data simultaneously, typically scores
one characteristic (e.g., the mean) but on a variable or variables that fall into
not others (e.g., the VARIANCE). Com- different categories. For example, it might
pare SIMPLE HYPOTHESIS. be used to show the number of males
composite r e l i a b i l i t y 1. the aggre- and females enrolled in each major at a
gate reliability of two or more similar particular college (see below). The differ-
items, such as judges' ratings. 2. in ent majors would be given along the
STRUCTURAL EQUATION MODELING, the
horizontal X-AXIS and two columns
extent to which the set of constructs rep- would rise above eachone for males
resented in the model relate to a given and one for femalesto the appropriate
latent variable, CRONBACH'S ALPHA is an
height as given by the frequencies along
index of such reliability. the vertical Y-AXIS. See also COMPONENT
BAR GRAPH; DUAL BAR GRAPH.
composite score a single value ob-
tained by combining the scores on two c o m p o u n d event an event that is the
or more component measures. For ex- intersection of two or more simpler
ample, a teacher may average the scores events. For example, being a man with
from separate tests throughout a semes- schizophrenia is the intersection of the
ter to calculate a final exam grade for a event classes being male and having
student. schizophrenia.

composite variable a variable con- c o m p o u n d p r o b a b i l i t y the likeliness

stmcted from the weighted or averaged that two unrelated (independent) events
combination of two or more component will occur together, either simulta-
variables. For example, a figure skater's neously or in succession. For example,
overall skill during a performance might one could calculate the compound prob-
be determined by judges' evaluations of ability of rolling a 6 twice in a row on a
his or her technical merit, program com- die, of rolling a 6 on the first turn and a 2
plexity, and presentation. on the second, and so forth.
compositional variable a variable c o m p o u n d symmetry a quality of
describing one or more substrata of a data in which the values within each
larger group. An example would be sample or subsample have the same de-
mathematics scores of boys in a class of gree of variability (i.e., have HOMOGENE-
both boys and girls. ITY OF VARIANCE) and the corresponding

Female
a Male

Humanities Math Science Technology

College Major
compound bar graph

55
comprehensive process analysis

pairs of values across the different a related measurement obtained at about

groups have the same COVARIANCE. the same point in time. In testing, the
VALIDITY of results obtained from one
comprehensive process analysis a
method for the study of qualitative data test (e.g., self-report of job performance)
relating to a process, such as change that can often be assessed by comparison
occurs during psychotherapy. Impor- with a separate but related measurement
tant moments that occurred during the (e.g., supervisor rating of job perfor-
process are identified and then described mance) collected at the same point in
in sequence to help inform existing the- time. See also CRITERION VALIDITY.
ories. c o n d i t i o n n. a category or level of a vari-
c o m p u t a t i o n a l f o r m u l a the equa- able .whose values are manipulated by
tion used to calculate values for a statisti- a researcher. Study participants are
then assigned to receive or be exposed
cal concept. This contrasts with the
to one or more of the different condi-
DEFINITIONAL FORMULA, w h l c h iS the
tions.
formal verbal definition of the concept.
c o m p u t a t i o n a l table a matrix that c o n d i t i o n a l d i s t r i b u t i o n a set of
shows the different steps and incremen- scores on one variable expressed as a
tal results involved in performing a cal- function of one or more other variables;
culation. such a distribution gives the score on
one variable given each value of a sec-
concealed measure see UNOBTRUSIVE ond variable.
MEASURE.
c o n d i t i o n a l event an outcome that
conceptual f o r m u l a see DEFINITIONAL may occur given that another event has
FORMULA. occurred.
conceptual r e p l i c a t i o n see REPLICA-
c o n d i t i o n a l expectation the long-
TION. run average value of a RANDOM VARI-
concomitance n. a co-occurrence be- ABLE as a function of one or more other
tween a response or outcome variable variables.
and a variable other than a predictor of c o n d i t i o n a l independence the situ-
interest. concomitant adj. ation in which two or more variables are
concomitant variable see COVARI- unrelated to one another only when
ATE. each is a function of one or more others.
concomitant v a r i a t i o n variation of c o n d i t i o n a l l i k e l i h o o d see CONDI-
two phenomena at the same time, in TIONAL PROBABILITY.
which the variables may be causally re- c o n d i t i o n a l model a model represent-
lated or both may be influenced by a ing a pattern of relationships among a
third variable. set of variables that is dependent on
concordance n. the state or condition given levels or categories of one or more
of being in harmony or agreement. In other variables. Compare UNCONDI-
TWIN STUDIES, for example, it is the pres- TIONAL MODEL.
ence of a given trait or disorder in both conditional odds a ratio of the like-
members of the pair. Compare DISCOR- liness of an event's occurrence to the
DANCE. concordant adj. likeliness of nonoccurrence, where the
concordance coefficient see COEFFI- outcome is dependent on values of one
CIENT OF CONCORDANCE. or more other variables. See ODDS.
concurrent v a l i d i t y the extent to c o n d i t i o n a l p r o b a b i l i t y the likeli-
which one measurement is backed up by ness that an event will occur given that
 confidence level

another event is known to have oc- confidence i n t e r v a l (CI) a range of

curred. Conditional probability plays an values for a population PARAMETER that
important role in statistical theory. The is estimated from a sample with a preset,
probability of observing a particular out- fixed probability (known as the CONFI-
come given that another outcome is DENCE LEVEL) that the range will contain
known to have occurred can be derived the true value of the parameter. The
from a CONTINGENCY TABLE. Also called width of the confidence interval pro-
conditional likelihood. vides information about the precision of
the estimate, such that a wider interval
c o n d i t i o n a l variance a measure of
indicates relatively low precision and a
the fluctuations among the individual
narrower interval indicates relatively
observations in a TIME SERIES. The value
high precision. For example, a confi-
at each point in the sequence thus is a
dence interval for the population mean
function of one or more previous points.
could be calculated with data obtained
Compare UNCONDITIONAL VARIANCE.
from a sample and would provide an es-
confederate n. in an experimental situ- timated range of values within which
ation, an aide of the experimenter who the actual population mean is believed
poses as a participant but whose behav- to lie. A confidence interval often is re-
ior is rehearsed prior to the experiment. ported in addition to the POINT ESTI-
MATE of a population parameter.
confidence b a n d a series of connected
CONFIDENCE INTERVALS around a RE- confidence level a value expressing
GRESSION LINE, as shown in the hypo- the frequency with which a given CON-
thetical example below. A confidence FIDENCE INTERVAL contains the true
band may be simultaneous (e.g., cover- value of the parameter being estimated.
age rate of entire band is .95) or non- For example, a 95% confidence level as-
simultaneous (e.g., pointwise coverage sociated with a confidence interval for
rate is .95). In either case it is a represen- estimating a population mean indicates
tation of the uncertainty in estimating a that in 95% of all estimates based on a
function value (e.g., salary) for a popula- random sample of a given size the confi-
tion based on data (e.g., job satisfaction) dence interval will contain the true
from a sample of that population. value of the population mean. The par-

High confidence interval

forming band

regression line

n confidence interval
n forming band

n I I T"
Higti
Level of Job Satisfaction
confidence band

57
confidence limit

ticular confidence level used is up to the fest) variables are examined in relation to
researcher but generally is 95% or 99%. all underlying (latent) variables, confir-
matory factor analysis imposes explicit
confidence limit either of the values theoretical restrictions so that observed
at the upper and lower ends of a CONFI- measures relate with some (often just
DENCE INTERVAL, which provide an es-
one) latent factors but not others.
timated range for the value of a
population PARAMETER. Also called con- c o n f i r m a t o r y research research con-
fidence bound. ducted with the aim of testing one or
more preexisting hypotheses. Compare
confidentiality n. a principle of ethics
EXPLORATORY RESEARCH.
requiring mental health care or medical
care providers to limit the disclosure of a confound n. in an experiment, an inde-
patient's identity, his or her condition pendent variable that is conceptually
or treatment, and any data entrusted to distinct but empirically inseparable
professionals during assessment, diag- from one or more other independent
nosis, and treatment. Similar protection variables. Confounding makes it impos-
is given to research participants and sur- sible to differentiate that variable's ef-
vey respondents against unauthorized fects in isolation from its effects in
access to information they reveal in con- conjunction with other variables. For
fidence. See INFORMED CONSENT. con- example, in a study of high-school stu-
fidential adj. dent achievement, the type of school
(e.g., private vs. public) a student at-
c o n f i g u r a t i o n a l frequency analy- tended prior to high school and his or
sis a nonparametric method of compar- her prior academic achievements in that
ing observed counts within the context are confounds. Also called con-
categories of a variable with the frequen- founder; confounding factor; con-
cies that are theoretically expected. It is founding variable.
an alternative to similar methods (e.g.,
CHI-SQUARE TESTS) that require a mini- confounded comparison a compari-
mum expected count within each cate- son of values obtained by different ex-
gory in order for the results to be valid. perimental groups on an outcome or
dependent variable when two or more
c o n f i r m a t i o n bias the tendency to
predictor or independent variables vary
gather evidence that confirms preexist-
simultaneously across the groups. In
ing expectations, typically by emphasiz-
such cases it is impossible to differenti-
ing or pursuing supporting evidence
ate the effects of the independent vari-
while dismissing or failing to seek con-
ables. For example, consider a researcher
tradictory evidence.
studying how material presentation for-
c o n f i r m a t o r y data analysis a statis- mat (lecture vs. computer) and teacher
tical analysis designed to address one or sex (male vs. female) affect student
more specific research questions, gener- learning. If the investigator were to ex-
ally with the aim of confirming precon- amine data for only two groups of stu-
ceived hypotheses. Compare EXPLORA- dents (those who had a male teacher and
TORY DATA ANALYSIS. lecture presentation format vs. those
who had a female teacher and computer
confirmatory factor analysis (CFA) presentation format), he or she would
any method of testing A PRIORI hypoth- have created a confounded comparison.
eses to the effect that the relationships Compare UNCONFOUNDED COMPARI-
among a set of observed variables are SON.
due to a particular set of unobserved
variables. Unlike EXPLORATORY FACTOR confounded effects the indivisible ef-
ANALYSIS, in which all measured (mani- fects of two or more predictor or inde-

58
 consistency check

pendent variables on a single response measuring the joint effect of two or

or dependent variable. That is, the more predictor or INDEPENDENT VARI-
unique influence of one predictor can- ABLES on the ordering of an outcome or
not be separated out from that of the DEPENDENT VARIABLE. It is used with
others. For example, assume a researcher NONMETRIC data, notably in ITEM RE-
is examining the effectiveness of a hun- SPONSE THEORY.
ger-reduction pill for weight loss. If consensual d r i f t the tendency for two
some of the participants receiving the or more observers who are working to-
pill discover they are part of the treat- gether to depart graduallyfromaccuracy
ment group and concurrently adopt or an agreed standard, such that they are
better eating habits and increase their making errors in the same direction and
exercise levels, these additional healthy to the same extent. See OBSERVER DRIFT.
activities will also affect weight loss. Since
those in the control group (receiving a consensual v a l i d i t y the degree to
PLACEBO sugar pill) are not engaging in which self-reports made by an individ-
these same supplemental activities, ef- ual coincide with scores or ratings made
fects of eating habits, exercise levels, and by other people who know that individ-
the actual treatment will have become ual (e.g., family members, work col-
confounded. leagues). See VALIDITY.
c o n f o u n d i n g f a c t o r see CONFOUND, consent n. see INFORMED CONSENT.

confounding variable seeCON- consequential v a l i d i t y the extent to

FOUND. which a research project has wider social
benefits, such that results from the pro-
c o n f u s i o n m a t r i x see CLASSIFICATION ject influence decisions that affect peo-
TABLE. ple's lives. Many researchers question
congruence coefficient in FACTOR whether this is truly a form of VALIDITY.
ANALYSIS, an index ofthe degree of simi-
consequent variable a variable that is
larity between two underlying or LA-
correlated with a primary outcome or
TENT VARIABLES that are postulated to
DEPENDENT VARIABLE and whose value
account for the data pattern among ob- is predicted by that outcome variable.
served or MANIFEST VARIABLES. It ranges For example, in studies examining the
from -1 to +1 in value. influence of teaching style on student
congruent v a l i d i t y see CONVERGENT academic achievement, student motiva-
VALIDITY. tion may be a consequent variable of
achievement. Compare ANTECEDENT
conjoint analysis a method of assess- VARIABLE.
ing behavioral choices in which partici-
pants are compared on pairs of attributes conservative adj. describing statistical
that have been selected by the researcher methods that have a reduced likelihood
as potential explanations or predictors of error, particularly of incorrectly re-
of such choices. For example, an investi- jecting the NULL HYPOTHESIS, but that
gator looking at vacation preferences are less capable of detecting significant
might choose to examine such attributes relationships among variables. In other
as warm versus cold climates, hotels ver- words, conservative approaches have
sus bed-and-breakfast accommodations, less POWER and thus tend to underesti-
destinations in the United States versus mate associations or effects.
destinations in Europe, and so forth.
Conjoint analysis is used particularly in consistency check a method often
studies of consumer decision making. built into an experimental or statistical
procedure to verify that data are accu-
conjoint measurement a method for rate or that computations have been

59
consistent estimator

done correctly. For example, question- of 1 kg is judged on average to be 1.5 kg,

naire items with slightly different word- and a weight of 2 kg is judged to be 2.5
ing but identical meaning should be kg, the constant error is 500 g. See also
answered by individuals in the same ABSOLUTE ERROR; RANDOM ERROR.
way, and respondents claiming to have a
high educational degree should be of a construct n. a concept or theoretical en-
certain age. tity. It may be derived from empirically
verifiable and measurable events or pro-
consistent estimator an ESTIMATOR cessesan empirical constructor via
calculated from sample data whose processes inferred from data of this kind
value becomes more and more represen- but not themselves directly observable
tative of the true quantity in the larger a hypothetical construct.
population as the sample size increases.
That is, a consistent estimator is one for construct v a l i d i t y the degree to
which larger samples result in a reduc- which a test or instrument is capable of
tion of BIAS. An example is the MEAN: AS measuring a concept, trait, or other the-
the size of the sample from which it is oretical entity. For example, if a re-
calculated increases, the obtained value searcher develops a new questionnaire
more closely approximates the actual to evaluate respondents' levels of aggres-
value in the population it is intended to sion, the construct validity of the instru-
represent. ment would be the extent to which it
actually assesses aggression as opposed
constant n. in mathematics, a fixed to assertiveness, social dominance, and
quantity that remains unchanged dur- so forth. A variety of factors can threaten
ing a specified operation or series of op- the basic construct validity of an experi-
erations. For example, the element a in ment, including (a) mismatch between
the REGRESSION EQUATION the construct and its OPERATIONAL DEFI-
(b) use of a single method to col-
NITION;
y= a bx e lect information on all variables in the
is a constant. Compare VARIABLE. study, thereby potentially introducing
bias; (c) failure to distinguish between
constant comparative analysis a constructs and levels of constructs, so
procedure for evaluafing QUALITATIVE that observations made on only a few
DATA in which the information is coded levels or conditions are extended by in-
and compared across categories, pat- ference to other levels; and (d) various
terns are identified, and these patterns EXPERIMENTER EFFECTS and other partic-
are refined as new data are obtained. For ipant reactions to aspects of the experi-
example, a researcher might use con- mental situation. There are two main
stant comparative analysis to assess re- forms of construct VALIDITY in the social
sponses to interview questions, creating sciences: CONVERGENT VALIDITY and
categories of answers according to the DISCRIMINANT VALIDITY. See also
perspectives expressed, examining their MULTITRAIT-MULTIMETHOD MATRIX.
different items, and integrating catego-
ries or revising their properties before contaminated normal distribution
formulating a theory. Also called con- a distribufion that appears to be a NOR-
stant-comparison method. MAL DISTRIBUTION but Is in fact character-
ized by multiple such distributions, each
constant error a SYSTEMATIC ERROR in of which has different parameters (e.g.,
some particular direction. Constant different STANDARD DEVIATIONS).
error is computed as the average positive
or negative difference between the ob- c o n t a m i n a t i o n n. in testing and exper-
served and actual values along a dimen- imentation, the situation in which prior
sion of interest. For example, if a weight knowledge, expectations, or other fac-

60
 contingency table

tors relating to the variable under study CONTINGENCY TABLE. 2. more generally,
are permitted to influence the collection the extent to which one event or out-
and interpretation of data about that come is dependent on another.
variable. contingency analysis an approach to
content analysis a procedure for as- measuring the association between two
signing codes to identify themes in writ- variables whose values are represented
ten or spoken records. A content analy- by unordered categories. For example, a
sis of a speech, for example, may involve researcher may use contingency analysis
a count of the number of times a particu- to determine whether children who are
lar behavior occurs or the number of undergoing a painful medical procedure
times a particular idea is mentioned. become less distressed when nurses reas-
sure them or whether nurses reassure
content-referenced test see CRITE- children more when they are distressed.
RION-REFERENCED TEST.
See COEFFICIENT OF CONTINGENCY.
content v a l i d i t y the extent to which a contingency coefficient see COEFFI-
test measures a representative sample of CIENT OF CONTINGENCY.
the subject matter or behavior under in-
vestigation. For example, if a test is de- contingency table a two-dimensional
signed to survey arithmetic skills at a table in which frequency values for cate-
third-grade level, content validity indi- gories of one variable are presented in
cates how well it represents the range of the rows and values for categories of a
arithmetic operations possible at that second variable are presented in the col-
level. Although researchers traditionally umns: Values that appear in the various
relied upon their own subjective impres- CELLS then represent the number or per-
sions or the judgments of experts as the centage of cases that fall into the two
primary means of determining content categories that intersect at this point.
validity, modern approaches involve the For example, the sex and geographical
use of EXPLORATORY FACTOR ANALYSIS
locations of a sample of individuals ap-
and other multivariate statistical proce- plying for a particular job may be dis-
dures. played in a contingency table.
contextual analysis any method of Sex
evaluating data that takes into account
the characteristics of the environment Location Female Male Total
in which the information was collected Los Angeles 19 14 33
and their influence upon study units.
New York 19 15 34
contextual effect the influence on an
outcome or response variable of the en- Seattle 15 10 25
vironment in which data are collected. Tampa 13 12 25
For example, a study participant may be-
have differently in a laboratory than in a Washington, DC 17 16 33
public setting. Total 83 67 150
contextual variable any variable that
describes the environment in which
Thus, the number of women from Los
data are collected, such as its location or
Angeles are given, the number of men
the degree of social interaction.
from Los Angeles are given, the number
contingency n. 1. the degree to which of women from New York City are given,
one CATEGORICAL VARIABLE is associ- the number of men from New York City
ated with another such variable, as rep- are given, and so on. Also called cross-
resented by the frequencies recorded in a classification table.

61
contingent probability

contingent p r o b a b i l i t y the probabil- tinuum (e.g., a line) rather than accord-

ity, expressed as a number between 0 ing to categories. Such ratings are made
and 1, that one event or category will by making a mark on the scale to indi-
occur if another one does. An example cate the exact "placement" of the rating
would be the probability that the child or by assigning a precise numerical
of a drug user will become a drug user value. Also called continuous scale.
himself or herself. Unusually high or continuous-time survival model
low contingent probabilities (compared
in SURVIVAL ANALYSIS, a method for pre-
to the general population) may, but do
dicting time until the occurrence of an
not necessarily, imply a causal relation-
event in which time is treated as a CON-
ship between the two events.
TINUOUS VARIABLE. For example, the time
c o n t i n u i t y correction see CORREC- a recovering alcoholic can expect to go
TION FOR CONTINUITY. without a drink could be expressed as a
continuous-time survival model. Com-
continuous adj. describing a variable, pare DISCRETE-TIME SURVIVAL MODEL.
score, or distribution that can take on
any numerical values within its range. continuous variable a variable that
Compare DISCONTINUOUS; DISCRETE. may in theory have an infinite number
See also CONTINUOUS VARIABLE. of possible values. For example, time is a
continuous variable because accurate in-
continuous d i s t r i b u t i o n a DISTRIBU- struments will enable it to be measured
TION in which values can occur any- to any subdivision of a unit (e.g., 1.76
where along an unbroken continuum. seconds). By contrast, number of chil-
An example would be any distribution dren is not a continuous variable as it is
showing variation in human height or not possible to have 1.76 children. In
weight. A continuous distribution can practice, a continuous variable may be
be plotted as a single smooth line, and it restricted to an artificial range by instru-
may be used to display the likelihood of mentation constraints, practical hmita-
specific values occurring (a continuous tions, or other reasons. For example, a
probability distribution) or the actual researcher assessing the influence of a
number of times they have been ob- new technique on student study time
served to occur, such as in a research may only be able to observe a group of
sample (a continuous frequency distri- individuals for 1 hour per day, such that
bution). Compare DISCRETE DISTRIBU- the range of time in the data he or she
TION. See also CONTINUOUS VARIABLE. collects may span 0 minutes to 60 min-
continuous f u n c t i o n a function utes, even though some people will in
f(x)in which an infinitely small change actuality have exceeded that upper figure.
Compare DISCONTINUOUS VARIABLE.
in input has an infinitely small change
on output. When depicted graphically, contour plot a two-dimensional dis-
the values of a continuous function play of a three-dimensional relationship.
form a smooth, unbroken curve. Com- The j-axis and y-axis of a CARTESIAN CO-
pare DISCONTINUOUS FUNCTION. ORDINATE SYSTEM are used to depict the
values of two of the variables while the
continuous r a n d o m variable a vari- values of the third variable are repre-
able that may assume any of an infinite sented by raised or shaded lines, similar
range of values according to a particular to how elevations are shown on a topo-
PROBABILITY DISTRIBUTION. Height and graphical map. For example, the graphic
weight are examples. Compare DISCRETE opposite is a hypothetical contour plot of
RANDOM VARIABLE. age and conservative attitudes in a sam-
continuous r a t i n g scale a scale on ple and their DENSITY ESTIMATION val-
which ratings are assigned along a con- ues in the larger population of interest.

62
 controlled sampling

contrast n. see COMPARISON. ING, the WEIGHT assigned to each value

contrast analysis comparisons be- in a set of values to be compared.
tween two or more groups that address c o n t r o l n. 1. the regulation of all extra-
specific questions about the precise pat- neous conditions and variables in an ex-
terns of difference between the MEANS of periment so that any change in the
these groups. Where several groups are DEPENDENT VARIABLE can be attributed
being compared, CONTRAST CODING can solely to manipulation of the INDEPEN-
be used to assign weights to the different DENT VARIABLE and not to any other fac-
means. tors. 2. see STATISTICAL CONTROL.
contrast c o d i n g in ANALYSIS OF VARI- c o n t r o l c o n d i t i o n in an experiment
ANCE and MULTIPLE REGRESSION, a or research design, a condition that does
method of assigning weights to values so not involve exposure to the treatment or
that specific comparisons of interest can intervention under study. For example,
be made. To ensure the statistical inde- in an investigation of a new drug, partic-
pendence of the particular quantities ipants in a control condition may re-
being evaluated, the process requires ceive a pill containing some inert
that the weights in each comparison substance, whereas those in the EXPERI-
sum to 0. For example, consider a re- MENTAL CONDITION receive the actual
searcher investigating four different drug of interest.
methods for teaching course content. If
he or she wanted to compare the average c o n t r o l experiment a follow-up ex-
performance of the first group to the av- periment designed to check that the
erage performance of the remaining findings of an initial study were in fact
three groups, the investigator would caused by the factor under investigation
apply contrast weights of +3, -1, -1, and and not by some other variable. For ex-
-1 to the respective group means. If, ample, consider a case in which initial
however, he or she wanted to compare research appeared to demonstrate the
the average performance of the first beneficial effect of a new treatment for
group to the average performance of the depression. To confirm these results a
fourth group, the investigator instead control experiment might be devised in
would apply contrast weights of +1, 0, 0, which a similar group of individuals was
and -1 to the respective group means. tracked to find the number of individu-
Also called orthogonal coding. Com- als whose depression was alleviated
pare DUMMY VARIABLE CODING; EFFECT without use of the treatment.
CODING.
c o n t r o l group a comparison group in a
contrast weight in CONTRAST COD- study whose members receive either no
intervention at all or some established
intervention. The responses of those in
the control group are compared with the
responses of participants in one or more
EXPERIMENTAL GROUPS that are given
the new treatment being evaluated. See
also NQ-TREATMENT CONTROL GROUP.
controlled observation an observa-
tion made under standard and system-
atic conditions rather than casual or
incidental conditions.
controlled s a m p l i n g a method of
choosing cases for research in which cer-

63
controlled trial

tain characteristics are considered unde- means (e.g., a PARTIAL CORRELATION).

sirable and the probability of their being Also called controlled variable.
selected is therefore minimized. For ex-
ample, an educational researcher may convenience sampling any process
need to adjust his or her sampling tech- for selecting a sample of individuals or
nique if certain subsets of students are cases that is neither random nor system-
costly to find or if their obtained data atic but rather governed by chance or
may exert undue influence on study re- ready availability. Interviewing the first
sults or otherwise present logistical diffi- 50 people to exit a store is an example of
culties. convenience sampling. Data obtained
from convenience sampling do not gen-
controlled trial a study in which pa- eralize to the larger population; there
tients with a particular condition, dis- may be significant SAMPLING BIAS, and
ease, or illness are assigned either to a SAMPLING ERROR cannot be estimated.
treatment group, which receives the Also called accidental sampling; op-
new intervention under investigation, portunity sampling.
or to a CONTROL GROUP, which receives
either no intervention or some standard convergent validity the extent to
intervention already in use. If individu- which responses on a test or instrument
als are allocated to the different groups exhibit a strong relationship with re-
at random, then the design is a RAN- sponses on conceptually similar tests or
DOMIZED CLINICAL TRIAL. Also Called instruments. This is one of two aspects
comparative trial. of CONSTRUCT VALIDITY, the other being
DISCRIMINANT VALIDITY. Also Called Con-
controlled variable 1. see CONTROL gruent validity.
VARIABLE. 2. an infrequent synonym of
INDEPENDENT VARIABLE, the Variable in conversation analysis a specialty
a study whose values are manipulated by within DISCOURSE ANALYSIS that focuses
the researcher. upon casual discussions as well as other
more formal extended verbal exchanges
c o n t r o l procedure a procedure ap- between two or more speakers.
plied before commencing a study in
order to equate participants with regard Cook's flistance (Cook's D) in an
to some variable that is not of research analysis of the relationship between a re-
interest but nonetheless may influence sponse variable and one or more predic-
the outcome, MATCHING is an example tor variables, a measure of the difference
of a control procedure. that is made to the result when a single
observation is dropped from the analy-
c o n t r o l series a series of CONTROL EX- sis. Cook's D thus indicates the degree of
PERIMENTS devised to eliminate any al- influence of a particular data value. An
ternative explanation of the findings observation typically is considered in-
obtained from a test. fluential if it has a Cook's D larger than
c o n t r o l treatment an existing or stan- 4/(n - k - 1), where n is the sample size
and k is the number of terms in the
dard intervention administered to mem-
model. [R. Denis Cook (1944- ), U.S.
bers of a CONTROL GROUP in a study.
statistician]
c o n t r o l variable a variable that is con-
sidered to have an effect on the response cooperative study see MULTISITE
measure in a study but which itself is not STUDY.
of particular interest to the researcher. coordinate axis see AXIS.
To remove its effects a control variable
may be held at a constant level during coordinate system see CARTESIAN CO-
the study or managed by statistical ORDINATE SYSTEM.

64
 correlational research

c o r r e c t e d m o m e n t see U N C O R R E C T E D or her available sample is taken from or-

MOMENT. ganizations that require individuals to
have obtained at least a bachelor's de-
c o r r e c t i o n n. a quantity that is added,
gree in order to be employed, then a cor-
subtracted, or otherwise introduced to
rection for range restriction would be
remove inaccuracy from a measure, cal- appropriate. See RESTRICTION O F R A N G E .
culation, or analysis. Corrections are
made, for example, when the ASSUMP- c o r r e l a t e n. a variable that is related to
TIONS about data that accompany a sta- another variable. See CORRELATION.
tistical procedure do not correspond to
c o r r e l a t e d g r o u p s see DEPENDENT
the actual data being analyzed.
SAMPLES.
c o r r e c t i o n f o r a t t e n u a t i o n in anal-
correlated-groups design see
yses estimating the relationship be-
WITHIN-SUBJECTS DESIGN.
tween two variables, an adjustment for
error introduced during the process of c o r r e l a t e d - g r o u p s t test see DEPEN-
obtaining the measures, where such DENT-SAMPLES T TEST.
error serves to underestimate the mea-
c o r r e l a t e d s a m p l e s see DEPENDENT
sured effect.
SAMPLES.
c o r r e c t i o n f o r c o n t i n u i t y a proce-
correlated-samples design see
dure that is applied to adjust for the fact
WITHIN-SUBJECTS DESIGN.
that a statistical method is based on an
assumption that the data have a C O N - c o r r e l a t e d - s a m p l e s t test see D E -
TINUOUS DISTRIBUTION when, in fact, PENDENT-SAMPLES T TEST.
they have a DISCRETE DISTRIBUTION.
c o r r e l a t i o n n. the degree of a relation-
YATES'S CORRECTION FOR CONTINUITY IS
ship (usually linear) between two vari-
an example. Also called continuity
ables, which may be quantified as a
correction.
C O R R E L A T I O N COEFFICIENT.
c o r r e c t i o n f o r g u e s s i n g a scoring rule
correlation analysis any of various
for MULTIPLE-CHOICE TESTS SUCh that
statistical procedures for identifying re-
the EXPECTED VALUE of getting an item
lationships among variables and de-
correct under the assumption of no
termining their strength or degree, CA-
knowledge is 0 rather than 1/n, where n
NONICAL ANALYSIS and FACTOR ANALY-
is the number of alternatives. Thus, if a
SIS are examples. Correlation analyses
test taker inconectly answered four ques-
make no inferences about causality
tions, a further 4/n would be deleted
(compare CAUSAL ANALYSIS). Also called
from his or her score. By thus penalizing
association analysis.
inaccurate responses, the correction dis-
courages a test taker from guessing and c o r r e l a t i o n a l f a l l a c y the fallacy that
provides a better estimate of that per- association implies causation: the prac-
son's tme ability. See also F O R M U L A tice of drawing conclusions about cause
SCORING. and effect based solely on observations
of a relationship between variables. For
c o r r e c t i o n f o r r a n g e r e s t r i c t i o n in
example, assume a researcher found that
estimations of the relationship between
dieters tend to weigh more than other
two variables, an adjustment for error
people. If the investigator then were to
arising from the use of a sample that is
conclude that the diet is responsible for
not representative of the larger popula-
the additional weight, then he or she has
tion of interest. For example, consider a
committed a correlational fallacy.
researcher who is investigating how cog-
nitive ability and absenteeism are re- c o r r e l a t i o n a l r e s e a r c h a type of
lated in the general U.S. workforce. If his study i n which relationships between

65
correlation coefficient

variables are simply observed without sider the hypothetical example below de-
any control over the setting in which picfing correlations among anhedonia,
those relationships occur or any manipu- hopelessness, low self-esteem, and suicidal
lation by the researcher, FIELD RESEARCH ideation in a sample of college students.
often takes this form. For example, con-
correlation ratio (symbol: r|) a mea-
sider a researcher assessing teaching
sure of the strength or degree of relation-
style. He or she could use a correlational
ship between two variables whose
approach by attending classes on a col-
association is nonlinear (i.e., it cannot
lege campus that are each taught in a dif-
be depicted graphically by a straight
ferent way (e.g., lecture, interactive,
line). Often referred to as eta, the corre-
computer aided) and noting any differ-
lation ratio is the nonlinear equivalent
ences in student learning that arise. Also
called correlational design; correla- of the CORRELATION COEFFICIENT.
tional method; correlational study. correlogram n. see AUTOCORRELATION.
Compare EXPERIMENTAL RESEARCH.
correspondence analysis a statistical
correlation coefficient a numerical technique in which information from a
index reflecting the degree of linear rela- two-way CONTINGENCY TABLE is trans-
tionship between two variables. It is formed into a graphical display. This can
scaled so that the value of +1 indicates a then be analyzed to understand the as-
perfect positive relationship (such that sociations between the various catego-
high scores on variable x are associated ries of the different variables involved.
with high scores on variable y), -1 indi-
cates a perfect negative relationship c o t w i n c o n t r o l see TWIN CONTROL.
(such that high scores on variable x are count data information about the fre-
associated with low scores on variable y, quency of an attribute or event. An ex-
or vice versa), and 0 indicates no rela- ample would be the number of times a
tionship. The most commonly used type third-grader acts out in class during a 30-
of correlation coefficient is the Pearson minute observation period.
PRODUCT-MOMENT CORRELATION COEF-
FICIENT.
counterbalancing n. arranging a se-
ries of experimental conditions or treat-
correlation m a t r i x a symmetrical ments in such a way as to minimize the
SQUARE MATRIX displaying the degree of influence of extraneous factors, such as
association between all possible pairs of practice or fatigue, on experimental re-
variables contained in a set. A CORRELA- sults. In other words, counterbalancing
TION COEFFICIENT between the ith and is an attempt to reduce or avoid CARRY-
/'th variables in a set of variables is dis- OVER EFFECTS and ORDER EFFECTS, A
played in the intersection of the ith row simple form of counterbalancing would
and the /th column of the matrix. Con- be to administer experimental condi-

Variable 1 2 3 4

1. Anhedonia

2. Hopelessness .82

3. Low self-esteem .77 .31

4. Suicidal ideation .23 .90 .54

correlation matrix
 covariance matrix

tions in the order A-B to half of the par- pothesis on the basis of a significance
ticipants and in the order B-A to the test with having demonstrated a scien-
other half; a LATIN SQUARE would be a tifically important effect.
more complex form.
c o u n t i n g process a type of STOCHAS-
counterfactual n. an alternative out- TIC PROCESS used to describe or analyze
come: a consideration of what would data that represent an ordered series of
have been observed had something taken the times to multiple events, such as a
place that in fact did not. Counter- study of the time to the first event, time
factuals are used primarily in RUBIN'S to the second event, and so on.
CAUSAL MODEL. For example, consider a
covariance n. a scale-dependent mea-
researcher seeking to determine the ef-
sure of the relationship between two
fect of a new depression treatment as
variables such that corresponding pairs
compared to an existing standard treat-
of values of the variables are studied
ment. According to Rubin's model, the
with regard to their relative distance
investigator would need to determine
from their respective means. A positive
the average difference for all participants
covariance results when values of one
between (a) the outcome after adminis-
variable that lie above the mean of that
tering the new treatment and (b) the
variable tend to be paired with values of
outcome after administering the alterna-
the second variable "that also lie above
tive treatment. Because the same partici-
the mean of that variable. A negative
pants cannot simultaneously be given
covariance results when values of one
the new and old treatments, the situa-
variable that lie above the mean tend to
tion arises in which certain experimen-
be paired with values of the second vari-
tal outcomes cannot be observed and
able that lie below the mean. A co-
become counterfactuals that instead must
variance is equal to 0 when two variables
be estimated.
are independent, or unrelated to one an-
c o u n t e r n u l l value the magnitude of other.
an effect (such as a difference in averages
between two groups or a measure of the covariance analysis see ANALYSIS OF
COVARIANCE.
strength of the relationship between
two variables) that carries with it a level covariance m a t r i x a SQUARE MATRIX
of statistical support that is equivalent to that represents how variance in each
the statistical support attributed to the variable in a set is related to variance in
value specified under a competing hy- all other variables in the set. Consider
pothesis (i.e., the NULL HYPOTHESIS). the following hypothetical example for
The counternull value helps eliminate five variables pertaining to perceived
two common errors: (a) equating failure health. The covariances between pairs of
to reject a null hypothesis with the esti- variables are located at the intersection
mation of the effect size as equal to zero of the row and column that correspond
and (b) equating rejection of a null hy- to the two variables. The quanfifies along

Variable 1 2 3 4 5

1. Ul enough to see doctor in past year .24 -

2. Hospitalized or disabled last year .06 .14 -
3. Hospitalized or disabled prior 4 years .03 -.03 .16 -
4. Self-rated health -.09 -.08 .00 .65 -
3. Health satisfaction -.18 -.13 -.02 .62 .91

covariance matrix

67
covariance model

the diagonal of the matrix are variances in the investigation. See DECEPTION RE-
rather than covariances. Also called <lis- SEARCH.
persion matrix; varianee-covari-
ance matrix. COVRATIO a numerical value corre-
sponding to a unit (e.g., individual, ani-
covariance model a mathematical mal) that indicates the effect of
specification of the pattern of relation- removing the unit from a REGRESSION
ships among a set of quantitative vari- ANALYSIS. More technically, it is a ratio
ables that expresses how variance in of the DETERMINANTS Of COVARIANCE
each variable may be related to variance MATRICES with and without a given ob-
in each of the other variables. servation. Values less than 1 indicate
that removal of the unit has little influ-
covariance structure analysis see
ence on the precision of the estimates
ANALYSIS OF COVARIANCE STRUCTURES. obtained from the analysis.
covariate n. a variable that exhibits
Cox-Mantel test see LOG-RANK TEST.
COVARIATION with a measured outcome
[David Cox (1924- ), British statisti-
or DEPENDENT VARIABLE: It is Often in-
cian; Nathan Mantel (1919-2002), U.S.
cluded in an analysis so that its effect
biostatistician]
may be taken into account when inter-
preting the effects of the INDEPENDENT Cox regression analysis a statistical
VARIABLES of interest. For example, co- technique used to build multivariate
variates are used in ANALYSES OF CO- models that relate one or more continu-
VARIANCE to statistically adjust groups ous or categorical variables to SURVIVAL
so that they are equivalent with regard TIMES, without requiring researchers to
to these variables; they may also be used specify in advance the form or nature of
in MULTIPLE REGRESSION to minimize such relationships. For example, one
error that may arise from omitting any might use Cox regression to determine
noncentral but potentially influential how likely it is that alcoholics who are
variables. Also called concomitant abstinent at three months and at six
variable. months will relapse. A key methodologi-
cal concept in Cox regression analysis is
covariation n. a relationship between
the hazard, that is, the immediate po-
two quantitative variables such that as
tential or "risk" of event occurrence. It is
one variable tends to increase (or de-
computed from the baseline hazard rate,
crease) in value, the corresponding val-
which estimates the overall risk of event
ues of the other variable tend to also
occurrence as a function of time; various
increase (or decrease). For example, if a
coefficients, which describe the relation-
person's weight consistently rises as he
ship between each predictor variable
or she grows older, then the two vari-
and the rate of event occurrence; and
ables would be exhibiting covariation.
each individual's values on the predictor
cover story a plausible but false state- variables. There are two types of Cox
ment about the purpose of a research regression: the simpler standard Cox
study that is given to participants to regression model (or [Cox's] propor-
avoid disclosing to them the true hy- tional hazards model) and a more com-
pothesis being investigated. Such decep- plex generalization known as the ex-
tion may be practiced when the par- tended Cox regression model (or time-
ticipants' behavior in the study is apt to dependent Cox regression model). The
be affected by knowledge of the experi- standard model requires that the PRO-
ment's true purpose. For ethical reasons, PORTIONAL HAZARDS ASSUMPTION be
the decepfion should not flagrantly vio- met and thus is used when the risk of
late the participants' right to know what event occurrence for the reference and
they will be getting into by taking part comparison groups remains constant

68
 criterion contamination

relative to one another over all time the degree of association between two
points. The extended model is used variables that have two or more unor-
when the effect of particular variables dered response categories. More specifi-
on the occurrence of the event of inter- cally, it is an omnibus EFFECT SIZE that
est changes over time, such that one quantifies the overall association among
group has a higher risk of event occur- the rows and columns in a CONTIN-
rence at early time points but a lower GENCY TABLE. Also called Cramer's
risk at later time points. [David Cox] phi. [Harald Cramer]
Cox-Snell residual the discrepancy Cramer-von Mises goodness-of-fit
between observed and predicted values test a statistical procedure to evaluate
for a variable defined as the time to a how well the distribution of an observed
particular event (e.g., time to death). set of scores matches a theoretical distri-
Used in cox REGRESSION ANALYSIS, such bution (e.g., a NORMAL DISTRIBUTION).
RESIDUALS provide a means for evaluat- It examines the discrepancies between
ing how a theory about changes in SUR- obtained scores and their corresponding
VIVAL TIMESfitsan observed set of data. expected scores to derive a single statis-
[David Cox; E.Joyce Snell] tic (W^), the larger the value of which
the poorer the model fit. [Harald
Cox's proportional hazards model Cramer; Richard E. von Mises (1883-
see cox REGRESSION ANALYSIS. [David 1953), Russian-born U.S. mathematician
Cox] and engineer]
Cox-Stuart test f o r t r e n d a statisti-
credible interval in BAYESIAN meth-
cal procedure for evaluating whether a
ods, a range within which a particular
sequence of independent scores follows
value of a population characteristic has a
a systematic pattern, for example,
specified probability of falling. It is the
whether earlier observations have an
Bayesian equivalent of the CONFIDENCE
overall tendency to be larger or smaller
INTERVAL used in FREQUENTIST ap-
than later observations. [David Cox; Alan
proaches. Also called Bayesian confi-
Stuart (1922-1998), British statistician]
dence interval; credible region.
Cramer-Rao l o w e r b o u n d a lower
c r i t e r i o n n. a standard against which a
limit on the extent to which estimated
judgment, evaluation, or comparison
values of a population parameter may
can be made. For example, a well-
vary, given that the ESTIMATOR is unbi- validated test of creativity might be used
ased. For example, suppose a researcher as the criterion to develop new tests of
examines a sample to derive possible es- creativity.
timates of the incidence of a rare disor-
der in the larger population. He or she c r i t e r i o n analysis a method of analyz-
could obtain a Cramer-Rao lower bound ing responses to a test in which it is as-
and compare it to the variance of the es- sumed that some unobserved variable
timates obtained using a particular UN- underlies the target behavior or concept
BIASED ESTIMATOR. Such a comparison of interest and accounts for the pattern
indicates whether the estimator in ques- of relationships (CORRELATIONS) among
tion is efficient, that is, whether given the test items.
alternative estimators it is the one with
the lowest degree of variance. Also called c r i t e r i o n c o n t a m i n a t i o n a situation
m i n i m u m variance bound. [Harald in which a response measure (the crite-
Cramer (1893-1985), Swedish mathe- rion) is influenced by factors that are not
matician; Calyampudi R. Rao (1920- ), related to the concept being measured.
Indian-born U.S. statistician] Evidence of this may be observed
through correlations of the response
Cramer's V (symbol: V; a measure of measure with variables that are concep-

69
criterion cutoff

tually distinct from that measure. For ex- that serves as a standard against which
ample, performance discrepancies (in other scores may be judged in a CRI-
dollars sold) among insurance agents TERION-REFERENCED TEST. 2. in REGRES-
may arise not from any actual differ- SION ANALYSIS, a predicted score on an
ences in ability but rather from socioeco- attribute or variable.
nomic differences in territories assigned
to the salespeople. c r i t e r i o n v a l i d i t y an index of how
well a test correlates with an established
c r i t e r i o n c u t o f f see CUTOFF SCORE. standard of comparison (i.e., a CRI-
TERION). Criterion validity is divided into
c r i t e r i o n group a group tested for three types: PREDICTIVE VALIDITY, CON-
traits its members are already known to CURRENT VALIDITY, and RETROSPECTIVE
possess, usually for the purpose of dem- VALIDITY. For example, if a measure of
onstrating that responses to a test repre- criminal behavior is valid, then it should
sent the traits they were intended to be possible to use it to predict whether
represent. For example, a group of chil- an individual (a) will be arrested in the
dren with diagnosed visual disabilities future for a criminal violation, (b) is cur-
may be given a visual test to assess its va- rently breaking the law, and (c) has a
lidity as a means of evaluating the pres- previous criminal record. Also called
ence of visual disabilities. criterion-referenced validity; cri-
c r i t e r i o n index an index that mea- terion-related validity.
sures a specific quality on which a set of c r i t e r i o n variable see DEPENDENT
scores may be evaluated. Examples in- VARIABLE.
clude indices of RELIABILITY and VALID-
ITY. c r i t i c a l difference in comparisons of
means or other statistics obtained from
criterion-referenced test an exam two or more samples, the minimum dif-
from which decisions are made about an ference that is deemed necessary to
individual's absolute level of accom- judge a test result as having STATISTICAL
plishment (i.e., mastery or nonmastery) SIGNIFICANCE. The Critical difference
of the material covered in that exam ac- will vary according to the procedure
cording to some standard reference point. used.
For example, if a student obtains a score
of 70% on a reading exam and a passing c r i t i c a l experiment see CRUCIAL EX-
score is 65%, then he or she has done PERIMENT.
well. Also called content-referenced
test. See also DOMAIN-REFERENCED TEST; c r i t i c a l r a t i o the result of dividing a
NORM-REFERENCED TEST. particular quantity resulting from a sta-
tistical test by a measure of the error re-
c r i t e r i o n scaling a method of chang- lated to the test result. For example, a
ing the values of a predictor and a re- calculated sample mean may be divided
sponse variable to make the relationship by its corresponding STANDARD ERROR
between the two variables more closely to derive a critical ratio, which may then
resemble a straight line. In criterion scal- be used to evaluate the STATISTICAL SIG-
ing, one calculates the averages of the re- NIFICANCE of that mean. Thus, if one ob-
sponse variable for subsets of individuals tains a critical ratio of 2.0, the observed
who have similar values on the predictor mean difference is twice as large as that
variable, replaces the predictor values expected on the basis of sampling error.
with these subset averages, and then
c r i t i c a l region a range of values that
uses the substituted averages to predict
may be obtained from a statistical proce-
the response variable.
dure that would lead to rejecting a spe-
c r i t e r i o n score 1. a score on a variable cific claim about a population. More

70
 crossed-factor design

specifically, it is the portion of a PROBA- cross-correlation n. a measure of the

BILITY DISTRIBUTION containing the val- degree of association between corre-
ues for a test statistic that would result in sponding values from a series of values
rejection of a NULL HYPOTHESIS in favor for two or more variables.
of its corresponding ALTERNATIVE HY-
POTHESIS. Also called rejection re- cross-cultural research the system-
gion. Compare ACCEPTANCE REGION. atic study of human psychological pro-
cesses and behavior across multiple
c r i t i c a l value a value used to make de- cultures, involving the observation of
cisions about whether a test result is sta- similarities and differences in values,
tistically meaningful. For example, to practices, and so forth between different
evaluate the result of a T TEST to deter- societies. Cross-cultural research offers
mine whether a sample mean is signifi- many potential advantages, informing
cantly different from the hypothesized theories that accommodate both indi-
population mean, a researcher would vidual and social sources of variation,
compare the obtained test statistic to the but also involves numerous risks, nota-
values from a T DISTRIBUTION at a given ble among them the production of cul-
PROBABILITY LEVEL. If the Statistic ex- tural knowledge that is incorrect because
ceeds the critical value within that distri- of flawed methodology. Indeed, there
bution, the NULL HYPOTHESIS is rejected are a host of methodological concerns
and the result is considered significant. that go beyond monocultural studies,
Also called rejection value. See also including issues concerning translation,
CRITICAL REGION. measurement, equivalence, sampling,
data analytic techniques, and data re-
Cronbach's a l p h a a measure of the av- porting. Also called cross-cultural
erage strength of association between all method; cross-cultural study.
possible pairs of items contained within
a set of items. It is a commonly used cross-cultural testing the assessment
index of the INTERNAL CONSISTENCY of a of individuals from different cultural
test and ranges in value from 0, indicat- backgrounds. The use of instruments
ing no internal consistency, to 1, indi- that are free of bias is essential to valid
cating perfect internal consistency. Also cross-cultural testing, as it provides for
called alpha coefficient; coefficient the measurement equivalency necessary
alpha. [Lee J. Cronbach (1916-2001), to ensure outcomes have the same
U.S. psychologist] meaning across diverse populations of
interest. For example, scores on a coping
cross-case analysis an examination to questionnaire that possesses bias may be
identify similarities and differences be- a legitimate measure of coping if they
tween units of study, such as organiza- are compared within a single cultural
tions or people. Such an analysis group, whereas cross-cultural differences
generally follows an initial case-specific identified on the basis of this question-
analysis in which each unit is examined naire may be influenced by other fac-
individually to identify its unique char- tors, such as translation issues, item
acteristics. inappropriateness, or differential response
styles. See also CULTURE-FAIR TEST.
cross-classification n. the placing of
observations or individuals into classes crossed-factor design a study that in-
based on the features of two or more volves two or more conditions or treat-
variables. This is usually carried out by ments in which each level of one
means of a CONTINGENCY TABLE. condition or treatment is combined
with each level of every other condition
cross-classification table see CON- or treatment. For instance, a pharmaceu-
TINGENCY TABLE. tical treatment (Drug A vs. Drug B) may
cross-fostering

be combined with a biofeedback treat- time. Measures on one variable at one

ment (biofeedback vs. no biofeedback) time point are used to predict values on
so that each dmg is combined with each a second variable at a later time point,
type of biofeedback to form four possi- and conversely measures on the second
ble interventions or treatment plans. variable at the earlier time point are used
The individual conditions being manip- to predict the values on the first variable
ulated are known as crossed factors or at the later time point.
crossed treatments. Also called crossed
design. cross-level inference in studies in-
volving NESTING, such as students
cross-fostering n. a technique in which nested within schools, a conclusion
very young animals are placed with a drawn about the relationship between
nonbiological parent for the purpose of variables at one level of the data (e.g.,
studying environmental and genetic ef- student level) based on an analysis per-
fects on development. For example, this formed at the other level of the data
technique has been used to study the (e.g., school level). Cross-level infer-
role of genetic factors in the origins of ences are often problematic as the na-
disorders. Such a study may involve ei- ture of the relationships involved can
ther (a) having the offspring of biologi- change depending on the level at which
cal parents who do not show the they are studied.
disorder reared by adoptive parents who
do show the disorder or (b) having off- crossover design a study in which dif-
spring of parents who show the disorder ferent treatments are applied to the same
reared by parents who do not. individuals but in different sequences.
cross-lagged panel design a study of In the most basic crossover design, a
the relationships between two or more two-by-two or two-period crossover de-
variables across time in which one vari- sign, a group of participants receives
able measured at an earlier point in time Treatment A followed by Treatment B
is examined with regard to a second vari- while a second group receives Treatment
able measured at a later point in time, B followed by Treatment A. For example,
and vice versa. For example, suppose an a researcher could use such a design to
organizational researcher measures job assess the effect of attending a day ser-
satisfaction and job performance at the vice on stroke survivors, randomly
beginning of one fiscal year and at the assigning participants to one of two
beginning of a second fiscal year. Exam- groups, the first of which would attend
ining the correlations between satisfac- the service for six months and then not
tion and performance at the different attend for six months, and the second of
times provides information about their which would not attend the service for
possible causal association: If the corre- six months and then would attend for
lation between high performance at six months. As with WITHIN-SUBJECTS
Time 1 and satisfaction at Time 2 is sig- DESIGNS, the benefit of this design is the
nificantly stronger than the correlation reducfion in ERROR VARIANCE. Also
between satisfaction at Time 1 and per- called crossover study; crossover
formance at Time 2, it is potentially the trial. Compare PARALLEL-GROUPS DE-
case that those who perform better are SIGN. See also GRAECO-LATIN SQUARE;
the ones who subsequently are more sat- LATIN SQUARE.
isfied with their jobs.
crossover interaction see DISORDI-
cross-lagged regression a type of RE- NALINTERACTION.
GRESSION ANALYSIS used to model the
relationship between two or more vari- cross-product n. the set of values ob-
ables observed at two or more points in tained by multiplying each value of one

72
 cross-validation

variable (x) by each value of a second cross-sectional sampling may involve

variable (y). collecting data from individuals of vari-
ous ages or developmental levels so as to
cross-product r a t i o an ODDS RATIO study behavioral or other differences
computed from the values in a two-by- among them.
two CONTINGENCY TABLE. For example,
consider a researcher examining the oc- cross-sequential design a study in
currence of depression in stroke survi- which two or more groups of individuals
vors. He or she could calculate a cross- of different ages are directly compared
product ratio to determine the relative over a period of time. It is thus a combi-
risk of depression occurring in such indi- nation of a CROSS-SECTIONAL DESIGN
viduals as follows: (1) Multiply the num- and a LONGITUDINAL DESIGN. For exam-
ber (or percentage) of people who have ple, an investigator using a cross-
had a stroke and have been diagnosed sequential design to evaluate children's
with depression by the number of peo- mathematical skills might measure a
ple who have not had a stroke and who group of 5-year-olds and a group of 10-
have not been diagnosed with depres- year-olds at the beginning of the re-
sion; (2) multiply the number of people search and then subsequently reassess
who have had a stroke and who have not the same children every six months for
been diagnosed with depression by the the next five years. Also called cross-
number of people who have not had a sequential study.
stroke and who have been diagnosed
with depression; and (3) divide the value cross-tabulation n. the number, pro-
obtained in Step 1 by the value obtained portion, or percentage of cases that have
in Step 2. specific combinations of values on two
variables that each have multiple cate-
cross-sectional analysis the exami- gories. Such information may be pre-
nation of data that have been collected sented visually in a cross-tabulation
at a single point in time. For example, a table. For example, a cross-tabulation
researcher might conduct a cross- table could be used to show the number
sectional analysis after measuring the in- of study participants according to both
come of people in different professions their sex and marital status.
at the end of a particular year. Compare
TIME-SERIES ANALYSIS. cross-validated m u l t i p l e correla-
t i o n (symbol: R^yy) the degree of associ-
cross-sectional design a research de- afion between the results obtained from
sign in which individuals, typically of two different portions of the same data
different ages or developmental levels, sample during a CROSS-VALIDATION test.
are compared at a single point in time. The correlation expresses the relation-
An example is a study that involves a di- ship between the actual scores in the
rect comparison of 5-year-olds with 8- CROSS-VALIDATION SAMPLE and the
year-olds. Given its snapshot nature, how- predicted scores in the DERIVATION SAM-
ever, it is difficult to determine causal rela- PLE.
tionships using a cross-sectional design.
Moreover, a cross-sectional study is not cross-validation n. a procedure used to
suitable for measuring changes over time, assess the utility or stability of a statisti-
for which a LONGITUDINAL DESIGN is re- cal model. A data set is randomly di-
quired. See also CROSS-SECTIONAL ANALY- vided into two subsets, the first of which
SIS. (the DERIVATION SAMPLE) Is used to de-
velop the model and the second of
cross-sectional s a m p l i n g a sampling which (the CROSS-VALIDATION SAMPLE)
method in which scores are obtained at is used to test it. In regression analysis,
a single point in time. For example, for example, the first subset would be

73
cross-validation sample

analyzed in order to develop a REGRES- grew up in American families that could

SION EQUATION, which would then be afford to have computers and a variety
applied to the remaining subset to see of books at home. In contrast, poorer in-
how well it predicts the scores that were dividuals who emigrated to America and
actually observed. were without computers or many books
in the home might find that the
cross-validation sample in CROSS- exam had some degree of cultural test
VALIDATION, the subset of data that is
bias. See also CULTURALLY LOADED ITEM;
used to test the model or procedure de- TEST BIAS.
veloped using the DERIVATION SAMPLE.
Also called holdout sample. culture-fair test a test based on com-
crucial experiment an experiment mon human experience and considered
constructed such that its result will aid to be relatively unbiased with respect to
in determining which of two opposing special background influences. Unlike
theories has made the correct predic- some standardized intelligence assess-
tion, thus refuting the other. Also called ments, which may reflect predominantly
critical experiment. middle-class experience, a culture-fair test
is designed to apply across social lines
CTT abbreviation for CLASSICAL TEST and to permit equitable comparisons
THEORY. among people from different back-
grounds. Nonverbal, nonacademic items
cubic spline a SPLINE FUNCTION used to are used, such as matching identical
join various third-degree POLYNOMIAL forms, selecting a design that completes
segments in order to produce a smoothed a given series, or drawing human figures.
representation of the overall shape of a Studies have shown, however, that any
distribution. assessment reflects certain socioethnic
c u l t u r a l l y loaded i t e m in testing and norms in some degree and hence may
assessment, a question that is more rele- tend to favor people with certain back-
vant to the background and experience grounds rather than others. For exam-
of the individual who wrote the ques- ple, an item that included the phrase
tion than to the individual who is trying "bad rap" could be unclear, as the phrase
to answer the question, such that the re- could refer to unjust criticism or to rap
sponse may not be a fair indication of music that was either not very good or
the responder's knowledge. For exam- rather good, depending on an individ-
ple, if a question included content about ual's common use of the word "bad." See
a particular religious tradition it might also CROSS-CULTURAL TESTING.
be perceived as being a culturally loaded
c u m u l a n t n. one of a set of values that
item if the test taker were an atheist or of
describes the basic nature of a DISTRIBU-
a completely different religion.
TION. A cumulant is similar to and a log-
c u l t u r a l test bias partiality of a test in arithmic function of a MOMENT of a
favor of individuals from certain back- distribution, especially for the first few
grounds at the expense of individuals values. The first cumulant (like the first
from other backgrounds. The partiality moment) is concerned with the MEAN or
may be in the content of the items, in average of a set of numbers, the second
the format of the items, or in the very act cumulant (like the second moment) re-
of taking a test itself. For example, sup- fers to the VARIANCE or degree of spread
pose a verbal comprehension exam was in a set of numbers, and the third
delivered on a computer and incorpo- cumulant (like the third moment) indi-
rated passages, pictures, and questions cates the SKEWNESS or degree of lopsid-
drawn from American literature. The edness in a set of numbers. Cumulants
exam is likely to favor individuals who and moments may differ beyond these
 cumulative frequency polygon

initial values, and statistical methods sible values for a variable, the second
generally are concerned only with the column (labeled f, for frequency) lists
first three cumulants or moments of a the number of scores that occur at each
distribution, as well as the fourth mo- of the possible values given in the first
ment, KURTOSIS. column, and the third column (labeled
CF, f o r C U M U L A T I V E FREQUENCY) giveS
cumulant generating function the running total of each of the values in
(CGF) a formula for obtaining values the second column. For example, a
that describe the basic nature of a DISTRI- teacher administers a test and the stu-
BUTION. A cumulant generafing function dents' scores are 1 F, 2 Ds, 4 Cs, 3 Bs, and
is generally the logarithm of a MOMENT 2 As. In a cumulative frequency distribu-
GENERATING FUNCTION, producing val- tion, the first column (X) represents
ues related to the MEAN, VARIANCE, and exam scores, with F, D, C, B, and A listed
SKEWNESS of a set of variables. from the bottom to the top. In the sec-
cumulative distribution function ond column (/) are the values of 1, 2, 4,
(CDF) a formula that gives the PROBA- 3, and 2 to indicate 1 F, 2 Ds, 4 Cs, 3 Bs,
BILITY, from 0 to 1, that a variable will and 2 As. In the final column (CF) are
have a score less than or equal to a spe- running totals of the second column
cific value. When this is plotted on a from the bottom up, listing 1, 3, 7, 10,
graph, the vertical y-axis will indicate and 12 to indicate the summed total of
the probability value from 0 to 1 for each scores at each of the grades, with the
possible score of the variable listed along total number of scores listed at the top
the horizontal ;if-axis. This type of graph (i.e., there are 12 total scores in this cu-
usually shows a pattern with a curve that mulative frequency distribution).
rises from the lower left up to a peak at
the upper right, where for the last score Score Frequeticy Cumulative
the probability that a variable will have a (X) (/) frequency (CP)
value equal to or less than the highest A 12
possible score is 1. Also called distribu-
tion function. B 10

C 7
cumulative frequency (CF) a run-
ning total of how often specific values D 3
occur. Cumulative frequencies are used F I
in DESCRIPTIVE STATISTICS when listing
the number of participants who fall into
each of several categories of a variable This type of table is useful in DESCRIP-
that can be ordered from low to high. TIVE STATISTICS to depict the number of
For example, if test scores in a particular scores at or below each score level, and
classroom are 1 F, 2 Ds, 4 Cs, 3 Bs, and 2 to provide an organized display of data
As, the cumulative frequency is obtained that could also be graphed in a CUMULA-
by successively adding the number of TIVE FREQUENCY POLYGON. AISO Called
students at each score from an F to an A. cumulative distribution; cumula-
Thus, the cumulative frequency values tive frequency table.
from the lowest to the highest would be
1, 3, 7, 10, and 12 for F, D, C, B, and A, cumulative frequency polygon a
respectively. See also CUMULATIVE FRE- graphical representation of a CUMULA-
QUENCY DISTRIBUTION. TIVE FREQUENCY DISTRIBUTION, With
points along the vertical y-axis depicting
cumulative frequency flistribu- the number of scores falling at or below
t i o n a table with three columns where the particular values given along the
the first column (labeled X) lists the pos- horizontal .x-axis. The resulting points

75
cumulative hazard function

D C

Exam Score

cumulative frequency polygon

are connected with straight lines, reveal- mulative odds ratio is an extension of
ing a pattern that tends to increase or re- the single ODDS RATIO, which is the value
main constant as one moves from the formed at just one time point.
lower left to the upper right portion of
the graph (see example above). Also cumulative outcome analysis a
called cumulative frequency curve; procedure for assessing the PROPORTION
cumulative frequency diagram; cu- of occurrence of an event, usually across
mulative frequency graph. time. This procedure can be performed
as part of SURVIVAL ANALYSIS, which
cumulative h a z a r d f u n c t i o n in a studies the rate of survival or success (or
SURVIVAL ANALYSIS, a mathematical de- conversely the rate of death or failure)
scription of the accumulating number of over a period of time. For example, a re-
failures or deaths over time: It can be in- searcher may want to assess patients
terpreted as giving the probability of who remain alive at one-year intervals
death or failure at a specific time, given across afive-yearperiod, based on an ini-
survival until that time. For example, a tial sample of 100 patients. Before the
cumulative hazard function might be start of the study, the proportion is
used in a research project to examine the equal to 1.0, as all of the patients that
mnning pattern of deaths due to cancer will be in the study are now alive. At the
over a five-year period. It is an extension end of the first year, if 85 patients are
of a HAZARD FUNCTION beyond a specific still alive, the proportion would be
individual or single occurrence. 85/100 or .85. At the end of the second
year, if 62 patients are still alive, the pro-
cumulative odds ratio a WEIGHTED portion would be 62/85 = .73; and so on.
AVERAGE of a Set of values formed by Cumulative outcome analysis may also
calculating the probability of an event be used in examining LIFE TABLES.
occurring in one group over the proba-
bility of the event not occurring in the cumulative percentage a mnning
same group, then dividing this by the total of the PERCENTAGE values occur-
probability of the same event occuring ring across a set of responses. The total
in a second group over the probability of will either remain the same or increase,
the event not occurring in this group. reaching the highest value of 100% after
For example, a medical researcher may totaling all of the previous percentages.
examine the probability of cancer occur- For example, if the percentage of fresh-
rence in a treatment group versus a con- men, sophomores, juniors, and seniors
trol group every month for six months. among all of the students at a college
The researcher then finds the average of were 40%, 25%, 20%, and 15%, respec-
these values, after weighting them for tively, the cumulative percentage values
the number of individuals involved in would be 40%, 65%, 85%, and 100%
each of the monthly calculations. A cu- when summing from the percentage of
 cumulative relative frequency

freshman up to the percentage of se- running sum of the PROPORTION of

niors. times specific scores occur compared to
the total number of individual score oc-
cumulative p r o b a b i l i t y a running
currences in a set of data. For example, a
total of the chance of a score occurring
teacher may find that there was just one
at or below a certain point, where the
score of F out of a total of 12 scores, such
largest running total is a value of 1.0. For
that the first cumulative relative fre-
example, a college teacher may want to
quency distribution value was .083 (i.e.,
know the cumulative probability values
1/12). If there were two scores of D
showing the chance that a grade selected
added to the one score of F, the second
from the entire set of grades on a re-
value would be .250 (i.e., 1/12 + 2/12 =
cently given test would be an F, D, C, B,
3/12); and so on up through the values
or A. If there were 8.3% Fs, 16.7% Ds,
to a score of A, which would equal 1.0 as
33.3% Cs, 25% Bs, and 16.7% As, the cu-
all of the scores would be taken into ac-
mulative probability values would be
count at that point (i.e., 12/12 = 1.0). A
.083, .250, .583, .833, and 1.000, respec-
collection or table of cumulative relative
tively, when totaling from the F to the A
frequencies is called a cumulative rela-
values.
tive frequency distribution.
cumulative probability distribu-
tion a graphical representation of a set
Score Frequency Cumulative relative
of data that displays along the vertical y- (X) frequency
axis the chance that a case picked at ran- w
dom from that set will have a value less A 2 I.OOO
than or equal to the corresponding value B 3 .833
on the horizontal ;t-axis. When the vari-
able of interest has DISCRETE or CATE- C 4 .583
GORICAL whole-integer values, as with D 2 .250
the number of students in a class, the cu-
mulative probability distribution also F 1 .083
may be called a cumulative probability
function. This can be graphed as the intersection
of the running proportion on the verti-
cumulative record a continuous tally cal y-axis and the individual score op-
or graph to which new data are added. In tions along the horizontal ;if-axis; the
conditioning, for example, a cumulative plotted pattern is often called a cumula-
record is a graph showing the running tive relative frequency curve, cumula-
total of the number of responses over a tive relative frequency diagram, or
continuous period of time. cumulative relative frequency graph.
cumulative relative frequency a For the exam score example, propor-

Exam Score

cumulative relative frequency graph

cumulative scale

tions from 0 to 1.0 would be listed verti- cure rate model a representation of
cally and the grades F, D, C, B, and A the PROPORTION of individuals who
would be listed horizontally. When plot- have survived an illness after treatment,
ted, the running total of the proportions which includes those who have survived
of each score that occurred would have a up to a certain point but have not been
height of .083 for the score of F, on up to cured. This proportion is important in
a height of 1.0 for a score of A. Also SURVIVAL ANALYSES of complex ill-
called cumulative proportion. nesses, such as cancer. For example, it
cumulative scale see GUTTMAN SCALE. might be useful to construct a cure rate
model that estimates the number of pa-
cumulative sum chart (CuSum tients who survive cancer and appear to
chart) a graph of a group of values that no longer have the disease as well as the
are calculated by setting an initial value number who still have cancer but might
of 0 and obtaining each subsequent be expected to still be alive at specific
value by adding the previous value to times after treatment.
the difference between the current score
and the mean of all of the scores. For ex- curse o f dimensionality the problem
ample, if there were three IQ scores of that the volume of a mathematical space
92, 103, and 105, the mean is equal to increases exponentially with every new
(92 + 103 + 105)/3 = 100. Then, the ini- facet that is added, such that it becomes
tial value of the cumulative sum chart increasingly difficult or even intractable
would be set to equal 0, and the next to study spaces as the number of the
value would equal 0 + (92 - 100) = -8. variables increases.
The next value would equal -8 + (103 - curve f i t t i n g any of various statistical
100) = -5; and the last value would equal techniques for obtaining a function that
-5 + (105 - 100) = 0. graphically represents a given set of data
as closely as possible, with minimal
error. A simple example of curve fitting
occurs in LINEAR REGRESSION, where a
straight line function is said to provide
the closest approximation to the ex-
pected values of the outcome with a spe-
cific predictor and have the fewest
discrepancies between the predicted val-
ues and the actual, observed values.
c u r v i l i n e a r adj. describing an associa-
tion between variables that does not
consistently follow an increasing or de-
creasing pattern but rather changes di-
As shown, the cumulative sum values
rection after a certain point (i.e., it
are given along the vertical y-axis and the
involves a curve in the set of data
IQ scores are given along the horizontal x-
points). For example, the relationship
axis. between anxiety and achievement often
cumulative test a type of test that as- has a curvilinear pattern of increasing
sesses all of the information taught up to achievement with increasing anxiety
the point that it is taken. For example, (i.e., motivation to study) up to a certain
cumulative tests in school include an in- point when there is so much anxiety
tegration of all of the material that a that achievement tends to decrease.
teacher covered over the entire term, Thus, individuals who are not at all anx-
rather than just the most recent infor- ious and those who are extremely anx-
mation that was taught. ious would both be expected to have

78
 cycle plot

poor performance, whereas moderately SERIES of data for a specific variable. The
anxious individuals would be expected pattern can be evenly spaced (e.g., as with
to have reasonably high performance. seven-day weekly salesfigures)or irregu-
See also NONLINEAR. lar (e.g., as with shifts in sales due to
weather changes or economic factors).
curvilinear regression see NONLIN-
EAR REGRESSION. cycle plot a graph that highlights regu-
lar patterns in TIME-SERIES data by dis-
CuSum chart abbreviation for CUMU- playing a separate line for each phase
LATIVE SUM CHART.
(e.g., week, month) for every major point
cutoff score a value or criterion that is (e.g., day of the week) along the series.
held to mark the lowest point at which a For example, a professor may want to ex-
certain status or category is attained. For amine the number of student e-mails he
example, the cutoff score for passing a or she receives each week over a 14-week
course is often 60%. Similarly, the cutoff semester by charting the 14 points for
score for being considered overweight is specific numbers of e-mails received on
a body mass index of 25 to 29. Also Sundays in the first line of the graph, fol-
called cutoff point; criterion cut- lowed by the 14 points for the number of
off. e-mails received on Mondays in the sec-
ond line, and so forth. Whenfinished,the
Cuzick's t r e n d test an extension of graph displays the number of e-mails on
the WILCOXON RANK-SUM TEST: a the verfical y-axis and the days of the
nonparametric procedure used with OR- week on the horizontal x-axis, with the
DINAL DATA for estimating the pattern of specific patterns over the 14-week semes-
relationship between rankings across ter for each day clearly visible (see below).
three or more independent random
In this way, it would be more obvious
samples. For example, Cuzick's trend
on what days of the week and during
test could assess the similarity of sets of
which weeks of the semester (e.g., just
standardized achievement test scores for
before exams, after holidays) more time
applicants to private, Ivy League, and
would need to be designated for attend-
state colleges, [jack Cuzick, U.S. epidemi-
ing to student questions and feedback.
ologist and statistician]
A cycle plot is thus more revealing than
cycle n. a distinct pattern within a TIME a standard time series graph that, for

Tues Wed Thurs

Day of Week
cycle plot

79
cyclic component

this example, would list all of the days of cyclic data a set of information, espe-
the week over each of the 14 weeks (i.e., cially a TIME SERIES, in which a recurring
7 X 14 = 98 points) along the x-axis, with distinct pattern is identifiable. For exam-
the latter making it difficult to perceive ple, a psychiatrist may note that patients
daily or weekly patterns as clearly. tend to require more visits or medica-
cyclic component the portion of a tion during regular time periods around
plot of TIME-SERIES data that shows a no- holidays or the anniversary of the death
ticeable, distinct pattern, presumably of a loved one.
due to some identifiable factor. For ex-
ample, it could be important to identify cyclic v a r i a t i o n a change in the pat-
the cyclic component in car sales due to tern of data across a TIME SERIES that is
long-term fluctuations in the economy, recurring and identifiable. For example,
in addition to taking into account a sea- sales, doctor visits, and student e-mails
sonal pattern showing more predictable may show regular patterns over time
increases at the end of the year and dur- that are useful to identify. See also CY-
ing holidays. CLIC COMPONENT.

80
 Dd
d 1. see COHEN'S D. 2. see GLASS'S D. haviors such as frequency and amount
of exercise, number of calories con-
d' symbol for D PRIME.
sumed per day, number of cigarettes
D 1. symbol for DIFFERENCE SCORE. 2. see smoked per day, number of alcoholic
COOK'S DISTANCE. drinks per day, and so forth.
symbol for MAHALANOBIS DISTANCE. data analysis the process of applying
graphical, statistical, or quantitative
DAG abbreviation for DIRECTED ACYCLIC techniques to a set of observations or
GRAPH. measurements in order to summarize it
D'Agostino test 1. a statistical assess- or to find general patterns. For example,
ment of the degree of SKEWNESS (lopsid- a very basic data analysis would involve
edness) in a distribution. Its full name is calculating DESCRIPTIVE STATISTICS (e.g.,
the D'Agostino test for skewness. 2. (sym- MEAN, MEDIAN, MODE, STANDARD DEVI-
bol: K^) a statistical assessment of the ATION) and possibly graphing the obser-
degree to which a distribution departs vations with a HISTOGRAM or BAR GRAPH.
from a NORMAL DISTRIBUTION, in which
data augmentation (DA) the use of
skewness and KURTOSIS are expected to
algorithms to simulate data so as to esti-
be zero. Its full name is the D'Agostino-
mate functions that are difficult to cal-
Pearson omnibus test for normality.
culate directly from the existing data.
[Ralph B. D'Agostino (1940- ), U.S. bio-
Data augmentation has been used in
statistician]
MARKOV CHAIN MONTE CARLO METHODS
damped regression see RIDGE REGRES- and other approaches.
SION.
database n. a large, structured collec-
Danieli weights values that can be tion of information stored in retrievable
used to adjust the beginning, end, and form on a computer. For example, most
middle of a MOVING AVERAGE form of researchers keep a record of information
TIME SERIES, allowing a clearer view of collected from their studies on a com-
the nature of the trends in the data that puter to allow for easy access, manipula-
are distinct from RANDOM ERROR. tion, and analysis in testing relevant
D a r l i n g test an assessment, more for- HYPOTHESES.
mally known as the Anderson-Darling data capture the automatic collection
test, of whether a sample of data came of data, often via scanning or computer
from a particular probability distribu- devices that record information quickly
tion. It is considered a very powerful and unobtrusively. For example, infor-
test of whether data differ from a NOR- mation is retained about a person's on-
MAL DISTRIBUTION. [Donald A. Darling line purchases with a particular credit
(1915- ), U.S. mathematician] card, which in turn may be used by a
company to help determine that per-
data pl. n. (sing, datum) observations or
measurements, usually quantified and son's preferences and make suggestions
obtained in the course of research. For about other items he or she might be in-
example, a researcher may be interested terested in purchasing.
in collecting data on health-related be- data c o d i n g the process of putting in-

81
data collection

formation into understandable, usually tions, and other regularities that can be
Q U A N T I T A T I V E Or Q U A L I T A T I V E , f o r m S tO used for predictive purposes. Although a
allow analyses that can summarize the relatively new discipline, data mining
main themes that emerge or test rele- has become a widely utilized technique
vant HYPOTHESES. For example, a re- within commercial and scientific re-
searcher may collect information on a search. For example, retailers often use
20-item survey that contains questions data mining to predict the future buying
on a number of areas of substance use. trends of customers or design targeted
Based on this information, several quan- marketing strategies, whUe clinicians
titative variables can be formed assess- may use it to determine variables pre-
ing, for example, the amount of alcohol dicting hospitalization in psychological
use, the frequency of alcohol use, the disorders. Data mining incorporates
amount of hard drug use, and the fre- methods from statistics, logic, and artifi-
quency of hard drug use. cial intelligence.
data collection a systematic gathering data p o i n t a specific piece of informa-
of information for research or practical tion derived from a larger set of data. A
purposes. Examples include mail sur- data point is often formed by the inter-
veys, interviews, laboratory experi- section of two other pieces of informa-
ments, and psychological testing. tion, as in the intersection of a person's
score on one variable with his or her
data dredging the inappropriate prac- score on a second variable in a SCATTER-
tice of searching through large files of PLOT. For example, 10 people could be
information to try to confirm a precon- asked their average number of hours of
ceived HYPOTHESIS or belief without an sleep per night and the number of colds
adequate design that controls for possi- they tend to have in a year. Then, the
ble CONFOUNDS or alternate hypotheses. number of sleep hours could be recorded
Data dredging may involve selecting along the horizontal A;-axis and the
which parts of a large data set to retain in number of colds per year recorded along
order to get specific, desired results. An the vertical y-axis, resulting in 10 sepa-
extreme example might occur if a mar- rate data points falling, most likely, in a
keting researcher found that 91 out of negative pattern from the upper left por-
100 people surveyed were opposed to a tion to the bottom right portion of the
certain product and then chose to only scatterplot (i.e., the more hours of sleep,
focus on the last 10 people in order to the fewer colds).
state that 9 out of 10 people prefer this
product, when in fact there were only 9 data p o o l i n g combining the informa-
out of 100 who preferred the product. tion from two or more studies or
See DATA SNOOPING. substudies, for example, by averaging
STANDARD DEVIATIONS Or VARIANCES
data m a t r i x (symbol: X) an arrange- across groups to form a single value for
ment of data in a MATRIX, usually with use in a T TEST or ANALYSIS OF VARIANCE.
one row for each participant and one Although it can be helpful to synthesize
column for each piece of information information in this way, data pooling
gathered from the participants. A data can sometimes lead to misleading con-
matrix may be used in analyses to test clusions, as in SIMPSON'S PARADOX.
HYPOTHESES and examine patterns.
data record a set of information, often
data m i n i n g the automated (comput- relating to a single participant, that is
erized) examination of a large set of stored in a larger file or DATABASE. For
observations or measurements, particu- example, a researcher may want to focus
larly as collected in a complex database, on the specific information or data re-
in order to discover patterns, correla- cord gathered from the 10th participant,

82
 decision function

which would most likely be found in the symptom as evidence that a treatment is
10th row of the larger file. working effectively. Compare ACCELER-
ATION.
data reduction the process of reducing
a set of measurements or variables into a deception by commission see AC-
smaller, more manageable, more reli- TIVE DECEPTION.
able, or better theoretically justified set
or form. For example, a researcher may deception by omission see PASSIVE
conduct a FACTOR ANALYSIS on a set of DECEPTION.
50 items on well-being and satisfaction
to determine whether the information deception research any study in which
could be summarized more efficiently participants are deliberately misled or
on UNDERLYING DIMENSIONS of rela- not informed about the purpose of the
tionship satisfaction, degree of meaning investigation in order to avoid the possi-
in life, job satisfaction, and general bility that responses may be given to
health. meet the perceived expectations of re-
searchers. For example, a social psychol-
data screening a procedure in which ogist may use a deception experiment in
one subjects a large set of information to which participants are randomly as-
preliminary review for any of a variety of signed to either of two scenarios that
reasons: to check for accuracy, to iden- each describe job applicants in identical
tify unusual patterns (e.g., OUTLIERS terms except that one is said to be a male
that are very different from the informa- and the other is said to be female. The re-
tion from most participants) or any searcher may then assess any gender bias
missing information, to determine in the participants by asking how likely
whether the information would meet a the applicant is to be hired, instead of
statistical ASSUMPTION (e.g., of normal- directly asking participants about their
ity), to reduce the data to more manage- attitudes toward gender. See ACTIVE DE-
able dimensions, and so forth. CEPTION; DOUBLE DECEPTION; PASSIVE
DECEPTION.
data set a collection of individual but re-
lated observations or measurements decile n. one of a series of values that di-
considered as a single entity. For exam- vide a statistical distribution into 10
ple, the entire range of scores obtained equal-sized parts. Thus, the first decile is
from a class of students taking a particu- the value below which lie 10% of cases,
lar test would constitute a data set. the second decile is the value below
data snooping 1. looking for unpre- which lie 20% of cases, and so on.
dicted, post hoc effects in a body of data. decision error a conclusion that a
2. examining data before an experiment study obtained positive results when in
has been completed, which can some- fact the research did not do so in 95% of
times result in erroneous or misleading similar studies or a conclusion that a
conclusions. See also DATA DREDGING. study was not effective when it actually
d e b r i e f i n g n. the process of giving par- produced effective results in most other
ticipants in a completed research project reported research. The former conclu-
a fuller explanation of the study in sion of a false positive is referred to as a
which they participated than was possi- TYPE I ERROR, whercas the latter conclu-
ble before or during the research. sion of a false negative is called a TYPE II
ERROR.
deceleration n. a decrease in speed of
movement or rate of change. For exam- decision f u n c t i o n a procedure or set
ple, a medical researcher may be inter- of procedures that determines the con-
ested in the deceleration of an illness or clusions to be drawn or actions to be

83
decision rule

taken on the basis of certain observed will occur with the probability equal to a
data. Also called decision rule. value designated as BETA (P); if it is re-
decision rule 1. in HYPOTHESIS TEST- jected, a correct decision has been made
ING, a formal statement of the set of val- with a probability equal to 1 - p, which is
labeled as POWER.
ues of the test statistic that will lead to
rejection of the NULL HYPOTHESIS that decision theory a broad class of pre-
there is no significant effect in the study sumptions in the quantitative, social,
being examined. For example, a com- and behavioral sciences that aim to ex-
mon decision rule is to reject the null plain the process and identify optimal
hypothesis when the value of a z TEST ways of arriving at conclusions in such a
stafisfic exceeds 1.96. 2. see DECISION way that prespecified criteria are met.
FUNCTION. decision tree a diagram that uses a tree-
decision table a table that sets out the like structure of branches and nodes to
key conditions and actions involved in delineate conditions, action choices,
coming to a conclusion. In statistics, a and the further conditions these choices
widely used decision table sets out the give rise to. A decision tree starts small
conditions of there being either a true and grows in size and complexity as it
NULL HYPOTHESIS or a false null hypoth- advances from left to right. For example,
esis, the actions of rejecting the null to decide which statistical analysis to
hypothesis or retaining the null hypoth- conduct in a given case, a diagram could
esis, and the corresponding PROBABIL- begin at the left with a single node ask-
ITY associated with each action. If the ing what the focus of the analysis is to
null hypothesis is true and it is rejected, be.
a TYPE I ERROR has occurred with a prob- A branch going off to the upper right
ability equal to ALPHA (a; often .05); if it could specify a node with the focus of
is retained, a correct decision has been obtaining mean differences between
made with a probability equal to 1 - a groups, and another branch going off to
(in this case, .95). If the null hypothesis the lower right could specify a node for
is false and it is retained, a TYPE II ERROR examining relationships among vari-

dedsion tree
84
 defender's fallacy

ables. Off of each of those branches is within each group (WITHIN-GROUPS

would be additional branches indicating MEAN SQUARE). If the between-group
the number of DEPENDENT VARIABLES, variance is significantly larger than the
with still further branches leading even- within-group variance, there is some ev-
tually to nodes for different types of idence for a difference in means across
tests, depending on further conditions groups. Likewise, in STRUCTURAL EQUA-
(e.g., population information, number TION MODELING, one may break down flf
and nature of samples). Similarly, the total effects into the additive direct
branches off of the relationships among effects of independent and intervening
variables node could eventually lead to variables on a dependent variable and
choices of different analytic techniques, the indirect effect of an independent
depending on the number, nature, and variable on a dependent variable
focus of the variables. A generic depic- through the intervening variable.
tion of such a decision tree is given on
the previous page. deduction n. 1. a conclusion derived
from formal premises by a valid process
decoding n. the process of translating of reasoning. 2. the process of inferen-
coded information back into its source tial reasoning itself. Compare INDUC-
terms or symbols. Decoding is used in TION. deductive adj.
information processing, communica-
tion, and computer science, for exam- defender's f a l l a c y the error of not tak-
ple. See CODING. decode vb. ing into account CONDITIONAL PROBA-
BILITY when drawing a conclusion about
decomposable model a system or rep- the possibility of an event occurring. For
resentation that can be broken down example, suppose that at the scene of a
into meaningful subsets. Decomposable crime a specific blood type was found
models are used in many areas, includ- that has only a 1 in 10,000 chance of oc-
ing BAYESIAN methods, MARKOV CHAINS,
curring, and that the defendant is found
GRAPHIC MODELS, and STRUCTURAL to have this blood type. The law^yer for
EQUATION MODELING. For example, a the defendant could argue that as the
full structural equation model with pre- population of the city was 8,000,000,
dictions from one or more INDEPENDENT there were 800 individuals with the
VARIABLES, One or more MEDIATORS,
same blood type who could be guilty of
and one or more DEPENDENT VARIABLES
the crime. The lawyer would be commit-
can be broken down into a direct effects
ting the defender's fallacy, as he or she
portion with predictions from the inde-
pendent variable(s) to the dependent has not taken into account the fact that
variable(s) but not involving the inter- there was not' any other evidence that
vening variables; and a mediational the other 799 individuals in the city
model with predictions from the inde- with that blood type had anything to do
pendent to the intervening variables, with the crime, whereas there was other
and in turn from the intervening to the evidence (e.g., from witnesses) that the
dependent variables. defendant was at the scene of the crime.
The defender's fallacy contrasts with the
decomposition n. in statistics, the prosecutor's fallacy, which would occur
breaking down of total effects or param- if the prosecuting lawyer claimed that as
eters into their different aspects. For ex- there was only a 1 in 10,000 chance of
ample, in an ANALYSIS OF VARIANCE one having this blood type the defendant
may break down the total variance into must certainly be guilty, without taking
the portion that concems how the means into account additional evidence of oth-
vary between groups (BETWEEN-GROUPS ers who have this blood type and who
MEAN SQUARE) and the portion that con- may have been at or near the crime
cerns how much random variance there scene.

85
definitional formula

definitional formula the formal ver- rather than randomly. This includes
bal definition of a statistical concept. For simple nonrandom techniques, such as
example, the definitional formula of JUDGMENT SAMPLING and CONVENIENCE
VARIANCE states that it is the mean SAMPLING. In more sophisticated meth-
squared difference between a score and ods, such as QUOTA SAMPLING and
the mean of all of the scores. This con- STRATIFIED SAMPLING, various Steps are
trasts with the COMPUTATIONAL FOR- taken with the aim of creating a sample
MULA, which is the equation used to that is representative of the larger popu-
calculate values for the concept. Also lation.
called conceptual formula; defini-
tion formula. delta (symbol: A) n. 1. a measure of the
change in a PARAMETER. For example,
d e f i n i t i o n a l v a l i d i t y the extent to AR'^ indicates how much of the change
which the methods or approaches used in R^ (the COEFFICIENT OF MULTIPLE DE-
by a researcher are consistent with the TERMINATION) was caused or explained
significance claimed (explicitly or im- by a given step (e.g., adding a specific
plicitly) for the research. See also EXTER- variable to an analysis). 2. see GLASS'S D.
NAL VALIDITY; INTERNAL VALIDITY; E M -
PIRICAL VALIDITY; FACE VALIDITY. delta method a procedure used to ar-
rive at the approximate PROBABILITY
degrees o f freedom (symbol: df) the DISTRIBUTION of a Variable that is ex-
number of elements that are allowed to pected to have a NORMAL DISTRIBUTION.
vary in a statistical calculation, or the For example, the delta method would
number of scores minus the number of allow a researcher to find the probability
mathematical restrictions. If the MEAN of having a low, medium, or large in-
of a set of scores is fixed, then the num- come if income was expected to be nor-
ber of degrees of freedom is one less than mally distributed in the population of
the number of scores. For example, if interest. The delta method is based on
four individuals have a mean IQ of 100, the CENTRAL LIMIT THEOREM. Also called
then there are three degrees of freedom, delta technique.
because knowing three of the IQs deter-
mines the fourth IQ. d e m a n d characteristics in an experi-
ment or research project, cues that may
dehoaxing n. DEBRIEFING participants influence or bias participants' behavior,
who have been involved in DECEPTION for example, by suggesting the outcome
RESEARCH to inform them that they or response that the experimenter ex-
were misled as part of the study. This pects or desires. Such cues can distort the
may involve desensitizing participants, findings of a study. See also EXPERI-
so that their self-image is not harmed by MENTER EFFECT.
having participated in a deceptive study.
Deming-Stephan a l g o r i t h m a pro-
deletion residual in a statistical proce- cedure for deriving a LOG-LINEAR ANALY-
dure, the difference that results from SIS or model from a CONTINGENCY
deleting data based on one of the partici- TABLE of frequencies for two CATEGORI-
pants in a sample. If this difference is CAL VARIABLES. For example, this proce-
large, there is some evidence that the re- dure might be used to examine whether
moved participant was an OUTLIER, who gender and college major are indepen-
was very different from the remaining dent (the NULL HYPOTHESIS expectation)
participants. Also called likelihood re- or related (the ALTERNATIVE HYPOTHESIS
sidual. See REGRESSION DIAGNOSTICS. expectation) in a sample of data from
deliberate s a m p l i n g any method of college students. [W. Edwards Deming
selecting individuals to participate in re- (1900-1993) and Frederick F. Stephan,
search on the basis of a specific plan, U.S. statisticians]

86
 dependent samples

Demmler-Reinsch spline see density f u n c t i o n see PROBABILITY

REINSCH SPLINE. DENSITY FUNCTION.
dendrogram n. a type of treelike dia- dependence n. 1. the state of having
gram used in AGGLOMERATIVE CLUSTER- some reliance on or ASSOCIATION with
ING and DIVISIVE CLUSTERING. It lists all another entity or event, as when one
of the participants at one end and then variable is formed from another variable
directs branches out from those partici- in an analysis. For example, dependence
pants who are similar and connects would be seen if a researcher included
them with a node that represents a clus- IQ formed from mental age over actual
ter, as in the generic illustration below. age, in an analysis that already has age as
A dendrogram could be used, for exam- a variable. 2. see STATISTICAL DEPEN-
ple, to cluster individuals into various DENCE.
risk categories depending on their num- dependent events in probability the-
ber of sexual partners, their frequency of ory, EVENTS that have a relafionship such
unprotected sex, and the perceived risk that the outcome of one affects the out-
of their partners. Individuals who had come of the other. For example, overeat-
few sexual partners with little or no un- ing and being overweight are dependent
protected sex and who perceived little or events, whereas shoe size and political
no partner risk of HIV infection would party preference are most likely not.
be branched into a cluster that could be Compare INDEPENDENT EVENTS.
labeled low risk, whereas individuals
with high values on these three variables dependent groups see DEPENDENT
would branch into a high-risk cluster, SAMPLES.
with other individuals presumably clus- dependent-groups analysis of
tering into a medium-risk group. v a r i a n c e see WITHIN-SUBJECTS ANALY-
density curve see PROBABILITY CURVE. SIS OF VARIANCE.
density estimation an estimate as to dependent-groups design see WITHIN-
the incidence of a specific occurrence in SUBJECTS DESIGN.
a POPULATION based on data from a ran- dependent samples sets of data that
dom SAMPLE of that population. For ex- are related owing to their having been
ample, a researcher could make a density collected from the same group on two or
estimation about the occurrence of an more occasions (as with a pre- and post-
illness in the population at large based test), or from two or more sets of individ-
on the incidence found in a random uals who are related or otherwise closely
sample of individuals who have the ill- associated (e.g., parents and their chil-
ness. dren). Also called correlated samples;

Agglomerative

2 3
Distance Measure

Divisive
dendrogram

87
dependent-samples analysis of variance

dependent groups; related sam- change in another variable. Compare IN-

ples. Compare INDEPENDENT SAMPLES. TEGRAL.

dependent-samples analysis o f derived score a score derived from

variance see WITHIN-SUBJECTS ANALY- initial RAW DATA to make it more inter-
SIS OF VARIANCE. pretable. For example, a score on a stan-
dependent-samples design see dardized test, such as the Scholastic
WITHIN-SUBJECTS DESIGN.
Assessment Test or the Graduate Record
Examinations,, is often converted from
dependent-samples t test a T TEST an initial value to a derived score with a
used to analyze data from two or more MEAN of 500 and a STANDARD DEVIA-
sets of associated individuals or from one TION of 100.
set of individuals measured on different
occasions (e.g., pre- and posttreatment). descriptive average an approximate
Also called correlated-groups t test; estimate of the center point of a set of
correlated-samples t test; related- values, indicating where information
samples t test; repeated measures t "piles up," which is sometimes made on
test. See WITHIN-SUBJECTS DESIGN. Com- the basis of imprecise or partial data. For
pare INDEPENDENT-SAMPLES T TEST. example, a teacher could provide a de-
scriptive average (e.g., 75%) that indi-
dependent variable (DV) the out- cates how most people performed on an
come that is observed to occur or change exam.
after the occurrence or variation of the
INDEPENDENT VARIABLE in an experi- descriptive discriminant analysis
ment, or the effect that one wants to a statistical procedure that distinguishes
predict or explain in CORRELATIONAL between two or more groups of a cate-
RESEARCH. Dependent variables may or gorical outcome on the basis of several
may not be related causally to the inde- initial explanatory or PREDICTOR VARI-
pendent variable. Also called criterion ABLES. The focus is on a set of standard-
variable; effect variable; outcome ized weights linking each initial variable
variable; response variable. to the grouping variable; high weights
(in absolute value) indicate variables
derivation sample a portion of an ini- that tend to differ across the groups. For
tial data set used to explore a model, test, example, a researcher may want to study
or analysis so as to assess the VALIDITY of what differentiates those who survive
its initial results. For example, a re- cancer from those who do not by using a
searcher may want to explore the factor set of background variables, such as age,
structure of a new test in an initial deri- weight, frequency of saturated fat in the
vation sample and then verify the find- diet, and family history of cancer. The
ings afterward using CONFIRMATORY descriptive discriminant analysis may
DATA ANALYSIS on a separate CROSS- reveal negative standardized weights in
VALIDATION SAMPLE. Likewise, a re- the moderate to high range, suggesting
searcher may want to use a derivation that within the sample examined more
sample to examine the predictions pro- family history of cancer, older age,
vided by a REGRESSION EQUATION. higher weight, and more saturated fat in
derivation study research that uses a the diet are significantly linked with a
DERIVATION SAMPLE to explore a new lower chance of surviving cancer. Com-
model or area of interest, before using pare PREDICTIVE DISCRIMINANT ANALY-
VALIDATION research to verify these SIS.
findings.
descriptive hypothesis a HYPOTHESIS
derivative n. in calculus, a measure of that states an expected outcome but
the change in a variable that results from does not offer any causal explanations.

88
 design effect

For example, a teacher may form a de- removing short-term cyclical variations
scriptive hypothesis that students will in a TIME SERIES to reveal the underly-
understand at least 70% of the course ing trend. For example, a marketing
material, without delving into the spe- researcher may want to investigate
cific reasons for such understanding. whether a particular advertising strategy
increased sales, after removing the fluc-
descriptive measure a variable that
tuations in the data that have to do with
measures something essential about a
the time of year (e.g., holiday sales, sum-
data SAMPLE, without making an infer-
mer spending). Compare DETRENDING.
ence as to the population PARAMETER
for that variable. For example, a teacher desensitizing n. see DEHOAXING.
may provide a descriptive measure to a
specific class that the average grade on design n. the format of a research study,
an exam was 75%, without inferring describing how it will be conducted and
that all students outside of the class also the data collected. For example, an EX-
would understand 75% of the exam ma- PERIMENTAL DESIGN involves an INDE-
terial. PENDENT VARIABLE and at least two
groups, a treatment or EXPERIMENTAL
descriptive research an empirical in- GROUP and a CONTROL GROUP, to which
vestigation designed to test prespecified participants are randomly assigned and
hypotheses or to provide an overview of then assessed on the DEPENDENT VARI-
existing conditions and, sometimes, re- ABLE. A variety of other design types
lationships, without manipulating vari- exist, including CORRELATIONAL RE-
ables or seeking to establish cause and SEARCH, QUASI-EXPERIMENTAL DESIGNS,
effect. For example, a survey undertaken LONGITUDINAL DESIGNS, NATURAL EX-
to ascertain the political party prefer- PERIMENTS, and OBSERVATIONAL STUD-
ences of a group of voters would be a de- IES, among others.
scriptive study because it is intended
simply to identify attitudes rather than designatory scale a system in which a
systematically infer or analyze influenc- number is assigned without that num-
ing factors. ber having an actual quantitative value.
For example, social security numbers use
d e s c r i p t i v e s t a t i s t i c s procedures for a designatory scale to uniquely identify
depicting the main aspects of sample each person, but a larger social security
data, without necessarily inferring to a number does not indicate anything dif-
larger population. Descriptive statistics ferent than a smaller social security
usually include the MEAN, MEDIAN, and number.
MODE to indicate CENTRAL TENDENCY,
as well as the RANGE and STANDARD DE- design effect an adjustment, used in
VIATION that reveal how widely spread GROUP-RANDOMIZED TRIALS Or HIERAR-
the scores are within the sample. De- CHICAL LINEAR MODELS, for the depen-
scriptive statistics could also include dence in the data when groups of similar
charts and graphs such as a FREQUENCY individuals (e.g., those in particular
DISTRIBUTION or HISTOGRAM, among classrooms or hospitals) are randomly
others. Compare INFERENTIAL STATIS- assigned to treatment conditions. The
TICS. formula for the design effect takes into
account the number of individuals in
descriptive survey a questionnaire de- each cluster or group and the degree of
signed to assess a specific sample on a dependence among scores (the INTRA-
given set of items without implying or CLASS CORRELATION). STANDARD ERRORS
inferring any causal link between vari-
are adjusted with the design effect to
ables.
correct the negative BIAS that occurs
deseasonalization n. the process of when using dependent data.

89
 design matrix

design m a t r i x a grid of data whose ele- ing no difference from normal. See also
ments denote the presence or absence of NORMAL PROBABILITY PLOT.
each participant (row) in a treatment
(column) of an experimental study. detrending n. the practice of removing
a specific existing pattern from data,
determinant n. a value that represents often collected in a TIME SERIES, in order
the generalized variance in a MATRIX of to reveal other expected patterns or se-
JM numbers, with large values indicating quences in the data, DESEASONALIZ-
that the matrix numbers are very dissim- ATION is a similar process but removes
ilar and thus have more varied informa- the effect of seasonal shifts in the data.
tion. A determinant can be calculated by
multiplying the EIGENVALUES of a ma- developmental survey an assessment
trix by one another. For example, a 2 x 2 of information across time on a set of in-
CORRELATION MATRIX with two OR- dividuals. For example, a researcher may
THOGONAL (i.e., unrelated) variables wish to assess the lifestyle habits and at-
would have eigenvalues of 1 and 1, titudes of adolescents over a 10-year pe-
which would yield a determinant of 1 x riod.
1, or 1. In contrast, a correlation matrix deviance n. in statistics, a measure of
with two collinear or completely related the GOODNESS OF FIT between a smaller
variables would have eigenvalues of 1 HIERARCHICAL MODEL and a fuller model
and 0, yielding a determinant of 0. Thus, that has all of the same parameters plus
the determinant of an orthogonal set of more. The difference or deviance be-
variables is very high (i.e., 1), as the tween these models follows a CHI-
two variables provide very different in- SQUARE DISTRIBUTION, with the DE-
formation. In contrast, the determinant GREES OF FREEDOM equal to the number
of the correlation matrix with com- of parameters that are added by the
pletely collinear variables would be as fuller model. If the deviance reveals a
low as possible (i.e., 0), as the two vari- SIGNIFICANT DIFFERENCE, then the
ables are redundant. Also called gener- larger model is needed. If the deviance is
alized variance. not significant, then the smaller, more
determination coefficient see COEF- parsimonious model is retained as more
FICIENT OF DETERMINATION. appropriate.
deterministic model a mathematical deviance information criterion
function in which the outcome can be (DlC) an index that reveals the DEVI-
exactly established. In other words, the ANCE between HIERARCHICAL MODELS,
model explains all of the variance in a where one model has a larger number of
DEPENDENT VARIABLE and nO ERROR parameters than the other. It is similar to
TERM is needed. This contrasts with a AKAIKE'S INFORMATION CRITERION and
STOCHASTIC MODEL, from which a range the BAYESIAN INFORMATION CRITERION
of possible values may result. but usually easier to calculate. In all
three indices or criteria, a smaller value
deterministic process see STOCHAS- indicates a more acceptable model.
TIC PROCESS.
deviance residual an index of the con-
detrended n o r m a l plot a graph in tribution made by a single observation
which the vertical y-axis represents devi- to the DEVIANCE between two models.
ations from normal and the horizontal
Ai-axis indicates the range of values for deviant case analysis a procedure
an observation. When observations are used first to understand why specific ob-
normally distributed, the points of a servations or data points differ from
detrended normal plot will fall on a what is expected in a specific model and
straight horizontal line at zero, indicat- then to adjust the model to accommo-

90
 diagonal matrix

date most of the data. Also called nega- number of observations or cases in the
tive case analysis. sample.
deviate n. in statistics, the extent to DFFITS in REGRESSION ANALYSIS, dif-
which a score differs from a specified ference in fits: an index of the influence
value, such as the MEAN. Thus, a nega- that a particular case (/) has upon the fit-
tive deviate from the mean indicates ted value y,. It is one of several indices
how much lower than average the score that are useful in diagnosing problems
is, while a positive deviate from the in regression analysis, and shows how
mean indicates how much higher than much a predicted y (y) changes when a
average the score is. particular case is excluded from the cal-
culation of the weights in the regression
deviation n. a significant departure or equation. One rule of thumb is to delete
difference. This conceptually broad term a case if its DFFITS absolute value is
has a variety of applications in psychol- greater than 2/V(p/n), where n is the
ogy and related fields but in statistics number of observations or cases in the
refers to the arithmetical difference be- sample and p is the number of predictor
tween one of a set of values and some variables plus 1.
fixed amount, generally the MEAN of the
set or the value predicted by a model. See diachronic adj. see SYNCHRONIC.
STANDARD DEVIATION.
diagnostics pl. n. procedures for evalu-
deviational f o r m u l a an equation ating how much a model differsfromex-
that computes the average difference of pected patterns and ASSUMPTIONS. For
each score in a set of data from the MEAN example, diagnostics may reveal how
score. In other words, it is the equation much a set of data departs from assump-
to calculate STANDARD DEVIATION, tions of NORMALITY, LINEARITY, and
which is equal to the square root of the HOMOGENEITY OF VARIANCE, possibly
VARIANCE. showing skewed, curvilinear, and un-
equal variances across levels of another
deviation score the difference between variable, respectively. See also REGRES-
an observation or value x and the MEAN SION DIAGNOSTICS.
value (i.e., x - mean) in a set of data. The
sum of the deviation scores for a given diagnostic test any examination or as-
data set will equal zero, as approxi- sessment measure that may help reveal
mately half of the values will be less than the nature and source of an individual's
the mean and half will be greater than physical, mental, or behavioral problems
the mean. Also called deviation value. or anomalies. In medical research, for
example, a diagnostic test would be ex-
d f symbol for DEGREES OF FREEDOM. pected to show SENSITIVITY (i.e., cor-
rectly identifying individuals with a
DFBETAS in REGRESSION ANALYSIS, certain illness) and SPECIFICITY (i.e., cor-
di/ferences in beta values: an index that rectly identifying those who do not have
describes the influence of the ith case a specific illness).
upon the estimates of REGRESSION COEF-
FICIENTS. It is one of several indices that d i a g o n a l m a t r i x a SQUARE MATRIX in
are useful in diagnosing problems in re- which all of the values in the lower and
gression analysis, showing how much a upper triangular sections are equal to
coefficient would change if a single zero. Probably the most common diago-
given case (e.g., an OUTLIER) were nal matrix is the IDENTITY MATRIX,
dropped from the data. One mle of which has values of one on the main di-
thumb is to delete a case if its DFBETAS agonal (upper left to lower right) and
value is greater than 2/V, where n is the values of zero elsewhere.

91
diagonals model

diagonals model a statistical represen- possible answers, each of which is

tation that draws on those values that completely separate from the other. The
appear on the main diagonal (upper left answers themselves are referred to as di-
to lower right) in a SQUARE MATRIX. This chotomous responses.
kind of representation is useful if data
dichotomous variable a variable that
are believed to be ORTHOGONAL or inde-
can have only two values (typically, 0 or
pendent, such that the values in the
1) to designate membership in one of
lower and upper triangular portions of a
two possible categories, such as female
matrix are not needed as they are close
versus male or Republican versus Demo-
to zero. Also called independent diag-
crat. Also called binary variable.
onals model.
dichotomy n. in statistics, the division
diary method a technique for compil- of scores into two categories (e.g., above
ing detailed data about an individual vs. below the median). See also DICHOT-
who is being observed or studied by hav- OMOUS EVENTS; DICHOTOMOUS ITEMS;
ing the individual record his or her daily DICHOTOMOUS VARIABLE.
behavior and activities. Also called
diary survey. Dickey-Fuller test see UNIT ROOT
TEST. [David Alan Dickey and Wayne Ar-
D i e abbreviation for DEVIANCE INFOR- thur Fuller (1931- ), U.S. statisticians]
MATION CRITERION.
DIF abbreviation for DIFFERENTIAL ITEM
dichotomized variable an item or FUNCTIONING.
score that initially had a set of continu-
ous values (e.g., age) but was then sepa- difference between means in re-
rated into two possible values (e.g., search involving two or more groups,
younger and older). It may be useful to the difference between the mean score
create a dichotomized variable when on a DEPENDENT VARIABLE in One group
there are TRUNCATED DATA. and the mean score on the same variable
in another group. The mean is not ex-
dichotomous data items or scores that pected to be the same across the groups,
have two categories. For example, gen- particularly when the study involves an
der (male or female) and success (pass or EXPERIMENTAL GROUP and a CONTROL
fail) are instances of dichotomous data. GROUP. In such cases, tests are used to
dichotomous events EVENTS in which confirm that the difference between
means has STATISTICAL SIGNIFICANCE.
there are only two possible outcomes,
which are independent and mutually difference score (symbol: D) an index
exclusive of one another. For example, of dissimilarity or change between ob-
health science researchers may study servations from the same individual
whether patients have a disease or do across time, based on the measurement
not have a disease, and also whether pa- of a construct or attribute on two or
tients survive an illness or do not sur- more separate occasions. For example, it
vive. Both disease status and survival would be helpful to calculate a differ-
status are dichotomous events with just ence score for a person's weight at the
two independent and mutually exclu- beginning of a diet and exercise program
sive outcomes (i.e., one cannot fall into and the final weight six months later.
both categories). Also called change score; gain score.
dichotomous items measures that differences vs. totals plot a graph
have only two independent and mutu- used to analyze results from a 2 x 2
ally exclusive possible responses. For ex- CROSSOVER DESIGN. The difference be-
ample, true-false questions are dichoto- tween the scores from one treatment to
mous items because there are just two the second is tracked along the vertical
 differential prediction

o 0

O B
b

~1 1 1 1 1 1
4 8 t2 16 20 24
Sum

differences vs. totals plot

y-axis, and the sum of the two treatment d i f f e r e n t i a l carryover effect in

scores is tracked along the horizontal A;- a 2 X 2 CROSSOVER DESIGN in which
axis, with the order of treatments indi- all participants get both treatments at
cated by separate symbols on the graph. different times, an apparent difference
Consider the hypothetical example in outcomes across the groups who had
above. different orderings of the treatments.
If there appears to be a clear difference Differential carryover effects limit the
between sums along the A-axis in the extent of causal conclusions. See CARRY-
two different ordered treatment pairs, OVER EFFECT; DIFFERENCES VS. TOTALS
there is some evidence for a CARRYOVER PLOT.
EFFECT, which would not allow straight-
forward conclusions. In contrast, if there differential item functioning
are not clear differences along the x-axis (DIF) the circumstance in which two in-
and there are more obvious differences dividuals of similar ability do not have
along the y-axis, this would be evidence the same probability of answering a
for a TREATMENT EFFECT, which would question in a particular way. This often
be desirable. is examined to assess whether men and
women or individuals of different eth-
d i f f e r e n c i n g n. a type of adjustment nicity are likely to provide disparate an-
used to achieve STATIONARITY when ad- swers on a test. If so, the fairness of the
jacent scores are correlated in a TIME SE- test can be called into question.
RIES. Differencing helps remove the
AUTOCORRELATION SO that other trends d i f f e r e n t i a l m o r t a l i t y the differ-
in the data may be examined more ences in the death rate between various
clearly. See also DESEASONALIZATION; subgroups of a population, such as those
DETRENDING. defined by sex, age, income bracket, life-
style, and so forth.
d i f f e r e n t i a l n. in psychometric testing,
any lack of similarity in responding to d i f f e r e n t i a l p r e d i c t i o n the situation
an item observed between two catego- in which a REGRESSION EQUATION calcu-
ries of respondent. In ITEM RESPONSE lated from the total data in a SAMPLE
THEORY the term is used to indicate that yields systematically different results for
items are responded to differently by in- individual subgroups in the sample. For
dividuals with similar abilities but differ- example, an overall regression equation
ent characteristics (e.g., different gender that is useful for an entire group com-
or ethnicity). See DIFFERENTIAL ITEM prising both males and females may un-
FUNCTIONING. derestimate or overestimate scores when

93
 differential scoring

it is applied separately to the males or to in which research participants adopt a

the females of that group. different intervention from the one they
were assigned because they believe the
d i f f e r e n t i a l scoring a method of different intervention is more effective.
forming a COMPOSITE VARIABLE in When diffusion of treatments occurs,
which the individual items that contrib- the INTERNAL VALIDITY of a study is
ute to the composite are given different called into question, as it would be diffi-
i M WEIGHTS. For example, a composite cult to attribute a specific outcome to a
based on a clinical diagnostic test could specific intervention. For example, par-
use differential scoring by applying FAC- ticipants in a weight-reduction program
TOR SCORES or REGRESSION COEFFI- may be assigned to a low-carbohydrate
CIENTS to individual items, such as intervention, a low-fat intervention, a
family history, degree of symptoms, and low-calorie intervention, or a control
patient age, so as to take into account group that only receives educational
how much each of these variables is material on losing weight. If some par-
linked to a diagnostic outcome. ticipants in the control condition found
out that those in one of the other condi-
d i f f e r e n t i a l v a l i d i t y 1. the accuracy
of a battery of tests in discriminating be- tions were losing more weight and they
tween a person's subsequent success in adopted that intervention, there would
two or more different criterion tasks. 2. be diffusion of treatments.
differences in VALIDITY COEFFICIENTS
d i f f u s i o n process a specific type of
across groups. For example, the correla-
STOCHASTIC PROCESS In a MARKOV
tion between test scores and job perfor-
CHAIN.
mance may differ for males and females.
D i g g l e - K e n w a r d model f o r drop-
d i f f i c u l t y scale an index of the com-
outs a procedure for assessing the na-
plexity of a test item. It is related to an
ture of missing values and estimating
ABILITY SCALE, which refers to the level
scores for individuals who did not re-
of competence required to complete items
main in a LONGITUDINAL DESIGN, based
successfully.
on their scores while in the study. [Peter
d i f f i c u l t y value the complexity of a J. Diggle (1950- ), Australian-born Brit-
test item as measured by the percentage ish statistician; Michael G. Kenward,
of participants or students in a desig- British biostatistician]
nated class, age level, or experimental
d i g i t a l data information that is repre-
group who respond to the item cor-
sented by a series of numerical values,
rectly. See DIFFICULTY SCALE.
usually so that it can be stored in a com-
d i f f u s e comparison a procedure often puter. Digital data may also be informa-
used in META-ANALYSIS to assess whether tion from sound or light waves that is
SIGNIFICANCE LEVELS Or EFFECT SIZES are recorded as signals or numbers (e.g., a
different across studies. Values are stan- digital recording of music or a movie).
dardized across studies and compared to
d i g i t preference the practice of round-
the mean over all of the studies. If the re-
ing a number to a preferred value instead
sulting diffuse comparison score (usu-
of keeping the original value. For exam-
ally obtained via a CHI-SQUARE TEST) is
ple, health researchers may prefer to
significantly different from zero, a meta-
round the weight of study participants
analyst could conclude that the signifi- to end in a .0 or .5 to accommodate
cance levels or effect sizes showed HET- those scales that do not provide more ac-
EROGENEITY across studies and could
curate information on weight. For exam-
not be seen as producing a similar value. ple, 140.0 to 140.4 pounds in actual
d i f f u s i o n o f treatments a situation body weight may register as 140.0 on a
 directional hypothesis

scale, whereas a weight of 145.5 to 145.9 starting point. The following is a generic
may register as 145.5. Thus, the scale example.
and researchers using it can be said to
have a digit preference for weight re-
corded in decimals ending in 0 or 5.
digram-balanced adj. describing an
experimental design in which treat-
ments are arranged so as to avoid a SE-
QUENCE EFFECT. For example, a research It may be used to represent data from
study that had four possible treatment any of various statistical methodssuch
conditions (A, B, C, D) that were all as- as STRUCTURAL EQUATION MODELING,
signed to each participant would be PATH ANALYSIS, and REGRESSION ANALY-
digram-balanced if different individuals SISin which one variable predicts one
received the treatments in varying or- or more other variables but the reverse is
ders: ABCD, DCBA, ACBD, DBCA. In such not tme; that is, prediction is not al-
a case, the research design itself is known lowed in both directions (e.g., if x pre-
as a B A L A N C E D L A T I N S Q U A R E . AlSO Called dicts y, then y cannot also predict x).
diagram-balanced. See also COUNTER- Also called acyclic digraph.
BALANCING; GRAECO-LATIN SQUARE. flirected g r a p h any display in which
f l i g r a p h n. a shortened name for DI- points (nodes) of information are con-
RECTED GRAPH. nected by lines (edges) showing direc-
tion (i.e., the edges have arrows or the
dimension n. in statistics, a factor or nodes are numbered). Consider the fol-
component that is applied in measuring lowing generic example.
a set of variables. For example, some in-
telligence scales are considered to have
two important dimensions consisting of
verbal and nonverbal intelligence.
dimensionality n. the number of
DIMENSIONS applied in measuring a
constmct. In FACTOR ANALYSIS and
PRINCIPAL COMPONENTS ANALYSIS, it iS
important to assess the dimensionality
of a set of items on a scale in order to
form cohesive subscales that each de- Such graphs are used to represent a di-
scribe a similar set of items. rectional process (e.g.. Point A predicts
Point B and Point B predicts Point C).
dip test o f u n i m o d a l i t y a statistical Also called digraph. Compare UNDI-
procedure to assess whether there is RECTED GRAPH.
more than one peak or MODE in a set of
data. For example, a dip test of uni- d i r e c t i o n a l hypothesis a scientific
modality for a bell-shaped NORMAL DIS- prediction stating (a) that an effect will
TRIBUTION would be nonsignificant, as occur and (b) whether that effect will
such a distribution has only one peak. specifically increase or specifically de-
crease, depending on changes to the IN-
direct correlation see POSITIVE COR-
DEPENDENT VARIABLE. For example,
RELATION.
a directional hypothesis could predict
directed acyclic g r a p h (DAG) a DI- that depression scores will decrease fol-
RECTED GRAPH in which there is no way lowing a six-week intervention, or con-
of moving through all of the points versely that well-being will increase
(nodes) in a sequence that returns to the following a six-week intervention. Also

95
directionality problem

called directional alternative hy- method of direct observation is useful in

pothesis; one-tailed hypothesis. EVALUATION RESEARCH Or FIELD RE-
Compare NONDIRECTIONAL HYPOTHE- SEARCH. See also NATURALISTIC OBSER-
SIS. VATION.
directionality problem in COR- direct observation coding system
RELATIONAL RESEARCH, the situation in in FIELD RESEARCH, a method of provid-
which it is known that two variables are ing descriptors and quantifying results
related although it is not known which so that the specific information to be re-
is the cause and which is the effect. corded and later analyzed is explicitly
identified.
directional test a statistical test of a DI-
RECTIONAL HYPOTHESIS. Also called di- direct relationship 1. an association
rectional alternative hypothesis between two variables such that they
test; directional hypothesis test; rise and fall in value together. For exam-
one-tailed test. Compare NONDIREC- ple, the number of hours studied and the
TIONAL TEST. level of test performance form a direct
relationship in that as the number of
directly p r o p o r t i o n a l describing a
study hours increases, the level of per-
correlation between two variables in
formance also increases, and vice versa.
which each increases or decreases with
the other. This contrasts with an .in- 2. in STRUCTURAL EQUATION MODEL-
versely proportional relation, in which ING, a correlation or prediction between
one variable increases as the other de- two variables that does not involve a ME-
creases. An example of a directly propor- DIATOR or intervening variable. Also
tional relationship is weight and amount called positive relationship. Com-
of calorie intake, while an example of an pare INDIRECT RELATIONSHIP.
inversely proportional relationship is direct r e p l i c a t i o n the process of re-
amount of alcoholic drinks in a two- peating a study with different data
hour period and visual perception while under similar conditions, or of conduct-
driving. Directly proportional relation- ing several different studies with the
ships may be indexed by POSITIVE COR- same data. Direct replication is useful for
RELATIONS. establishing that the findings of the
original study are reliable (see RELIABIL-
direct model a representation in which ITY). In contrast, SYSTEMATIC REPLICA-
interest centers on how one or more TION uses a different data set and also
variables specifically correlate with one adjusts the conditions in specific ways.
or more other variables, without the in-
clusion of a MEDIATOR or intervening D i r i c h l e t d i s t r i b u t i o n a group of
variable. For example, a researcher may PROBABILITY CURVES or patterns that en-
hypothesize that an individual's health compass MULTIVARIATE extensions
background relates to his or her coping of a beta probability pattern, such that
style, which in turn may be related to the marginals of a Dirichlet distribution
the individual's sense of well-being. A are BETA DISTRIBUTIONS. A Dirichlet dis-
direct model of this research would tribution may be used in BAYESIAN sta-
focus on how health background relates tistics when estimating a PRIOR DISTRI-
to well-being without considering the BUTION. [Johan Peter Gustav Lejeune
mediator of coping style. Dirichlet (1805-1859), German mathe-
matician]
direct observation a method of col-
lecting data in which a researcher simply d i r t y data a set of observations or mea-
views or listens to the subjects of the re- surements that contains errors. For ex-
search, without asking specific ques- ample, information may have been
tions or manipulating any variables. The inaccurately reported by a participant or

96
 discourse analysis

may have been incorrectly entered into connected. For example, a sequence of
computer records. ORDINAL numbers is discontinuous be-
cause it does not have a smooth flow
disaggregation n. the process of break- from one value to the next. See also DIS-
ing down data into smaller units or sets CRETE. Compare CONTINUOUS.
of observations. For example, faculty sal-
ary data initially may show a signiflcant discontinuous f u n c t i o n a graph in
difference between men and women. which there is a sudden break or discon- W
After disaggregating the data into sepa- nection in the flow of points that are
rate levels (e.g., assistant, associate, full plotted. Compare CONTINUOUS FUNC-
professor), however, one may find that TION.
there are no significant differences in discontinuous variable a variable
salary among men and women at the as- that has distinct, DISCRETE values but no
sistant professor level but that there are precise numerical flow. For example,
differences at the full professor level. gender can be thought of as a discontin-
Thus, disaggregating the data reveals a uous variable with two possible values,
finer-tuned pattern, suggesting that rnale or female. In contrast, a CONTINU-
long-term faculty differential compensa- OUS VARIABLE involves numerically pre-
tion practices applied in the past have cise information, such as height, weight,
been maintained over time. Compare and miles per hour. Also called discrete
AGGREGATION. variable.
disattenuated correlation a CORRE- discordance n. the state or condition of
LATION COEFFICIENT that removcs the being at variance. For example, in TWIN
effects of MEASUREMENT ERROR, on the STUDIES discordance refers to dissimilar-
principle that the obtained correlation ity between a pair of twins with respect
between x and y is limited by the RELI- to a particular trait or disease. Compare
ABILITY ofeach of the variables; an unre- CONCORDANCE. discordant adj.
liable measure (e.g., x) cannot predict
another unreliable measure (e.g., y). It is discordant s i b l i n g study research
formed by dividing a correlation be- that involves family members who differ
tween X and y hy the square root of the on specific characteristics or traits. For
product of the respective reliabilities for example, researchers may collect data
X and y. For example, assume the correla- on families that have children with dif-
tion between x and y was .30, the reli- fering gender, eye color, or blood types.
ability of A; was .70, and the reliability of discourse analysis a method of study-
y was .60. The disattenuated or corrected ing verbal communications that extends
correlation would be beyond the single sentence to encom-
.30/V(.70x.60) = .46. pass conversations, narratives, and writ-
ten arguments. Discourse analysis is
Thus, removing the effect of unreliabil- particularly concerned with the ways in
ity in both X and y would result in a which a sequence of two or more sen-
medium-size correlation of .30 becom- tences can produce meanings that are
ing closer to a large (i.e., .50) correlation different from or additional to any
of .46. Compare ATTENUATION. found in the sentences considered sepa-
rately. An important source of such
disclosure o f deceptions in DECEP-
meanings is the "frame" or format ofthe
TION RESEARCH, a participant's revela-
discourse (news item, fairytale, joke,
tion of the true nature of the study, even
etc.) and a recognition of the various
though he or she was asked not to give
norms that this implies. The norms and
others any information about it.
expectations that govern conversation
discontinuous adj. intermittent or dis- are a major concern of discourse analy-

97
 discovery research

sis, as is the structure of conversational nosis could be expressed as a discrete-

language generally. time survival model that would show
discovery research studies that focus the probability of living six years post-
on uncovering new treatments or inter- diagnosis, given that the cancer patient
has survived for five years. Compare
ventions, such as pharmaceutical studies
in which investigators are interested in CONTINUOUS-TIME SURVIVAL MODEL.
J M finding new drugs to treat a particular discrete variable see DISCONTINUOUS
illness. VARIABLE. See also DISCRETE RANDOM
VARIABLE.
discrete adj. separate or distinct; often
referring to CATEGORICAL or ORDINAL f l i s c r i m i n a b i l i t y n. the quality that
data that have names or ranks as values. enables an object, variable, or person to
In contrast, CONTINUOUS data have a be readily distinguished from something
potentially infinite flow of precise nu- or someone else.
merical information. See also DISCON-
TINUOUS. d i s c r i m i n a l dispersion in an experi-
ment involving the differentiation of
discrete d i s t r i b u t i o n a delineation of stimuli, the distribution of responses
separate, distinct values. For example, around a given MEAN.
the number of children in a family can
be displayed in a discrete distribution d i s c r i m i n a n t analysis a MULTIVARI-
with values of 0, 1, 2, 3, and 4. Compare ATE method of data analysis that uses a
CONTINUOUS DISTRIBUTION. linear combination of values from a set
of quantitative variables to predict dif-
discrete measure a measure of a DIS- ferences among a set of predefined cate-
CONTINUOUS value, for example, the gories or groups of another variable. For
grade level of a student. example, a researcher might use
discrete r a n d o m variable a variable a discriminant analysis to determine
having distinct values that follow a whether several measures of personality
known pattern or probability. For exam- can differentiate those who pursue dif-
ple, the outcome of rolling two dice is a ferent majors in college. When the out-
discrete RANDOM VARIABLE: The value come variable has only two categories
could be 2, 3, 4, and so forth up to 12, (i.e., is a DICHOTOMOUS VARIABLE), dis-
with a symmetric probability curve that criminant analysis results are similar to
peaks at the value 7, which is most likely those achieved with MULTIPLE REGRES-
to occur as it has the most combinations SION as far as the overall significance be-
available to produce it (i.e., with the two tween the set of predictors and the
dice having a 1 and 6, or a 6 and 1, or a 2 grouping variable, and the magnitude
and a 5, or a 5 and a 2, or a 3 and a 4, or a and importance of each of the predic-
4 and a 3). Compare CONTINUOUS RAN- tors. If there are more than two out-
DOM VARIABLE. come categories, MULTIPLE DISCRIMI-
NANT ANALYSIS is used. Also called ca-
discrete-time M a r k o v c h a i n see nonical discriminant analysis. See
MARKOV CHAIN. also DESCRIPTIVE DISCRIMINANT ANALY-
discrete-time stochastic process SIS; PREDICTIVE DISCRIMINANT ANALY-
see STOCHASTIC PROCESS. SIS.
discrete-time s u r v i v a l model a d i s c r i m i n a n t f u n c t i o n 1. any of a
model that depicts the probability of an range of statistical techniques used to
event occurring at a certain point, given situate an item that could belong to any
events at all previously countable points. of two or more variables in the correct
For example, the number of years a can- set, with minimal probability of error. 2.
cer patient can expect to live after a diag- in DISCRIMINANT ANALYSIS more specifi-

98
 disjoint sets

cally, a linear combination of predictor MATION when comparing PROBABILITY

variables that is used to categorize items DISTRIBUTIONS.
into distinct groups. The first discri-
minant function denotes the linear d i s c r i m i n a t i o n parameter in ITEM
combination that best distinguishes RESPONSE THEORY, the STANDARD DEVI-
among the discrete groups, the second ATION of the MEASUREMENT ERROR of an
discriminant function is the linear com- item in a test, which indicates how well
bination that is the next best in distin- that item distinguishes among test tak-
guishing among the groups, and so ers on the attribute being measured.
forth. Also called canonical discri- discrimination range the range of
minant function. item DIFFICULTY VALUES, usually between
.40 and .60, that indicates an appropri-
discriminant loading a correlation ate proportion of people are getting an
between a variable and an underlying di- item correct, such that the item is mod-
mension found when trying to predict a erately differentiating among individu-
categorical outcome from a set of predic- als. If the item difficulty values are too
tor variables. It may range from -1 to +1, high, most people get the item wrong
where values near either extreme indi- and it thus is not particularly differen-
cate a variable that is able to clearly dif- tial. Conversely, a very low value indi-
ferentiate among the categories of the cates that too many people get the item
outcome. A discriminant loading is anal- correct, which also is not useful in dis-
ogous to a standardized weight in a MUL- tinguishing among test takers.
TIPLE REGRESSION analysis when the
outcome is dichotomous (i.e., has only d i s c r i m i n a t i o n value an index of
two possible categories, such as pass or how much a term or variable helps in
fail). Also called canonical loading. differentiating among two groups, dis-
tributions, or situations.
discriminant validity the degree to
which a test or measure diverges from discursive adj. referring or relating to
(i.e., does not correlate with) another the use of analytical reasoning as op-
measure whose underlying construct is posed to intuition.
conceptually unrelated to it. This is one discursive psychology the study of
of two aspects of CONSTRUCT VALIDITY, social interactions and interpersonal re-
the other being CONVERGENT VALIDITY. lationships with a focus on understand-
Also called divergent validity. ing the ways in which individuals
construct events via written, spoken, or
d i s c r i m i n a t i n g power 1. a measure symbolic communication.
of the ability of a test to distinguish be-
tween two or more groups being as- d i s h a b i t u a t i o n n. the reappearance or
sessed. 2. in DISCRIMINANT ANALYSIS, enhancement of a habituated response
the degree of accuracy with which a set (i.e., one that has been weakened follow-
of predictor variables differentiates out- ing repeated exposure to the evoking
comes into categories. stimulus) due to the presentation of a
new stimulus. Dishabituation can be in-
d i s c r i m i n a t i o n index see INDEX OF terpreted as a signal that a given stimu-
DISCRIMINATION.- lus can be discriminated from another
habituated stimulus and is a useful
d i s c r i m i n a t i o n i n f o r m a t i o n the method for investigating perception in
amount of additional evidence required nonverbal individuals or animals. Com-
to substantiate a tme ALTERNATIVE HY-
pare HABITUATION.
POTHESIS over that required to substanti-
ate a NULL HYPOTHESIS. This is sometimes disjoint sets in SET THEORY, two sets
referred to as KULLBACK-LEIBLER INFOR- that have no elements in common. The

99
dismantling study

intersection of the groups is empty, and MEAN. The larger the dispersion mea-
they are mutually exclusive. An example sure, the more spread out the scores.
of disjoint sets would be the set of all
dispersion parameter an index of the
males and the set of all females.
SKEWNESS (lopsidedness) or KURTOSIS
d i s m a n t l i n g s t u d y research that ex- (peakedness) of a distribution. The NOR-
amines whether individual components MAL DISTRIBUTION thus does not need a
of a larger protocol can be just as effec- dispersion parameter because skewness
tive and with less effort and expense. and kurtosis are not present (i.e., equal
For example, a condition traditionally zero) in such a distribution. Also called
treated by several months of therapy in scale parameter.
addition to medications could be inves- d i s p e r s i o n test an assessment of
tigated via a dismantling study to assess whether two samples have a similar
whether just the therapy, just the medi- spread of values on a variable of interest.
cation, or perhaps a reduced time i n This is useful in determining whether
therapy plus some medication would be there is evidence for H O M O G E N E I T Y OF
equally as effective as the full interven- VARIANCE, which allows for pooling or
tion. averaging variances across samples.
d i s o r d i n a l i n t e r a c t i o n in a FACTO- d i s p l a y n. the presentation of stimuli to
RIAL ANALYSIS OF VARIANCE, an INTER- any of the senses.
ACTION EFFECT in which the direction
of influence of one of the I N D E P E N D E N T
d i s s i m i l a r i t y c o e f f i c i e n t an index
VARIABLES differs depending on the of difference or distance between two
objects, variables, or samples. It may
level of a second independent variable.
be based on C O R R E L A T I O N , with lower
A disordinal interaction is indicated by a
values indicating more difference or dis-
pattern of crossed lines when plotting
tance, or VARIANCE, with higher values
the two MAIN EFFECTS from the analysis.
indicating greater difference or distance.
For example, if a researcher found that a
A dissimilarity coefficient often is used
particular treatment for dementia im-
to assess the difference between vari-
proved functioning in individuals with
ables or entities in multivariate proce-
mild and moderate degrees of the dis- dures such as CLUSTER ANALYSIS,
ease but hindered functioning in those CORRESPONDENCE ANALYSIS, MULTIDI-
with a severe degree, then a disordinal MENSIONAL SCALING, and PRINCIPAL
interaction would be present. Also called COMPONENTS ANALYSIS. Also Called dis-
crossover i n t e r a c t i o n . Compare OR- tance measure. Compare SIMILARITY
DINAL INTERACTION.
COEFFICIENT.
d i s p e r s i o n n. the degree to which a set d i s s i m i l a r i t y m a t r i x a M A T R I X used
of scores deviate from the mean. Also to indicate the distance or difference be-
called spread. See also RANGE; STANDARD tween pairs of objects, variables, or sam-
DEVIATION; VARIANCE. ples. Also called distance m a t r i x . See
d i s p e r s i o n m a t r i x see C O V A R I A N C E also DISSIMILARITY COEFFICIENT. Com-
MATRIX. pare SIMILARITY MATRIX.
d i s t a l c a u s e see U L T I M A T E C A U S E .
d i s p e r s i o n m e a s u r e an index of vari-
ability, or how dispersed a set of values is d i s t a l c o n t r o l a C O N T R O L G R O U P for
for a given variable. Common measures an experiment or research project that is
of dispersion include the RANGE, which kept physically distant from the EXPERI-
is the highest minus the lowest score, M E N T A L GROUP. A distal control group
and the S T A N D A R D D E V I A T I O N , which is may permit a better experimental com-
the average distance of scores from the parison in that there is less chance of
 divisive clustering

contamination or DIFFUSION OF TREAT- specification of the relationship. For ex-

MENT. On the other hand, a distal con- ample, it would be helpful to examine
trol group may not be as closely the distribution of scores for a college
matched to the experimental group as a exam to view the frequency of students
NEIGHBORHOOD CONTROL group or as who. achieved various percentages cor-
central to the nature of the study. Be- rect on an exam. In a NORMAL DISTRIBU-
cause of these factors a design may use TION most of the scores would fall in the
both a local and a distal control group. middle (i.e., about 70% correct or a score
of C), with fewer students achieving a D
distance n. the disparity between two
(i.e., 60-69% correct) or a B (i.e., 80-89%
values or entities. In statistics, the term
correct) and even fewer earning 59% or
often refers to the difference or DEVIA-
less (i.e., an F) or 90% to 100% (e.g., an A
TION between a RAW SCORE and a mean
score). See also FREQUENCY DISTRIBU-
score. It may also refer to MAHALANOBIS
TION; PROBABILITY DISTRIBUTION.
DISTANCE, which provides an indication
of dissimilarity between two vectors or d i s t r i b u t i o n curve a graph depicting
data sets. the possible values of a variable along
the horizontal ;t-axis and the frequency
distance f u n c t i o n a FUNCTION de- of occurrences for each value along the
scribing how disparate two points, vari- vertical y-axis. The normal BELL CURVE is
ables, or other entities are from one a common example of a distribution
another. A distance function is often curve.
used in multivariate methods such as
CLUSTER ANALYSIS. distribution-dependent test see
PARAMETRIC TEST.
distance m a t r i x see DISSIMILARITY
MATRIX. distribution-free statistics see NON-
PARAMETRIC STATISTICS.
distance measure see DISSIMILARITY
COEFFICIENT. distribution-free test see NONPARA-
METRIC TEST.
distractor n. a stimulus or an aspect of a
stimulus that is irrelevant to the task or d i s t r i b u t i o n f u n c t i o n see CUMULA-
activity being performed, used to divert TIVE DISTRIBUTION FUNCTION.
attention, prevent the occurrence of
some unwanted response, or interfere disturbance t e r m see ERROR TERM.
with some behavior. For example, in divergent v a l i d i t y see DISCRIMINANT
memory studies, the participant might VALIDITY.
be given some arithmetic problems to
solve as a distractor task between the diversity index a quantification of the
study and recall phases of an experiment variety among a species or population,
in order to minimize rehearsal of the used in fields such as biology and de-
material to be remembered. mography. The measure usually ranges
from 0 to 1, with values closer to 0 most
distributed-lag m o d e l a model in often indicating more HOMOGENEITY or
which past values of an independent similarity (i.e., less diversity) and values
variable have an ongoing effect on cur- closer to 1 most often indicating high
rent values of a dependent variable. HETEROGENEITY.
d i s t r i b u t i o n n. the relation between flivisive clustering a form of CLUSTER
the values that a variable may take and ANALYSIS that starts with all enfities,
the relative number of cases taking on persons, or variables in a single data set
each value. A distribution may be simply and then divides them into progres-
an empirical description of that relation- sively smaller groups that have similar
ship or a mathematical (probabilistic) characteristics within groups and differ-

101
domain

ent characteristics across groups until matrix could yield information on

each entity is in its own single cluster. which teams appeared to dominate over
Divisive clustering can be contrasted teams that had won over other teams
with AGGLOMERATIVE CLUSTERING, a (i.e., demonsttating two-step dominance).
type of HIERARCHICAL CLUSTERING in Dominance matrices frequently are used
which each entity starts out in its own in GAME THEORY research.
group and then groups are collapsed to-
gether if there is high similarity between dominance statistic an index that
them. Also called top-down (hierar- quantifies the number of cases in one
chical) clustering. group that outscore cases in another
group.
d o m a i n n. in SET THEORY, the set of ele-
ments over which a function is defined. Doolittle method a procedure for fac-
toring (decomposing) a SQUARE MATRIX
domain-referenced test an assess-
into the product of two other matrices:
ment that covers a specific area of study
a lower TRIANGULAR MATRIX with values
such that a score will reveal how much
of 1 along the diagonal and an upper tri-
of this area has been mastered. This
angular matrix. This procedure is similar
is similar to a CRITERION-REFERENCED
to the CHOLESKY FACTORIZATION
TEST, in which a content area is mapped
method, which yields a lower triangular
and scores reflect whether a particular
matrix multiplied by its TRANSPOSE.
standard or criterion has been achieved.
These kinds of tests are contrasted with D-optimal design a research design in
NORM-REFERENCED TESTS, in which
which only those conditions that create
scores indicate how well a test taker per- the least COVARIANCE among the pa-
formed on the items relative to others rameters are included, such that a full
who took the test. Thus, if an individual design of all variable levels is not imple-
got 90% of the items correct in a do- mented. To identify a D-optimal design,
main-referenced or criterion-referenced the number of levels and their arrange-
test, this would be a high score indica- ments are modified to determine how
tive of his or her deep knowledge and RELIABILITY changes with different
understanding of the content covered in kinds of designs. If two designs produce
the test. essentially the same reliability, the re-
d o m a i n s a m p l i n g the process of se- searcher can choose the one that is logis-
lecting the content of a test from a spe- tically easier or is less costly to conduct.
cific content area. For example, most Alternatively, for a fixed cost a re-
teacher-made tests use domain sampling searcher can identify the design that
to create items from the material cov- maximizes reliability. See also A-OPTI-
ered during a particular set of lectures. MAL DESIGN; E-OPTIMAL DESIGN.

d o m i n a n c e m a t r i x a SQUARE MATRIX D o r a n estimator a procedure to fill in

that maps the performance of a set of en-
tities by listing a 1 if a row entity exceeds
or wins over the column entity and a 0
otherwise. For example, a dominance
matrix might be used to rank college
basketball teams according to which col-
lege won over which other colleges.
Rankings could be formed by adding up
the row values in the dominance matrix
to see which college team had the most
wins. An examination of the body of the
missing data for a TIME-SERIES ANALYSIS
that is useful when there are more data
missing at the beginning of a series than
at the end. For example, assume a partic-
ular company only provides quarterly
sales data during the first half of the year
and both quarterly and monthly sales
data in the latter half of the year. One
could use the Doran estimator to guess
the missing monthly sales values by
using only the quarterly sales informa-
 double-dummy technique

tion. [Howard E. Doran, U.S. econo- pants, and it uses half of the participants
metrician] needed for a double BLIND between-
groups study in which separate groups
dose-ranging trial a clinical trial in
are randomly assigned to different treat-
which patients are randomly assigned to
ments. A possible problem with the
treatment groups to receive one of sev-
crossover facet of the design is that the
eral different dosages of a drug. A dose-
effects of an initial treatment may not
ranging trial provides information on
fully be gone by the time the second
how much medicine is needed for it to
treatment is administered, thus causing
be effective. Also called dose-ranging
potential bias in the interpretation of
study.
which treatment is causing the most ef-
dot p l o t see SCATTERPLOT. fective outcome. See also CROSSOVER
DESIGN.
double bar g r a p h see DUAL BAR
GRAPH. double bootstrapping a procedure in
which samples are randomly drawn,
double-barreled question an item with replacement, from an initial data
on a survey that simultaneously asks set and their PARAMETERS and STAN-
about two kinds of information instead DARD ERRORS estimated and averaged
of just one, thus making the responses across the set of samples. This is then fol-
difficult to interpret. An example of a lowed by another random sampling of
double-barreled question is "How often the data, again with replacement, after
do you use a condom after drinking alco- which parameter estimates and standard
hol?" If a respondent said very often, a errors are obtained a second time. Dou-
researcher does not know whether this ble bootstrapping usually is less biased
refers to the frequency of condom use, than using a single set of BOOTSTRAP-
the frequency of drinking alcohol, or PING samples; additional procedures
both. It would be best to ask two sepa- often are applied between the two boot-
rate questions to obtain this informa- strapping procedures in order to further
tion (e.g., "How often do you use a reduce BIAS.
condom?" and "How often do you drink
alcohol?"), with each question having double cross-validation a procedure
its own set of response choices (e.g., in which a sample is split into two por-
"Never" up to "Quite Often"). tions or subsamples and a statistical
model (e.g., a MULTIPLE REGRESSION
double b l i n d see BLIND. EQUATION) is estimated from the first
double-blind crossover t r i a l a type portion and verified with the second
of clinical trial in which participants re- portion. The process is then reversed: An
ceive two different treatments and nei- estimate of the model is obtained using
ther the researchers nor the participants the second subsample and verified using
are aware of which treatments are ad- the first subsample. If results from botfi
ministered to which individuals at what validations are similar, there is strong
time. Participants are randomly assigned evidence that the model is robust and
to a particular order of treatments generalizable across samples that are
(Treatment A followed by Treatment B, similar to those used in the two sets of
or Treatment B followed by Treatment analyses. See CROSS-VALIDATION.
A); between treatments they go through
a brief WASH-OUT PERIOD to eliminate double deception in DECEPTION RE-
the effect of the initial treatment before SEARCH, an apparent debriefing after the
the second treatment is administered. experiment, which is itself a further
This kind of design should neutralize po- ploy.
tential bias from researchers or partici- double-dummy technique a re-

103
double exponential distribution

search design used when two treatments initial start date of the study would have
are dissimilar enough for participants to to be substituted instead. Similarly, the
know which treatment they are receiv- quit date for those who had not yet quit
ing just by its form (e.g., a comparison of smoking by the conclusion of the study
tablet and capsule medication types). In also would not be known, and the end
the double-dummy technique, each par- date of the study would have to be sub-
ticipant is randomly assigned to at least stituted instead. Thus, both ends of the
one inactive PLACEBO (i.e., dummy) distribution of quit dates would be cen-
condition plus a regular treatment. sored. See CENSORED DATA.
Thus, with a double-dummy technique
participants are randomly assigned to doubly m u l t i v a r i a t e describing a de-
one of the following three conditions: sign in which two sets of multiple vari-
(1) Treatment A (e.g., active tablet) plus ables are assessed. For example, assume a
Placebo B (e.g., inactive capsule); (2) Pla- researcher evaluates several reading per-
cebo A (inactive tablet) plus Treatment B formance outcomes (e.g., number of
(active capsule); or (3) Placebo A plus books read, difficulty level of the books,
Placebo B, which serves as a CONTROL. vocabulary level) for a set of grade-
school students. Further, imagine the re-
double exponential d i s t r i b u t i o n a searcher assesses the students at the be-
graphed curve that rises very quickly to a ginning, middle, and end of the school
peak and then falls very quickly to the year. In assessing multiple reading out-
horizontal axis, forming a symmetrical comes at multiple time points, the re-
shape with the appearance of a witch's searcher would be using a doubly
hat. The pattern is similar to a normal multivariate design.
BELL CURVE except that it is not as full at
the peak and tails. Also called Laplace doubly stochastic 1. describing a
distribution. SQUARE MATRIX of ttonnegative num-
bers in which all of the rows sum to 1
double masked see BLIND. and all of the columns sum to 1. 2. de-
scribing a model in which observations
double s a m p l i n g a research process in of a RANDOM VARIABLE are modeled in
which data can be sampled a second two stages, in the second of which one
time if the results from a first statistical or more of the parameters used in the
test just missed significance due to low first stage are themselves treated as ran-
POWER. Because this process can result dom (STOCHASTIC) variables. Doubly
in a higher TYPE I ERROR rate, it should stochastic models are used to model pro-
be used only when needed to conserve cesses with a random probability pat-
costs and maximize power. tern, such as TIME SERIES.
doubly censored data data in which d p r i m e (symbol: d') a measure of an in-
exact measurements are not known for dividual's ability to detect signals; more
events at both ends of a distribution. For specifically, a measure of sensitivity or
example, assume a health researcher ex- discriminability derived from SIGNAL
amined the time at which a group of DETECTION THEORY that is unaffected by
individuals quit smoking cigarettes. Fur- response biases. It is the difference (in
ther, consider that there was a subgroup standard deviation units) between the
who had already quit smoking at the means of the noise and signal + noise
time of the initial assessment, as well as a distributions. A value ofrf'= 3 is close to
subgroup who had not yet quit smoking perfect performance; a value of d' = 0 is
at the end of the final assessment. Thus, chance ("guessing") performance.
the exact quit dates for those who quit
before the study began would not be d r i f t n. 1. a reduction in variation in ge-
known, and for purposes of analysis the netic traits that can occur when sam-
 dual scaling

pling from continually smaller groups, of intelligence) was dependable across a

such that some traits ultimately become variety of situations (e.g., geographical
excluded from possibility. 2. a reduction areas, age groups, content areas).
in the rehability of technical instm-
ments or the accuracy of observers over d u a l bar g r a p h a BAR GRAPH that de-
time. See INSTRUMENT DRIFT; OBSERVER picts thefrequenciesfor two CATEGORI-
DRIFT. CAL VARIABLES. One categorical variable
is represented by a series of possible re-
d r u n k a r d ' s w a l k any process in sponses (e.g., pizza choice: cheese, vegeta-
which a person or thing proceeds by ran- ble, or meat) listed across the horizontal
dom steps but still arrives at a specified ;ic-axis and the other categorical variable
point or goal, much as an intoxicated (e.g., gender: male or female) is depicted
person may choose randomly among by drawing separate rectangles that repre-
various streets along the way and still ar- sent thefrequencyor percentage, shown
rive home. It is a type of RANDOM WALK, along the vertical y-axis, associated with
that is, a series of haphazard events. For each of the responses on thefirstvariable.
example, a student who did not study Thus, if 30% of men prefer cheese
for an exam could be said to follow a pizza, 20% prefer vegetable pizza, and
random walk by unsystematically guess- 50% prefer meat pizza, three rectangles
ing the answers for each question. If the that extend to 30, 20, and 50, respec-
student, however, happened to pass the tively, on the vertical axis would be
exam despite randomly guessing at each shown for men. Similarly, the percent-
juncture, the process would be a dmnk- ages for women would be depicted by
ard's walk. a second set of rectangular lines above
DS abbreviation for DUAL SCALING. the three pizza choices, as shown in the
illustration below. Also called double
D study decision research performed bar graph.
upon variance components identified in
an initial G STUDY. A D study seeks to de- d u a l scaling (DS) a method of assessing
termine the number of conditions that relationships and patterns, whether lin-
need to be considered to maintain a rea- ear or nonlinear, among CATEGORICAL
sonable level of RELIABILITY for relative DATA that involves assigning scale val-
(norm-referenced) or absolute (criterion- ues to the elements of the data set. For
referenced) decisions about a measure. example, DS may be useful when assess-
For example, a D study could be con- ing the relationship between achieve-
ducted to analyze how many different ment motivation and performance.
conditions would need to be tested be- When achievement motivation is low or
fore stating that a specific measure (e.g., high, performance is expected to be low;

0)
o
C 40

Male
Pizza Choice

dual bar graph

105
Duhem-Quine thesis

however, when achievement moti- dummy variables. Also called indica-

vation is moderate, performance is ex- tor variable.
pected to be high. This nonlinear rela-
tionship between the two variables can d u m m y variable c o d i n g a method of
be analyzed more effectively with DS assigning numerical values to a CATE-
than with conventional statistical meth- GORICAL VARIABLE in such a way that
ods (e.g., correlation, regression), as the the variable reflects class membership.
latter require assumptions of LINEARITY The values of 0 and 1 often are used,
and would reveal no apparent associa- with 0 typically representing nonmem-
tion between the variables. Also called bership and 1 typically representing
optimal scaling. membership. Compare CONTRAST COD-
ING; EFFECT CODING.
Duhem-Quine thesis the proposition
that an individual hypothesis cannot be Duncan's multiple range test
empirically tested in isolation, because (Duncan's MRT) a MULTIPLE COMPAR-
any test involves assuming the tmth of ISONS test used to follow up on an ANAL-
a larger body of principles to which YSIS OF VARIANCE that yields significant
the hypothesis belongs. [Pierre Duhem results. After obtaining a significant F
(1861-1916), French physicist and math- RATIO, a researcher could conclude that
ematician; Willard Van Orman Quine at least one pair of samples is signifi-
(1908-2000), U.S. philosopher and logi- cantly different on the DEPENDENT VARI-
cian] ABLE and then conduct a Duncan
multiple range test on pairs of samples
d u m m y variable in REGRESSION ANAL- in order to identify that specific pair.
YSIS, a numerical variable that is created The overall TYPE I ERROR rate would be
to represent a QUALITATIVE fact, which protected, which would not be the case
is done by giving a variable a value of 1 if separate T TESTS were conducted on all
or 0 to indicate the presence or absence possible pairs. Other types of procedures
of a CATEGORICAL trait. A dummy vari- for following up a significant analysis of
able usually represents a DICHOTOMOUS variance include the SCHEFFE TEST and
VARIABLE, that is, one that can have TUKEY'S HONESTLY SIGNIFICANT DIFFER-
only two categories. For example, gen- ENCE TEST, among others. [David B.
der could be coded as a 1 to represent fe- Duncan (1917-2006), U.S. biostatistician]
male and a 0 to represent male. When a
measure has more than two categories, a D u n n - B o n f e r r o n i procedure a sta-
researcher creates / c - l dummy variables tistical method for assessing whether
to represent the total number of differ- multiple pairs of samples are signifi-
ent categories, k, with one of the catego- cantly different from each other, while
ries indicated by 0. Thus, to code a protecting the overall TYPE I ERROR rate
political party as either Democrat, Re- by dividing the alpha or SIGNIFICANCE
publican, or Independent/Other, the LEVEL by the number of comparisons
first dummy variable would be labeled made. For example, if a researcher wants
"Democrat" and allot a 1 to each person to investigate whether three samples are
who is a Democrat and a 0 to each per- significantly different using an alpha of
son who is not. The second dummy vari- .05, three T TESTS could be conducted
able would be labeled "Republican" and (i.e., between groups 1 and 2, 2 and 3,
allot a 1 to each person who is a Republi- and 1 and 3) using an alpha level of
can and a 0 to each person who is not. .05/3 = .0167 for each comparison; this
Those who are Independent/Other in approach would hold the overall Type I
political party affiliation would not need error rate at .05. Also called Bonferroni
a separate variable as they would be de- t test; Dunn's multiple comparison
noted by 0 on each of the two other test. See also BONFERRONI INEQUALITY.
 DV

[O. J. Dunn; Carlo Emilio Bonferroni Dunn's multiple comparison test

(1892-1960), Italian mathematician] see DUNN-BONFERRONI PROCEDURE.
Dunnett's m u l t i p l e comparison d u r a t i o n analysis see SURVIVAL ANALY-
test a statistical method for assessing SIS.
whether a CONTROL GROUP is signifi-
d u r a t i o n recording the process of
cantly different from each of several
tracking and measuring the amount of
treatment groups. It is similar to the
time an individual engages in a specific
DUNN-BONFERRONI PROCEDURE but not
behavior (e.g., completing homework
as stringent in that it does not require
assignments, interacting socially).
the SIGNIFICANCE LEVEL per comparison
to be as small. The index of difference D u r b i n - W a t s o n test a method for as-
that is obtained from this test is called sessing AUTOCORRELATION in a TIME SE-
Dunnett's t. [Charies W. Dunnett (1921- RIES (i.e., of determining the degree of
2007), Canadian statistician] association between values separated by
a given time lag). It is properly called the
D u n n - S i d a k procedure a statistical Durbin-Watson d test so as to distin-
technique that uses MULTIPLE COMPARI- guish it from the Durbin-Watson h test,
SONS to assess whether three or more
a further refined method for producing
samples are significantly different. More an unbiased estimate of autocorrelation
powerful (requiring a less stringent SIG- that is appropriate for use with large
NIFICANCE LEVEL) than the D U N N -
samples. Oames Durbin (1923-2012),
BONFERRONI PROCEDURE, it assumes that
British statistician; Geoffrey Watson
each of the comparison tests is inde- (1921-1998), Australian stafistician]
pendent of all the others. Also called
Dunn-Sidak method; Dunn-Sidak DV abbreviation for DEPENDENT VARIA-
test. [O. J. Dunn; Zbynek Sidak] BLE.

107
 Ee
e 1. abbreviation for EXPONENT. 2. in university and then find the association
measurement theory, abbreviation for between these sets of means. This would
ERROR. provide information at a departmental,
instead of an individual, level.
E symbol for EXPECTED VALUE.
ecological f a l l a c y a mistaken conclu-
Eberhardt's statistic an index of ran- sion drawn about individuals based on
domness in a spatial distribution. If it is findings from groups to which they be-
significant, a researcher may conclude long. For example, if a university admin-
that the results are not due to random- istrator found that the ECOLOGICAL
ness but rather have a meaningful pat- CORRELATION between faculty salary
tern. and number of publications at the de-
ECM a l g o r i t h m abbreviafion for EXPEC- partmental level was strong and positive
TATION-CONDITIONAL MAXIMIZATION AL-
(e.g., r = .60), it would be an ecological
GORITHM. fallacy to assume that for any particular
faculty member the correlation would
ECME a l g o r i t h m abbreviation for EX- be the same.
PECTATION-CONDITIONAL MAXIMIZATION
EITHER ALGORITHM. ecological inference an assumption
about individuals or subsets of individu-
ecological assessment the gathering als on the basis of aggregate data about
of observations in various environments the larger group. For example, a re-
to examine whether individuals or enti- searcher may wish to make an ecological
ties behave differently depending on the inference about the education of women
surroundings. For example, a teacher at from ethnic minorities on the basis of
a child-development center may investi- data from a report that provides infor-
gate whether children demonstrate mation by gender or ethnicity but not
more or less problem behavior in various both together. Various strategies have
settings at the center. Findings from the been suggested for making vahd ecologi-
ecological assessment may offer ideas on cal inferences from summary statistics
how to intervene to improve the overall (i.e., inferences that do not commit the
behavior of the children in multiple set- ECOLOGICAL FALLACY).
tings.
ecological momentary assessment
ecological correlation a correlation (EMA) the process of examining the be-
that is calculated between group means havior of individuals at random, multi-
instead of between individual scores, ple time points to get a clearer picture of
thereby showing the relationship be- how they behave in various real-world
tween variables at the group level. For settings. For example, a researcher may
example, a university administrator may conduct an EMA on individuals who are
want to assess the relationship between trying to quit smoking by contacting
faculty salary and productivity as mea- them randomly throughout a week to
sured by the number of publications. An inquire about urges to smoke, mood,
ecological correlation would find the av- other people who are interacting with
erage salary and average number of pub- them, and other relevant circumstances
lications for each department at the that may be occurring at these times.

108
 effective sample size

ecological v a l i d i t y the degree to effect coefficient a value indicating

which results obtained from research or the overall influence of one variable
experiments are representative of condi- upon another, often used in statistical
tions in the wider world. For example, methods such as PATH ANALYSIS or HIER-
psychological research carried out exclu- ARCHICAL LINEAR MODELING.
sively among university students might
have a low ecological validity when ap- effect i n d i c a t o r in STRUCTURAL EQUA-
plied to the population as a whole. Eco- TION MODELING, a measure used to rep-
logical validity may be threatened by resent an occurrence that is assumed to
EXPERIMENTER BIAS Or by naive sam- be the effect of a LATENT VARIABLE.
pling strategies that produce an unrepre- Compare CAUSAL INDICATOR.
sentative selection of participants. See effectiveness research reviewing past
also GENERALIZABILITY. studies or conducting new studies in
E D A abbreviation for EXPLORATORY order to evaluate how well current and
DATA ANALYSIS. new treatments work, with the aim of
improving the care and treatment of pa-
Edgeworth expansion a method of tients. The term is often used synony-
reframing the functional specification of mously with clinical utility research,
a distribution so that it has properties although some have suggested clinical
similar to a more common pattern, such utility research is a broader term that en-
as the NORMAL DISTRIBUTION. [Francis compasses studies of additional health-
Ysidro Edgeworth (1845-1926), Irish related topics, such as access to care.
economist] Compare EFFICACY RESEARCH.
EFA abbreviation for EXPLORATORY FAC- effective range a measure of the dis-
TOR ANALYSIS. tance between the highest score and the
effect n. 1. in ANALYSIS OF VARIANCE, a lowest score in a set of values that does
statistically significant relationship be- not take into account any extreme OUT-
tween variables, such that one variable is LIER scores. For example, assume an in-
held to be an outcome of another (or dividual kept track of how much he or
some combination of others). See MAIN she exercised over a 12-week period with
EFFECT; INTERACTION EFFECT. 2. short the number of hours per week as 3, 2, 0,
for EFFECT SIZE. 3, 2, 3, 2.5, 3, 2, 3, 2.5, and 3. The actual
range of scores would be the largest
effect c o d i n g in REGRESSION ANALYSIS, number of exercise hours minus the
a procedure in which values of 1 or -1 lowest number: 3 - 0 = 3 hours a week.
are assigned to represent the categories However, the one score of zero is an out-
o f a D I C H O T O M O U S V A R I A B L E (or 1, 0,
lier that may have occurred when
and -1 for a trichotomous variable). The this person was sick with the flu and
results obtained for the group indicated could not exercise. In that case, one
by values of 1 are then interpreted rela- would ignore the outlier score and
tive to the larger group comprising all recompute the effective range: 3 - 2 = 1
participants. For example, a researcher hour a week.
could assign values of 1 to individuals
in an EXPERIMENTAL GROUP and values effective sample size 1. the number
of -1 to those in a CONTROL GROUP. He of participants needed for a study to be
or she could conclude that there is a as valid as if participants were randomly
treatment effect if the experimental group sampled from the population, when
has a mean different from the GRAND such random collection of data is not
MEAN across all of the individuals. Com- possible (e.g., as with households or
pare CONTRAST CODING; DUMMY VARI- classrooms). 2. in POWER ANALYSIS, the
ABLE CODING. number of units required in a sample in

109
effect modifier

order to achieve a particular level of sta- sample to estimate a particular parame-

tistical POWER. ter. It is a measure of the optimality of an
effect m o d i f i e r a variable that changes estimator when comparing various sta-
tistical procedures or experimental de-
the way two other variables are related.
signs, efficient adj.
See MODERATOR.
effects analysis an assessment of e f f i c i e n t estimator an unbiased ESTI-
whether an intervention worked or not, MATOR that achieves minimum variance
particularly when there are more than for all possible values of a parameter. An
two groups involved. efficient estimator is sometimes referred
to as a minimum variance unbiased esti-
effect size any of various measures of mator (MVUE). The sample MEAN, for
the magnitude or meaningfulness of a example, is an efficient estimator of the
relationship between two variables. For population mean.
example, COHEN'S D shows the number
of STANDARD DEVIATION units between eigenvalue (symbol: X) n. a numerical
two means. Often, effect sizes are inter- index, commonly used in FACTOR ANAL-
preted as indicating the practical signifi- YSIS and PRINCIPAL COMPONENTS ANAL-
cance of a research finding. Additionally, YSIS, that indicates the portion of the
in META-ANALYSES, they allow for the total variance among several correlated
computation of summary statistics that variables that is accounted for by a more
apply to all the studies considered as a basic, underlying variable or construct.
whole. See also STATISTICAL SIGNIFI- An eigenvalue may be computed as the
CANCE. sum of the squared FACTOR LOADINGS
for all the variables. Eigenvalues are of
effect-size correlation (symbol: r^ffect central importance in linear algebra (i.e.,
size) an index of the association between matrix algebra). Also called character-
an INDEPENDENT VARIABLE and scores istic root; characteristic value; la-
on the DEPENDENT VARIABLE without re- tent root.
moving any other sources of variation in
the data. An effect-size correlation has eigenvector n. in linear algebra, any
values from -1 to +1, with absolute val- vector in a p-dimensional space that is
ues of .1, .3, and .5 representing small, associated with a given TRANSFORMA-
medium, and large relationships, respec- TION and is left invariant (except for
tively. stretching or shrinking) by that transfor-
mation. Eigenvectors are of basic impor-
effect variable see DEPENDENT VARI- tance in MULTIVARIATE statistics; their
ABLE. values are applied to variables to form a
efficacy research an empirical study, LINEAR COMBINATION that synthesizes
such as a RANDOMIZED CLINICAL TRIAL, much of the variance in a set of vari-
that examines whether a specific treat- ables. Also called characteristic vec-
ment or approach works when com- tor; latent vector.
pared to outcomes in a PLACEBO elaborative v a l i d i t y the extent to
CONTROL GROUP. Efficacy research is which a measure reveals useful informa-
sometimes contrasted with EFFECTIVE- tion about situations similar to those in
NESS RESEARCH, which uses NATURALIS- which it has been used. For example, a
TIC OBSERVATION to assess whether an measure of depression may be said to
approach or treatment is working. have elaborative validity if it also is re-
efficacy subset analysis see TREAT- lated to the extent of a person's hope-
MENT-RECEIVED ANALYSIS. lessness or suicidal ideation.
efficiency n. the degree to which an ES- element n. a member of a set, class, or
TiMATOR uses all the information in a group. For example, in the set of all PhD-

110
 empirically keyed test

conferring American universities, any correct is estimated from the observed

one of those American universities that data. This contrasts with conventional
confer the PhD is an element of the set. Bayesian methods, in which the proba-
See also SET THEORY. ble truth of the hypothesis (the PRIOR
DISTRIBUTION) is estimated before see-
elementary event in probability the-
ing the data. Also called empirical
ory, the fundamental outcome of an ex-
Bayes method; empirical Bayes
periment of chance. For example, in
procedure.
selecting an individual from a list of eli-
gible job candidates, the individuals e m p i r i c a l classification the process
who might be drawn from the list are the of grouping entities into categories
elementary events of the experiment. based on shared characteristics using
Also called simple event. statistical methods, such as CLUSTER
ANALYSIS or MULTIDIMENSIONAL SCAL-
elliptically symmetric distribu-
t i o n a DISTRIBUTION in which KURTOSIS ING. For example, a researcher could sug-
is permitted as long as it is equal across gest an empirical classification of sexual
variables. The NORMAL DISTRIBUTION is risk categories by clustering individuals
a type of elliptically symmetric distribu- based on their frequency of unprotected
tion, as kurtosis values are equal to zero sex, their number of sexual partners, and
for all variables. the perceived risk of their sexual part-
ners. It would be expected that the low-
E M A abbreviation for ECOLOGICAL MO- est risk category in the empirical
MENTARY ASSESSMENT. classification would be people who have
E M a l g o r i t h m expectation-maximiza- no or low frequency of unprotected sex,
tion algorithm: a statistical procedure one or few sexual partners, and no or
used to find PARAMETER estimates in low perceived risk of their sexual part-
models that incorporate LATENT VARI- ners.
ABLES. e m p i r i c a l construct see CONSTRUCT.
emics n. see ETICS. e m p i r i c a l curve a line on a graph that
emotional bias the tendency to ignore connects points formed from the values
reason when experiencing strong posi- of relevant variables.
tive or negative feelings, thus impairing e m p i r i c a l d i s t r i b u t i o n the cumula-
one's ability to accurately understand tive pattern of frequencies observed for a
and evaluate facts and to act accord-
variable in a sample. Compare THEORET-
ingly.
ICAL DISTRIBUTION.
e m p i r i c a l adj. 1. derived from or denot-
e m p i r i c a l l a w a law that is based on
ing experimentation or systematic ob-
facts, experimental evidence, or system-
servations as the basis for conclusion or
atic observations and expresses a general
determination, as opposed to specula-
relationship between variables, as op-
tive, theoretical, or exclusively reason-
posed to a law based only on theory.
based approaches. Many forms of re-
search attempt to gain empirical evi- e m p i r i c a l l y derived test a test devel-
dence in favor of a hypothesis by oped using procedures that maximize
manipulating an INDEPENDENT VARI- CONTENT VALIDITY, CRITERION VALID-
ABLE and assessing the effect on an out- ITY, or CONSTRUCT VALIDITY OI a combi-
come or DEPENDENT VARIABLE. 2. based nation of these.
on experience.
e m p i r i c a l l y keyed test an assessment
e m p i r i c a l Bayes estimation a varia- in which answers are scored in such a
tion of BAYESIAN INFERENCE in which way as to establish differences in re-
the probability of the hypothesis being sponses among groups already known to

111
empirical method

differ. For example, on a test measuring that the mind at birth is like a blank
problem solving, the correct alternative sheet of paper. 2. the view that experi-
among the response choices would be mentation and systematic observation is
the one preferred by members of a CRI- the most important, if not the only,
TERION GROUP who were administered foundation of scientific knowledge and
the test previously. the means by which individuals evalu-
ate truth claims or the adequacy of theo-
empirical method any procedure for ries and models. 3. in philosophy, the
conducting an investigation that relies position that all linguistic expressions
upon experimentation and systematic that are not tautologous must be empiri-
observation rather than theoretical spec- cally verifiable if they are to be deemed
ulation. The term is sometimes used as a valid or meaningful. This principle was
vague synonym for SCIENTIFIC METHOD. essential to the philosophy of LOGICAL
empirical probability the number of POSITIVISM, empiricist adj., n.
times a specific event occurs over the
empowerment evaluation see PAR-
total number of trials in a process. For
TICIPATORY EVALUATION.
example, the THEORETICAL PROBABILITY
of getting a head whenflippinga coin is empty cell a category of a specified vari-
equal to .5 or V2, whereas the actual em- able or CROSS-TABULATION of variables
pirical probability obtained in an experi- in which there is no membership. For
ment may be different. Also called example, a marketing researcher might
experimental probability. be interested in color preferences (pink,
blue) for baby clothes by gender (girl,
empirical question a problem that
boy). He or she could collect data from a
can be solved by conducting an experi-
random sample of 100 people in a baby
ment or investigation, rather than by
clothes department and then create a
logic or theory. For example, "How
CONTINGENCY TABLE of the findings. For
many people own a white car?" is an em-
example, 40 people may prefer pink
pirical question. Determining its answer
clothes for girl babies, 10 may prefer
would require collecting a rather large,
blue clothes for girl babies, 50 may pre-
random sample that is representative of
fer blue clothes for boy babies, and none
the population of car owners and asking
prefer pink clothes for boy babies. The
each person in the sample the color of
CELL representing the intersection of
his or her car.
"pink" and "boy" is an empty cell
empirical test the test of a hypothesis because there is no one who endorses
by means of experiments or other sys- that combination of categories.
tematic observations.
empty set in SET THEORY, a collection of
e m p i r i c a l v a l i d i t y the degree to entities with no ELEMENTS or members.
which the accuracy of a test, model, or Also called null set.
other construct can be demonstrated
end-aversion bias the tendency for in-
through experimentation and system-
dividuals to avoid the extreme choices
atic observation (i.e., the accumulation
on a scale and instead select a choice in
of supporting research evidence) rather
the middle of the scale, closer to neutral.
than theory alone.
Such an approach results in a narrower
empiricism n. 1. an approach to EPISTE- range of responses that most likely will
MOLOGY holding that all knowledge of not be an accurate representation of the
matters of fact either arises from experi- variable being measured. For example,
ence or requires experience for its valida- a supervisor assessing employees would
tion. In particular, empiricism denies show end-aversion bias if his or her rat-
the possibility of innate ideas, arguing ings generally were around 4 on a 7-point

112
 equal intervals

scale. Also called central-tendency FUNCTION for a RANDOM VARIABLE, de-

bias; end-of-scale bias. picting the possible values it may take
endogenous parameter a variable or and usually following an inverted U or
curvilinear pattern. [V. A. Epanechnikov,
value that is predicted from or depend-
Russian statistician]
ent on another variable or value. For ex-
ample, if the well-being of cancer epidemiological s a m p l i n g any of
patients depends on coping style, the several methods for enlisting a relevant
outcome of well-being would be an en- and sufficient sample of participants for
dogenous parameter. Endogenous pa- a study to understand a health risk in the
rameters are of particular importance in population.
STRUCTURAL EQUATION MODELING and
PATH ANALYSIS. Compare EXOGENOUS
epistemological r e f l e x i v i t y an
PARAMETER.
awareness of how one's own theoretical
assumptions or situational conditions
endogenous research research that is could affect a research scenario. It is a
based on information inherent in a spe- recognition that research is not con-
cific area or system, without taking into ducted in a vacuum but rather is affected
account outside, independent informa- by the outlook of the researcher and the
tion. context in which he or she is operating.
endogenous variable a DEPENDENT epistemology n. the branch of philoso-
VARIABLE whose valucs are determined, phy concerned with the nature, origin,
caused, or explained by factors within and limitations of knowledge. It is also
the model or system under study. Com- concerned with the justification of truth
pare EXOGENOUS VARIABLE. claims. epistemological adj.
entropy n. an indication of the degree of EPSEM abbreviation for EQUAL PROBA-
disorder, disequilibrium, or change of a BILITY OF SELECTION METHOD.
closed system. For example, an entropy
measure may be used to indicate the epsilon (symbol: e) n. a small value, par-
quality of classification in a CLUSTER ticularly a small degree of RANDOM ERROR
ANALYSIS. Clusters with high similarity in a measure or a REGRESSION, EQUA-
would have low entropy, whereas clus- TION.
ters with more DISPERSION would have epsilon squared (symbol: e^) in ANALY-
higher entropy. SIS OF VARIANCE Or MULTIPLE REGRES-
entropy group in CLUSTER ANALYSIS, a SION analysis, a measure of the pro-
cluster comprising OUTLIERS that are portion of shared variance between a
not readily placed in one of the main grouping variable and an outcome that
clusters. has been adjusted for the sample size
and number of variables. See also AD-
envelope n. in a SCATTERPLOT, a pair of JUSTED R^.
lines used to encompass most of the
points and portray the main pattern. equal intervals 1. the requirement for
an INTERVAL SCALE to have an equal dis-
E-optimal design an experimental de-
tance between numbers or units over all
sign that uses one or more EIGENVALUES parts of the scale. On a Celsius tempera-
to reduce variance and bias, thereby in- ture scale, for example, there would be
creasing the amount of information re- an equal distance between, for example,
lated to parameter estimates in an 20 and 23 and between 10 and 7. 2.
experiment. See also A-OPTIMAL DESIGN; in developing scales for qualitative char-
D-OPTIMAL DESIGN. acteristics such as attitudes, the assump-
Epanechnikov k e r n e l a formula used tion that anchor values should be
to estimate the PROBABILITY DENSITY equally distant from one another, so

113
equality constraint

that, for example, the difference be- equiprobable adj. describing two or
tween agree and strongly agree is the same more different events that have the
as that between agree and strongly dis- same probability of occurring. For exam-
agree. ple, when flipping a fair coin, the occur-
equality constraint a restriction re- rence of a head or a tail is equiprobable.
quiring that two or more PARAMETERS equivalence n. a relationship between
have the same value. This is sometimes two or more items, scales, variables, or
used in procedures such as CONFIRMA- stimuli that permits one to replace an-
TORY FACTOR ANALYSIS, for example, other. See also EQUIVALENCE COEFFI-
when testing whether a FACTOR LOAD- CIENT; MEASUREMENT EQUIVALENCE.
ING is the same across different measures
of the same construct. Compare IN- equivalence coefficient a numerical
EQUALITY CONSTRAINT. index of the similarity of two different
assessments of the same measure, often
equality of variance see HOMOGENE- used as an indication of ALTERNATE-
ITY OF VARIANCE. FORMS RELIABILITY for slightly different
equal p r o b a b i l i t y of selection forms of the same test given on two dif-
method (EPSEM) a procedure for ran- ferent occasions. For example, there is
domly choosing from a larger popula- often a need to calculate an equivalence
tion, such that each person or entity has coefficient for standardized tests admin-
an equal chance of being chosen. See istered to a group of persons more than
RANDOM SAMPLING. once, to ensure that each form of the test
is measuring the same construct, such as
equal p r o b a b i l i t y s a m p l i n g see UN- verbal or quantitative ability.
EQUAL PROBABILITY SAMPLING.
equal w e i g h t i n g see WEIGHTING. equivalence paradox the situation in
which two different procedures produce
equated score the score distribution very similar outcomes, despite having
from Measure B transformed to match different initial assumptions or features.
the distribution of Measure A in one or In statistics, the term refers to the fact
more features. See TRANSFORMATION. that PRINCIPAL COMPONENTS ANALYSIS
equation n. a formal, usually brief, and FACTOR ANALYSIS often produce a
statement showing that two expressions similar pattern of dimensions and load-
are the same or equal. In mathematics or ings, although they each have a different
statistics, an equation is often a state- premise. Whereas principal components
ment with y on the left of an equal sign analysis mathematically redistributes all
and an expression for delineating or of the variance in the items to form di-
modeling y on the right side. For exam- mensions or components, factor analy-
ple, a commonly used equation for a sis more realistically separates out
straight line isy=mx-^-b, where m stands unique variance in the items before
for the slope of the line, x stands for dif- forming dimensions or factors using
ferent values along the X-AXIS, and b only the variance that is shared across
stands for the point where the line the variables.
crosses the Y-AXIS.
equivalent f o r m see ALTERNATE FORM.
equipercentile method a procedure equivalent-forms r e l i a b i l i t y see AL-
for showing how two measures are simi- TERNATE-FORMS RELIABILITY.
lar, such that a shared value of x on the
two measurements implies that the equivalent-groups design a study in
probability of a person drawn at random which the groups of participants are as-
having a score greater than x is the same sumed to be the same on all possible
for both measures. variables at the beginning of a study,
 error of measurement

such that at the end of the study any dif- rate information (e.g., a wrong response,
ferences on the response measure can a mistaken belief). 2. in experimenta-
be attributed to the experimental ma- tion, any change in a DEPENDENT VARI-
nipulation. Although the best way to ABLE not attributable to the mani-
achieve an equivalent-groups design is pulation of an INDEPENDENT VARIABLE.
by randomly selecting a large and rele- 3. in statistics, a deviation of an ob-
vant sample from the population and served score from a true score, where
then randomly assigning participants true score is often defined by the mean
to either an EXPERIMENTAL GROUP or of the particular group or condition in
a CONTROL GROUP, sometimes this which the score being assessed for error
is not possible. In such situations, re- occurs, or from the score predicted by a
searchers can instead use MATCHING, model. Errors generally are categorized
with the expectation that the resulting as SYSTEMATIC ERROR Or RANDOM ERROR.
two participant groups then would be See also RESIDUAL.
approximately equivalent (i.e., come
from the same population) at the start of error bar an area on a graph used to de-
the study. pict how much uncertainty or slippage
For example, a reading researcher may there is around an estimated PARAMETER
want to implement a reading interven- value, such as the mean. An error bar
tion in one school and a control or PLA- usually is given as one STANDARD DEVIA-
CEBO condition in another similar TION above and below a point, but it also
school. Before analyzing results between can be a STANDARD ERROR (i.e., the stan-
the two schools, the researcher could ad- dard deviation divided by the square
minister an intelligence test to students root of the SAMPLE SIZE).
at both schools. Then, the researcher
error d i s t r i b u t i o n the pattern de-
could create similar sets by matching
scribing how scores can vary randomly
students with the same or highly similar
around a particular value, such as the
scores across the schools and conduct-
mean. It is often a NORMAL DISTRIBU-
ing the study with just those students.
TION with a mean of 0 and a STANDARD
See also MATCHED-PAIRS DESIGN. Com-
DEVIATION of 1.
pare NONEQUIVALENT-GROUPS DESIGN.
erf abbreviation for ERROR FUNCTION. error f u n c t i o n (erf) a roughly "S" or
half-mountain-shaped pattern on a
ergodicity n. a principle stating that the graph with values of a particular variable
average value of a variable over a set of (x) given along the horizontal axis and
individuals in a defined space or time, values from -1 to +1 given on the verti-
such as a SAMPLE, will be the same as the cal axis. Such a pattern indicates there
average across a long TIME SERIES of would be a high level of accuracy if using
points for a single individual. For exam- the mean to estimate a value for x, and
ple, if ergodicity held for a measure of conversely a large likelihood of error or
satisfaction in an organization, the aver- inaccuracy if using more extreme values.
age satisfaction score of all employees in
the organization would be the same as error mean square see MEAN
the average satisfaction score across a SQUARED ERROR.
one-year period for one employee. In re-
ality, ergodicity does not always hold, error o f expectation an error arising
thus giving rise to different streams because of a preconceived idea of the na-
of NOMOTHETIC and IDIOGRAPHIC re- ture of the stimulus to be presented or
search, which focus on the group or the the timing of the presentation.
individual, respectively.
error o f measurement see MEASURE-
error n. 1. a deviation from true or accu- MENT ERROR.

115
error of the first kind

error of the f i r s t k i n d see TYPE I explained by the INDEPENDENT VARI-

ERROR. ABLES. Also called residual sum of
error o f the second k i n d see TYPE II
squares; sum of squared errors
(SSE).
ERROR.
error rate the frequency with which er- error term the element of a statistical
rors are made. Examples include the pro- equation that indicates the amount of
portion of an experimenter's data change in the DEPENDENT VARIABLE that
recordings that are wrong or the number is unexplained by change in the INDE-
of TYPE I ERRORS that occur during SIG- PENDENT VARIABLES. Also called dis-
NIFICANCE TESTING. turbance term; residual term.
error-rate estimation the process of error variance the element of variabil-
predicting how often an error, especially ity in a score that is produced by extra-
one of classification, will be made for neous factors, such as measurement
a specific variable. For example, medical imprecision, and is not attributable to
researchers may use error-rate estima- the INDEPENDENT VARIABLE or other
tion to predict how often they may in- controlled experimental manipulations.
correctly diagnose a patient as having Error variance usually indicates how
a specific disease. This term is often much randomfluctuationis expected
used when conducting a DISCRIMINANT within scores and often forms part of the
ANALYSIS to assess how often cases will denominator of test statistics, such as
be put into the wrong category on the the F RATIO in an ANALYSIS OF VARIANCE.
basis of the predictor scores. Also called residual error; residual
variance; unexplained (or unpre-
error rate family-wise see FAMILY- dicted) variance.
WISE ERROR RATE.
ESM abbreviation for EXPERIENCE-
error rate per comparison see SAMPLING METHOD.
TESTWISE ERROR RATE.
essay test an examination in which
error score in CLASSICAL TEST THEORY, examinees answer questions by writing
the difference between a person's ob- sentences, paragraphs, or pages. The reli-
served measurement or score and his or abUity of grading such assessments is
her expected measurement or score. Also usually lower than that of OBJECTIVE
called residual score. TESTS.
errors-in-variables p r o b l e m the estimable f u n c t i o n a function of the
commonly occurring situation in which parameters of a model that can be
PREDICTOR VARIABLES in a REGRESSION
uniquely approximated from the data.
ANALYSIS are not perfectly reliable. A
The estimable function is important in
LEAST SQUARES REGRESSION approach
GENERAL LINEAR MODEL applications.
will have bias in the parameter estimates
when there is MEASUREMENT ERROR in estimate 1. n. a best guess of the value of
the predictors. The problem has given a parameter of a DISTRIBUTION on the
rise to ATTENUATION procedures that basis of a set of empirical observations.
correct for unreliability of the estimates. 2. vb. to assign a value to a parameter in
this way.
error sum o f squares the difference
between a score and the mean score, estimating f u n c t i o n a mathematical
multiplied by this same difference. It is equation used to approximate values
part of the denominator of the F RATIO and properties (e.g., bias, consistency,
in an ANALYSIS OF VARIANCE and repre- efficiency) for a specific parameter (e.g.,
sents the amount of variation in the DE- regression coefficient) by using informa-
PENDENT VARIABLE that cannot be tion from the sample data and the pa-
 ethology

rameter of interest. This approach is dures. In an ANALYSIS OF VARIANCE, for

particularly suitable when a LIKELIHOOD example, eta squared is calculated as the
FUNCTION is not easily known. SUM OF SQUARES for the treatment (i.e.,
sum of squared differences between a
estimation n. in statistics, the process group mean and the GRAND MEAN) di-
of approximating a population PARAME- vided by the TOTAL SUM OF SQUARES. Eta
TER from sample data while allowing for squared is often used as an EFFECT SIZE
some degree of uncertainty by giving a measure.
range of values within which the param-
eter will most likely fall. Estimation usu- ethics n. the principles of morally right
ally requires a large random sample, conduct accepted by a person or a group
from which one can calculate POINT ES- or considered appropriate to a specific
TIMATE values, such as the mean and field. In psychological research, for ex-
STANDARD DEVIATION, in the larger pop- ample, proper ethics requires that partic-
ulation of interest. The process of esti- ipants be treated fairly and without
mation also involves building a harm and that investigators report re-
CONFIDENCE INTERVAL around the ob- sults and findings honestly. See RE-
tained sample value, plus and minus SEARCH ETHICS. ethical adj.
some MARGIN OF ERROR. Thus, if One cal-
culates the mean of scores on an ethics committee a designated group
achievement test to be 65, this value is a (often an INSTITUTIONAL REVIEW BOARD)
point estimate of the average test score that oversees the conduct of research to
in the population. A 95% confidence in- ensure that participants are treated fairly
terval might yield an interval of 60 to 70, and without harm.
which is likely to capture the tme mean ethnic bias 1. discrimination against
score for the population. individuals based on their ethnic group,
often resulting in inequities in such
estimation theory a system developed
areas as education, employment, health
to approximate population PARAMETERS
care, and housing. 2. in testing and mea-
from sample data. Ideally, it involves
surement, contamination or deficiency
large, random samples and other meth-
in an instrument that differentially af-
ods that minimize the amount of bias or
fects the scores of those from different
error between the true population value,
ethnic groups. Ideally, researchers strive
which usually is unknown, and the sam-
to create CULTURE-FAIR TESTS.
ple-based estimate.
ethnic research the body of studies
estimator n. a quantity calculated from
that investigates the backgrounds of
the values in a sample according to some
people from different races and cultures.
mle and used to give an approximation
of the value in a population. For exam- ethnography n. the descriptive study
ple, the sample mean or average is an es- of cultures or societies based on direct
timator for the population mean; the observation (see FIELD RESEARCH) and
value of the population mean is the esti- (ideally) some degree of participation.
mate. See EFFICIENT ESTIMATOR. Compare ETHNOLOGY. ethnogra-
pher n. ethnographic adj.
eta (symbol: Tj) n. 1. in STRUCTURAL EQUA-
TION MODELING, denoting a LATENT VARI- ethnology n. the comparative, analyti-
ABLE. 2. see CORRELATION RATIO. cal, or historical study of human cultures
or societies. Compare ETHNOGRAPHY. See
eta squared (symbol: ij^) a measure of
also ETics. ethnological adj. eth-
the amount of variance in a DEPENDENT
nologist n.
VARIABLE that can be explained by one
or more INDEPENDENT VARIABLES, as cal- ethology n. the comparative study of
culated in a variety of statistical proce- the behavior of animals, typically in

117
etics

their natural habitat but also involving ered a Euler diagram. The following is a
experiments both in the field and in cap- generic depiction of a Euler diagram.
tivity. Ethology was developed by be-
Set1
havioral biologists in Europe and
is often associated with connotations
of innate or species-specific behavior
patterns. The term increasingly is used
to describe research involving obser-
vation and detailed descriptions of
human behavior as well. ethological
adj. ethologist n.

etics n. the study of behaviors or situa-

tions that are held to be generalizable
across a wide range of cultures and back- [Leonhard Euler (1707-1783), Swiss-
grounds. This contrasts with emics, born Russian mathematician]
which refers to investigation of behav- evaluation apprehension uneasiness
iors and situations specific to a particu- or worry about being judged by others,
lar culture. For example, an etics-focused especially feelings of worry experienced
researcher might evaluate which facial by participants in an experiment as a re-
expressions are fairly universal across sult of their desire to be evaluated favor-
different countries and age groups, ably by the experimenter. Participants
whereas an emics-focused researcher experiencing evaluation apprehension
would investigate those expressions that may inhibit reactions (e.g., the display
are unique to a cultural group. of aggression) that they believe will lead
the experimenter to regard them as psy-
Euclidean distance a simple measure chologically unhealthy.
of the distance between two points, as
quantified by a ruler or other numerical evaluation research the use of scien-
gauge. Also called Euclidean metric. tific principles and methods to assess the
influence or effectiveness of social inter-
Euclidean space in statistics and ventions and programs, including those
mathematics, a three-dimensional area related to mental health, education, and
or plot. Algebra, geometry, and calculus safety (e.g., crime prevention, automo-
may be used to extend Euclidean spaces bile accident prevention). Evaluation re-
beyond three dimensions. search is thus a type of APPLIED RESEARCH.
See also ASSESSMENT RESEARCH.
Euler d i a g r a m a pictorial image of
how entities or situations intersect or event n. in probability theory, any of the
overlap. While a VENN DIAGRAM depicts namable things that can be said to result
all areas of intersection, a Euler diagram from a single trial of an experiment of
does not necessarily include all areas. chance. For example, in the roll of a sin-
For example, a researcher might create a gle die, the events could include (among
Venn diagram in which a large circle others) any of the six individual num-
represents a population, a circle inside bers, any even number, and any odd
that represents a sample or subset of the number.
population, and additional overlapping event history analysis see SURVIVAL
circles inside the previous one indicate ANALYSIS.
additional subsets. If the image were to
leave out some of the inner circles, such event history data information col-
as the ones representing different kinds lected over a time period, often using di-
of random samples, it would be consid- aries or other recording methods that
 exogenous variable

can accurately map what occurred at priate on different versions of a test and
various points in time. Event history methods or parameters that work well
data often are used in psychology, soci- under different conditions show ex-
ology, health sciences, education, and changeability.
engineering.
exclusive adj. completely separate and
event record a log containing the de- incompatible. For example, gender has
tails of a particular situation or occur- two exclusive categories of male and fe-
rence as assessed over a period of time. male. See also MUTUALLY EXCLUSIVE
For example, an event record may col- EVENTS.
lect information over numerous time
points throughout a week on the mood exclusive u n i o n in probability studies,
of a patient who is enrolled in a study to the combination of two or more events
reduce depression. that have no overlap. For example, the
set of all hearts and spades in a deck of
event s a m p l i n g a strategy commonly cards forms an exclusive union of two
used in DIRECT OBSERVATION that in- separate events, as a card cannot be both
volves noting and recording the occur- a heart and a spade at the same time. In
rence of a carefully specified behavior probability, the chance of two separate
whenever it is seen. For example, a re- events occurring together (i.e., an exclu-
searcher may record each episode of sive union) is equal to the probability of
apnea that occurs within a 9-hour pe- the first event plus the probability of the
riod overnight while a person sleeps. second event.
event space a subset of the set of all pos- exhaustive adj. complete or all encom-
sible outcomes in a situation (the SAM- passing. For example, a set of 13 hearts,
PLE SPACE). For example, the sample 13 spades, 13 diamonds, 13 clubs, and
space of working days for a particular or- two jokers forms an exhaustive set of all
ganization might comprise Sunday, possible playing cards that could be
Monday, Tuesday, Wednesday, Thurs- drawn.
day, Friday, and Saturday for all employ-
ees, but the event space for working days exogenous parameter a PARAMETER
might include only Monday through that occurs outside of and is not ex-
Friday for some employees. plained by a system or model but never-
theless has an effect on variables within
events p a r a d i g m a research design or that system. For example, poor socioeco-
therapeutic approach based on the inter- nomic status is sometimes seen as lead-
pretation of EVENT HISTORY DATA. ing to lower levels of education, which
exact i d e n t i f i c a t i o n see JUST-IDENTI- in turn could lead to lower income lev-
FIED MODEL. els. Thus, in a study of the effect of edu-
cation on income level socioeconomic
exact r e p l i c a t i o n see REPLICATION.
status might be considered an exoge-
exact test any statistical test in which nous parameter: It predicts a mediator
the probability of finding a result as ex- (education), which then predicts an
treme or more extreme than the one ob- endogenous or DEPENDENT VARIABLE
tained, given the NULL HYPOTHESIS, can (income). Compare ENDOGENOUS PARA-
be calculated precisely rather than ap- METER.
proximated. The FISHER EXACT TEST iS
exogenous variable an INDEPENDENT
an example.
VARIABLE whose value is determined by
exchangeability n. the ability to be factors outside the model or system
used in different circumstances or situa- under study. Also called exogenous
tions. For example, in measurement and factor. Compare ENDOGENOUS VARI-
statistics, items that are equally appro- ABLE.

119
exp

exp abbreviation for EXPONENTIAL FUNC- first variable plus the long-term average
TION. of the second variable.
expansion n. a description of a mathe- expectation-conditional maximi-
matical or statistical function via a series zation algorithm (ECM algorithm)
of sums. The Taylor expansion (see TAY- a procedure for finding M A X I M U M LIKE-
LOR SERIES) is an example. LIHOOD estimates of PARAMETERS in sta-
tistical models. Its computations are
expansion f a c t o r the amount by easier than those of traditional MAXI-
which scores can change over time or M U M LIKELIHOOD-EXPECTATION MAXI-
with differences in testing conditions. It MIZATION ESTIMATION methods.
is often important to consider an expan-
sion factor in ITEM RESPONSE THEORY expectation-conditional m a x i m i -
when predicting a pattern of scores. zation either a l g o r i t h m (ECME
algorithm) an iterative procedure for
expectancy n. see EXPECTATION. ex-
finding MAXIMUM LIKELIHOOD estimates
pectant adj.
of PARAMETERS in Statistical models that
expectancy chart a table or graph that alternates between performing an ex-
provides estimates of the future outcome pectation step and a maximization
for a specific variable. For example, in- step. It is an improved extension of the
dustrial and organizational psycholo- EXPECTATION-CONDITIONAL MAXIMIZA-
gists may use expectancy charts based TION ALGORITHM that is easier and faster
on test scores to determine the likeli- to compute.
hood of an individual's success in a job,
and insurance companies may use ex- expectation surface the space or
pectancy charts to predict how long a plane that contains a series of points
person will live based on such factors as that are the likely values of an outcome
health, current age, financial resources, given a set of predictor or explanatory
and occupation. variables or parameters.

expectancy c o n t r o l design an exper- expected frequency a frequency pre-

imental design in which the EXPERI- dicted from a theoretical model, as op-
MENTER EXPECTANCY EFFECT operates
posed to an OBSERVED FREQUENCY.
separately from the INDEPENDENT VARI- Expected frequencies are chiefly used in
ABLE of primary substantive interest. the CHI-SQUARE TEST. For example, sup-
pose a researcher wishes to compare the
expectancy effect see EXPERIMENTER ethnic distribution of a sample of 100
EXPECTANCY EFFECT. participants with the ethnic distribution
of the entire United States. Based on
expectation n. the long-term average
data from the 2010 U.S. Census, the
of a RANDOM VARIABLE. For example, expected frequency values for ethnic
the expectation is that the mean for a distribution of the sample would be as
specific random variable obtained with follows: 63.7 White, 16.3 Hispanic/La-
an extremely large sample will equal the fino/Spanish, 12.6 Black/African Ameri-
mean of the population of interest. Also can, 4.8 Asian, 0.9 Alaskan/Native
called expectancy. See EXPECTED VALUE. American, 0.2 Hawaiian/Pacific Island,
expectation algebra the set of mathe- and 1.5 Other/Mixed. The researcher
matical rules that inform how to find would compare the observed frequency
the long-term average of a RANDOM from the sample of 100 to these ex-
VARIABLE. For example, one of the rules pected frequencies and calculate a CHI-
of expectation algebra is that the long- SQUARE GOODNESS-OF-FIT TEST. If the
term average of the sum of two variables chi-square is significantly larger than
is equal to the long-term average of the the CRITICAL VALUE obtained from a sta-

120
 experimental error

tistical chart on 6 DEGREES OF FREEDOM, relationship. An experiment involves

where the degrees of freedom equal the the manipulation of an INDEPENDENT
number of categories minus one, there is VARIABLE, the measurement of a DE-
evidence that the sample distribution PENDENT VARIABLE, and the exposure of
does not match that of the 2010 census. various participants to one or more of
If the calculated chi-square value is the conditions being studied, RANDOM
smaller than the one obtained from the SELECTION of participants and their
statistical chart (i.e., the p VALUE is RANDOM ASSIGNMENT to Conditions also
greater than .05), then the researcher are necessary in experiments. experi-
cannot reject the hypothesis that there mental adj.
is a good fit between thefittedfrequency
experimental a t t r i t i o n see ATTRI-
values in the U.S. population and the
TION.
sample ethnic distribution observed fre-
quency values. Also called fitted fre- experimental c o u f l i t i o n a level of
quency. the INDEPENDENT VARIABLE that is ma-
nipulated by the researcher in order to
expected t value the number that assess the effect on a DEPENDENT VARI-
would occur due to chance alone when ABLE. Participants in an experimental
examining the number of estimated condition receive some form of TREAT-
STANDARD DEVIATION unltS a SAMPLE MENT or experience whereas those in a
MEAN is from another sample or popula- CONTROL CONDITION do not. For exam-
tion mean. Under the NULL HYPOTHESIS, ple, patients in an experimental condi-
the expected t value is zero, indicating tion may receive a new drug, whereas
no difference between means. See T those in a control condition may receive
TEST. a pill that looks like the new drug but is
expected value the value of a random only a PLACEBO containing some inert
variable or one of its functions as de- substance.
rived by mathematical calculation experimental c o n t r o l see CONTROL.
rather than observation. It is symbolized
by E(A:), with x varying according to the experimental design an outline or
specific item of interest that is being cal- plan of the procedures to be followed
culated. Usually, the expected value is a in scientific experimentation in order
mean or weighted average. to reach valid conclusions, with consid-
eration of such factors as participant se-
experience-sampling method (ESM) lection, variable manipulation, data
a procedure for assessing research partic- collection and analysis, and minimiza-
ipants' thoughts and feelings at specific, tion of external influences.
predetermined points in time. For exam-
ple, researchers conducting a weight re- experimental error the various kinds
duction program may check in with of ERROR that can occur in a controlled
participants at various times to assess study. An experimental error may be a
their dietary intake of healthy food and SYSTEMATIC ERROR, referring to a prob-
any urges to eat unhealthy food. The ex- lem that can be identified, reproduced,
perience-sampling method is believed to and potentially corrected; or a RANDOM
provide more reliable input on the inner ERROR, referring to chance fluctuations
state of participants than may be ob- that cannot easily be corrected. For ex-
tained via after-the-fact feedback. ample, a researcher may be conducting a
study to assess the effectiveness of a new
experiment n. a series of observations treatment. If the treatment dose is
conducted under controlled conditions slightly off due to faulty measurements
to study a relationship with the purpose or ingredients, this is a systematic error
of drawing causal inferences about that that could be corrected. If participants

121
experimental ethics

react somewhat differently to the treat- tific study of behavior, motives, or cog-
ment, some more positively and some nition in a laboratory or other con-
more negatively, this is a random error trolled setting in order to predict,
that is not easily eradicated but that may explain, or influence behavior or other
be somewhat reduced (e.g., by randomly psychological phenomena. Experimen-
assigning a large set of individuals to tal psychology aims at establishing
EXPERIMENTAL GROUPS VS. CONTROL quantified relationships and explana-
GROUPS). tory theory through the analysis of re-
sponses under various controlled
experimental ethics see RESEARCH
conditions and the synthesis of ade-
ETHICS. quate theoretical accounts from the re-
experimental group a group of par- sults of these observations.
ticipants in a research study who are ex-
posed to a particular manipulation of experimental realism the extent to
the INDEPENDENT VARIABLE (i.e., a par- which a controlled study is meaningful
ticular treatment or TREATMENT LEVEL). and engaging to participants, eliciting
The responses of the experimental group responses that are spontaneous and nat-
are compared to the responses of a CON-
ural. See also MUNDANE REALISM.
TROL GROUP, other experimental groups, experimental r e p l i c a t i o n see REPLI-
or both. Also called treatment group. CATION.
experimental hypothesis a premise experimental research research uti-
that describes what a researcher in a sci- lizing randomized assignment of partici-
entific study hopes to demonstrate if pants to conditions and systematic
certain conditions are met, such as RAN- manipulation of variables with the ob-
DOM SELECTION of participants, RAN- jective of drawing causal inference. It
DOM ASSIGNMENT tO EXPERIMENTAL is generally conducted within a labora-
GROUPS or CONTROL GROUPS, and ma- tory or other controlled environment,
nipulation of an INDEPENDENT VARI- which in reducing the potential influ-
ABLE. ence of extraneous factors increases IN-
TERNAL VALIDITY but decreases EXTERNAL
experimental m a n i p u l a t i o n in an
VALIDITY. Compare CORRELATIONAL RE-
experiment, the manipulation of one or
SEARCH; FIELD RESEARCH; QUASI-EXPERI-
more INDEPENDENT VARIABLES in order
MENTAL RESEARCH.
to investigate their effect on a DEPEND-
ENT VARIABLE. An example would be the experimental series the trials admin-
assignment of a specific treatment or istered to an EXPERIMENTAL GROUP in 3
PLACEBO to participants in a research controlled research study, as opposed to
study in order to control possible CON- those administered to the CONTROL
FOUNDS and assess the effect of the treat- GROUP.
ment.
experimental treatment 1. In re-
experimental method a system of search, the conditions applied to one or
scientific investigation, usually based on more groups that are expected to cause
a design to be carried out under con- change in some outcome or DEPENDENT
trolled conditions, that is intended to VARIABLE. 2. an intervention or regimen
test a hypothesis and establish a causal that has shown some promise as a cure
relationship between independent and or ameliorative for a disease or condition
dependent variables. but is still being evaluated for efficacy,
experimental p r o b a b i l i t y see EM- safety, and acceptability.
PIRICAL PROBABILITY.
experimental u n i t the unit to which
experimental psychology the scien- an experimental manipulation is ap-
 experimenter modeling effect

plied. For example, if an intervention ther interaction with participants or un-

is applied to a classroom of students,
the classroom (not the individual stu-
dents) is the experimental unit. In
QUASI-EXPERIMENTAL RESEARCH and
Other nonexperimental contexts, the
term UNIT OF ANALYSIS is used instead.
experimental variable an INDEPEND-
ENT VARIABLE that is manipulated by the
researcher to determine its relationship
to or influence upon some outcome or
DEPENDENT VARIABLE.
intentional errors of observation, mea-
surement, analysis, or interpretation. In
the former, the experimenter's personal
characteristics (e.g., age, sex, race), atti-
tudes, and expectations directly affect
the behavior of participants. In the lat-
ter, the experimenter's procedural errors
(often arising from his or her expecta-
tions about results) have no effect on
participant responses but indirectly dis-
tort the research findings.

experimentation n. the carrying out experimenter expectancy effect a

of EXPERIMENTS. type of EXPERIMENTER EFFECT in which a
researcher's expectations about the find-
experimentee n. a research participant, ings of his or her research are inadver-
that is, a person who is studied under tently conveyed to participants and
controlled conditions by the researcher influence their responses. This distor-
or EXPERIMENTER. tion of results arises from participants'
experimenter n. in a controlled study, reactions to subtle cues (DEMAND CHAR-
the researcher who devises tasks for the ACTERISTICS) unintentionally given by
EXPERIMENTEE to perform, determines the researcherfor example, through
the conditions under which the perfor- body movements, gestures, or facial ex-
mance will take place, and monitors and pressionsand may threaten the ECO-
interprets the results. LOGICAL VALIDITY of the research. The
term is often used synonymously with
experimenter bias any systematic er-
ROSENTHAL EFFECT.
rors in the research process or the inter-
pretation of its results that are attributable experimenter intentional effect
to a researcher's preconceived beliefs, ex- the EXPERIMENTER EFFECT that OCCUrS
pectancies, or desires about results. For when a researcher deliberately tries to
example, a researcher may inadvertently manipulate the result in order to test
cue participants to behave or respond in whether certain behaviors or character-
a particular way. See EXPERIMENTER EF- istics of a researcher can alter the out-
FECT. come. The experimenter intentional
experimenter biosocial effect in a effect may be a positive source of infor-
controlled study, an unintended effect mation on factors that can influence an
outcome, or it may be a form of EXPERI-
on participants' responses that is associ-
MENTER BIAS that threatens the INTER-
ated with individual differences among
NAL VALIDITY of a study.
the biological characteristics of the re-
searcher (e.g., gender, age). Compare EX- experimenter interpreter effect
PERIMENTER PSYCHOSOCIAL EFFECT. a SYSTEMATIC ERROR that results when
experimenter d r i f t a gradual and un- a researcher's interpretation of the ob-
conscious change in the way a researcher served data is biased (e.g., by the re-
conducts or interprets a controlled study, searcher's expectation).
leading to SYSTEMATIC ERROR in the re-
experimenter m o d e l i n g effect a
sults.
SYSTEMATIC ERROR that results in partic-
experimenter effect any influence ipants responding to a task in a way that
a researcher may have on the results is too similar to the way in which the re-
of his or her research, derived from ei- searcher would respond to the task.

123
experimenter observer effect

experimenter observer effect a SYS- pants based on relevant criteria for a

TEMATIC ERROR by an investigator in the study.
perception or recording of data. This is
explained s u m o f squares see RE-
often predictable from a knowledge of
GRESSION SUM OF SQUARES.
the investigator's expectation.
explained variance the proportion of
experimenter psychosocial effect the variance in a DEPENDENT VARIABLE
in a controlled study, an unintended ef- that can be determined or explained by
fect on participants' responses associ- one or more other variables. In ANALYSIS
ated with individual differences among OF VARIANCE procedures the explained
the psychological characteristics of the variance usually is given by ETA SQUARED
researcher. Compare EXPERIMENTER BIO- or OMEGA SQUARED, whereas in predic-
SOCIAL EFFECT. tion methods, such as MULTIPLE REGRES-
SION, it usually is given by the
experiment o f nature see NATURAL COEFFICIENT OF MULTIPLE DETERMINA-
EXPERIMENT.
TION. Explained variance also is applied
experiment-wise a l p h a level the in procedures such as PRINCIPAL C O M -
PONENTS ANALYSIS or FACTOR ANALYSIS
SIGNIFICANCE LEVEL (i.e., the acceptable
risk of making a TYPE I ERROR) that is set to describe the proportion of the vari-
by a researcher for a set of multiple com- ance in a measure that can be under-
parisons and statistical tests. It is often stood or predicted from a component or
set at the conventional level of .05; by FACTOR. Compare ERROR VARIANCE.
setting the alpha for individual tests (the explanatory analysis an analysis that
TESTWISE ALPHA LEVEL) at a lower rate sets out to understand the causes of a
(e.g., .01) it can be ensured that the particular phenomenon, often involv-
whole set of tests does not produce error ing the specification and testing of ME-
greater than the desired experiment- DIATORS to assess how INDEPENDENT
Wise level. When conducting a one-way VARIABLES and DEPENDENT VARIABLES
or one-factor ANALYSIS OF VARIANCE, ex- are causally related.
periment-wise alpha usually is the same
as the FAMILY-WISE ALPHA LEVEL; in an explanatory research a study con-
analysis of variance involving two or ducted to assess why a particular finding
more factors, family-wise alpha refers to occurred. For example, one might con-
the probability of Type I error for each duct explanatory research to determine
test of a specific factor. why individuals who have been abused
as children tend to be at higher risk for
experiment-wise error rate in a test negative outcomes as adults. Compare
involving MULTIPLE COMPARISONS, the PREDICTIVE RESEARCH.
probability of making at least one TYPE I explanatory t r i a l a CLINICAL TRIAL
ERROR over an entire research study. The
conducted to understand the causal fac-
experiment-wise error rate differs from tors that make a particular treatment ef-
the TESTWISE ERROR RATE, which is the fective in bringing about a specific
probability of making a Type 1 error outcome. Explanatory trials contrast
when performing a specific test or com- with pragmatic trials, which are de-
parison. signed to find an effective treatment
expert-choice s a m p l i n g selecting without necessarily trying to understand
participants for a study by using the why it is effective.
opinions of an authority in the area of explanatory variable see INDEPEND-
interest. Expert-choice sampling is a ENT VARIABLE.
type of PURPOSIVE SAMPLING that inten-
tionally seeks specific kinds of partici- explicative research a study con-

124
 exponential notation

ducted to understand the relationships script of another number to indicate

or correlations among variables. how many times the base number is to
be multiplied by itself. For example, the
exploratory data analysis (EDA) the
base number 2 with an exponent of 3
analysis of data to generate new research (e.g., 2^) denotes the number formed by
questions or insights rather than to ad- multiplying 2 by itself three times (i.e., 2
dress specific preplanned research ques- X 2 X 2 = 8).
tions. Compare CONFIRMATORY DATA
ANALYSIS. exponential adj. having an increase or
exploratory f a c t o r analysis (EFA) a decrease that is proportional to the cur-
method for finding a small set of under- rent value of a function. For example,
lying dimensions from a large set of re- exponential growth of applications to a
lated measures. The observed data are school would occur if the number of ap-
freely explored in order to discover the plications received each year doubled, so
underlying (latent) variables that ex- that after four years applications had in-
plain the interrelationships among a creased by a factor of 16.
larger set of observable (manifest) vari- exponential curve a graph of a rapidly
ables. For example, exploratory factor growing FUNCTION in which the in-
analysis has been conducted to assess crease is proportional to the size of an x
whether there are one, two, or more di- variable and the SLOPE is equal to the
mensions underlying items used to as- value of the y variable.
sess intelligence. The analysis does not
yield a unique solution: Although points exponential d i s t r i b u t i o n a theoreti-
are fixed in a multidimensional space of cal distribution of survival times, used in
underlying factors, the axes according to parametric SURVIVAL ANALYSIS when the
which one should interpret the factors HAZARD RATE is thought to be constant
are not fixed and are infinite in number. over time. It is a specific type of GAMMA
These axes may be either ORTHOGONAL DISTRIBUTION.
(producing uncorrelated factors) or
exponential f a m i l y a set of PROBABIL-
OBLIQUE (producing correlated factors).
ITY DISTRIBUTIONS that include expo-
Compare CONFIRMATORY FACTOR ANAL-
nential terms and that have similar
YSIS.
characteristics. The NORMAL DISTRIBU-
exploratory research a study that is TION is a commonly known example.
conducted when not much is known exponential function (exp) a mathe-
about a particular phenomenon. In ex- matical expression of the type y = a",
ploratory research one typically seeks to where A is a constant. A particular type
identify multiple possible links between has the form y = e", where e is a funda-
variables. In contrast, in CONFIRMATORY mental mathematical constant that is
RESEARCH one tests whether a specific
the base of natural logarithms (with the
prediction holds under specific circum- value 2.718...). Often, functions of this
stances. type (e.g., e*"^") are written as exp(>: + a).
exploratory survey an initial study
exponential n o t a t i o n a shorthand
used to test whether a research area is
worth pursuing through more formal way of depicting a number with many
EXPERIMENTAL RESEARCH. An explana-
zero values. For numbers with many
tory survey is similar to but less formal zeros to the left of a decimal, exponen-
than a PILOT STUDY, which is conducted tial notation lists the number and multi-
to provide a preliminary run of a specific plies it by 10 raised to the number of
study on a smaller scale. zeros; for numbers with many zeros to
the right of a decimal, exponential nota-
exponent n. a number shown as a super- tion lists the number and multiplies it

125
exponential smoothing

by 10 raised to the negative of the num- group of participants who did not, yet
ber of zeros. For example, 5,000,000 the researcher has no control over which
written in exponential notation is 5 x specific people actually view the adver-
10^ and 0.0000005 written in exponen- tisement. Also called ex post facto
tial notation is 5 x 10"^. study.
exponential smoothing a method of ex post facto hypothesis a HYPOTHE-
weighting TIME-SERIES data using expo- SIS about what caused or brought about
nentially decreasing weights for older a condition that is made only after one
events. This reduces random fluctua- has examined information already col-
tions in the data so that one is better able lected about that condition.
to see the underlying trends in the series extended Cox regression model see
and gives recent data relatively greater
cox REGRESSION ANALYSIS.
importance in forecasting. Exponential
smoothing is widely used with financial external r e l i a b i l i t y the extent to
and economic data. which a measure is consistent when as-
sessed over time or across different indi-
exponential t r e n d a pattern in which viduals. External reliability calculated
data values increase over time propor- across time is referred to more specifi-
tional to a specific value. For example, cally as RETEST RELIABILITY; external re-
housing costs that increase annually by liability calculated across individuals is
a factor of 1.04 could be said to manifest referred to more specifically as INTER-
an exponential trend, such that one RATER RELIABILITY.
could use the trend to predict the future
value of any given house based on its external v a l i d i t y the extent to which
original cost. Thus, a house that cost the results of research or testing can be
$100,000 to purchase originally would generalized beyond the sample that gen-
cost an estimated $148,000 to purchase erated the results to other individuals.
10 years later, as determined by multi- Situations, and time periods. For exam-
plying the original house cost by 1.04, ple, if research has been conducted only
multiplying the resulting value by 1.04, with male participants, it cannot be as-
and then repeating the second multipli- sumed that similar results will apply to
cation step 10 times. female participants. The more special-
ized the sample, the less likely will it be
ex post facto data information about that the results are highly generalizable.
past occurrences. For example, a health Compare INTERNAL VALIDITY.
researcher could collect ex post facto
data on the number of hours per week a extraneous variable a measure that is
person believes he or she exercised over not under investigation in an experi-
the past month. ment but may potentially affect the out-
come or DEPENDENT VARIABLE and thus
ex post facto design research that ex- influence results. Such potential influ-
amines past occurrences in order to un- ence often requires that an extraneous
derstand a current state. Although this variable be controlled during research.
type of design involves both a DEPEND- See also CONFOUND.
ENT VARIABLE and an INDEPENDENT
VARIABLE, the investigator cannot ma-
extraneous variance variation in
nipulate the latter. For example, a re- scores that is' caused by CONFOUNDS,
searcher who is interested in deter- HIDDEN VARIABLES, or Other factors that
mining the effectiveness of a particular are not under investigation in a study.
television advertisement on consumer extrapolation n. the process of esti-
behavior may recruit a group of partici- mating or projecting unknown score
pants who saw the advertisement and a values on the basis of the known scores
 eyeballing

obtained from a given sample. For exam- least values of a large set of random vari-
ple, a researcher might estimate how ables that form an approximately
well students will do on an achievement straight line from the lower left to the
test on the basis of their current perfor- upper right of a graph. The GUMBEL DIS-
mance, or estimate how well a similar TRIBUTION is an example.
group of students might perform on the eyeballing n. slang for a preliminary
same achievement test. casual look at research results, usually
extreme-value d i s t r i b u t i o n a lim- prior to a more formal analysis of the
ited DISTRIBUTION of the greatest or data. Also called eyeball test.

127
 Ff
f 1. in an ANALYSIS OF VARIANCE or an pendent variable. 4. in FACTOR ANALY-
ANALYSIS OF COVARIANCE, an EFFECT SIZE SIS, an underlying, unobservable LATENT
index that represents the STANDARD DE- VARIABLE thought (together with other
VIATION of a set of group means divided factors) to be responsible for the interre-
by the average standard deviation of lations among a set of variables. 5. in
scores across the set of groups. Its value mathematics, a number that divides
ranges from 0 to infinity, with small, without remainder into another num-
medium, and large effect size values sug- ber.
gested as .10, .25, and .40, respectively.
2. in a table or distribution, symbol for f a c t o r analysis (FA) a broad family of
FREQUENCY. mathematical procedures for reducing a
set of intercorrelations among MANIFEST
F symbol for F RATIO. VARIABLES to a Smaller set of unobserved
FA abbreviation for FACTOR ANALYSIS. LATENT VARIABLES or factors. For exam-
ple, a number of tests of mechanical
facet n. 1. any component of an entity ability might be intercorrelated to en-
that itself may be assessed as an inde- able factor analysis to reduce them to a
pendent entity. For example, coopera- few factors, such as fine motor coordina-
tion and trust might be considered facets tion, speed, and attention. This tech-
of agreeableness in personality assess- nique is often used to examine the
ment. 2. in GENERALIZABILITY THEORY, common influences believed to give rise
an aspect of data that causes scores to to a set of observed measures (measure-
vary. For example, fluctuation in a set of ment structure) or to reduce a larger set
scores due to using different raters, dif- of measures to a smaller set of linear
ferent time points, or different items for composites for use in subsequent analy-
a particular test could be considered a sis (data reduction). See CONFIRMATORY
facet. 3. in FACTOR ANALYSIS, a primary FACTOR ANALYSIS; EXPLORATORY FAC-
or FIRST-ORDER FACTOR associated with TOR ANALYSIS; PRINCIPAL COMPONENTS
one or more SECOND-ORDER FACTORS. ANALYSIS.
face v a l i d i t y the apparent soundness f a c t o r i a l 1. adj. describing a design for
of a test or measure. The face validity of an ANALYSIS OF VARIANCE in which the
an instrument is the extent to which the effect of multiple INDEPENDENT VARI-
items or content of the test appear to be ABLES, each having two or more catego-
appropriate for measuring something, ries of response, is assessed. For example,
regardless of whether they actually are. a researcher may wish to examine the ef-
A test with face validity, however, may fect by gender (male vs. female) of a new
lack EMPIRICAL VALIDITY. drug treatment (drug vs. no drug). 2. n.
f a c t o r n. 1. anything that contributes to the value obtained when multiplying a
a result or has a causal relationship to a given integer by each of the positive in-
phenomenon, event, or action. 2. an tegers preceding it in value. Factorials
underlying influence that accounts in are indicated by a number followed by
part for variations in individual behav- an exclamation point. For example, 4! (4
ior. 3. in ANALYSIS OF VARIANCE and factorial) denotes 1 x 2 x 3 x 4 , which
other statistical procedures, an inde- equals 24.
 factorial validity

f a c t o r i a l analysis o f covariance a f a c t o r i a l invariance the situation in

statistical procedure to understand the which the pattern of FACTOR LOADINGS
effect on a DEPENDENT VARIABLE of two on a LATENT VARIABLE in a FACTOR
or more INDEPENDENT VARIABLES that ANALYSIS remains identical from sample
are CATEGORICAL in nature, plus one or to sample. For example, if the same fac-
more additional correlated variables that tor loading values were obtained for a
are QUANTITATIVE in nature. For exam- measure of general intelligence whether
ple, a researcher may use a factorial anal- a researcher tested men or women, the
ysis of covariance to understand reading general intelligence test would be said to
performance by examining the effect of demonstrate factorial invariance by gen-
phoneme training (i.e., yes vs. no) and der. In other contexts, factorial invariance
gender (male vs. female), after taking is referred to more broadly as MEASURE-
into account the intelligence scores of MENT INVARIANCE.
the research participants; The factorial
analysis thus provides a test of whether f a c t o r i a l r e p l i c a t i o n the appearance
there are treatment condition and gen- of the same underlying LATENT VARI-
der differences, as well as whether there ABLE (factor) across repeated analyses.
is an interaction after controlling for pre- For example, a researcher might investi-
existing intelligence differences among gate factorial replication for a new ver-
the participants. sion of an intelligence test by assessing
whether the FACTOR LOADINGS showing
f a c t o r i a l analysis o f variance a sta- how strongly items relate to the underly-
tistical procedure to understand the ef- ing intelligence dimension were the
fect on a DEPENDENT VARIABLE that IS same on two random samples of partici-
QUANTITATIVE in nature of two or more pants.
INDEPENDENT VARIABLES that are CATE-
GORICAL in nature. For example, a health f a c t o r i a l table a CROSS-TABULATION
researcher may use a factorial analysis of that shows the extent to which two IN-
variance to examine the effects of diet DEPENDENT VARIABLES affect a DEPEND-
(e.g., high vs. low carbohydrates) and ENT VARIABLE. For example, in an
exercise (e.g., 3 hours per week vs. 1 ANALYSIS OF VARIANCE, a factorial table
hour per week) on weight. might be used to display the findings
from a design that investigated how
f a c t o r i a l design an experimental much teaching style (lecture vs. hands
study in which two or more CATEGORI- on) and gender (male vs. female) relate
CAL VARIABLES are simultaneously ma- to student performance in a statistics
nipulated or observed in order to study class. The factorial table would show the
their joint influence (INTERACTION EF- mean performance for women in a lec-
FECT) and separate influences (MAIN ture class, women in a hands-on class,
EFFECTS) on a separate DEPENDENT men in a lecture class, and men in a
VARIABLE. For example, a researcher hands-on class.
could use a factorial design to investi-
gate treatment type (e.g., new exercise f a c t o r i a l v a l i d i t y confirmation that
procedure vs. traditional procedure) and the underlying structure of dimensions
age (< 40 vs. > 40). The primary advan- and loadings are as expected when con-
tages of factorial designs are that they ducting a FACTOR ANALYSIS on a set of
allow for the evaluation of interrelation- items. For example, a test of well-being
ships and that they are more efficient would be said to have factorial validity if
than conducting multiple studies with the items measuring satisfaction loaded
one variable at a time. See also FRAC- positively and the items measuring
TIONAL FACTORIAL DESIGN; TWO-BY-TWO meaninglessness loaded negatively, with
FACTORIAL DESIGN. similar values across several random sam-

129
factoring

pies of individuals. Also called struc- of the MANIFEST VARIABLE in terms of

tural validity. the underlying LATENT VARIABLES or fac-
tors. Consider the example below.
f a c t o r i n g n. 1. in FACTOR ANALYSIS, the
It provides information about the
process of extracting dimensions or un- unique relation of a measured item to
derlying LATENT VARIABLES (factors). 2. the factor in question (Factor 1), control-
in mathematics, the subdivision of a tar- ling for other factors (Factor 2). See also
get number into a series of numbers FACTOR STRUCTURE MATRIX.
whose product is the target number.
f a c t o r r e f l e c t i o n changing the signs
f a c t o r l o a d i n g in FACTOR ANALYSIS,
of a set of FACTOR LOADINGS from posi-
the correlation between a MANIFEST
tive to negative, or vice versa. For exam-
VARIABLE and a LATENT VARIABLE (fac-
ple, researchers studying a newly revised
tor). The factor loading reflects the de-
scale to measure happiness might use
gree to which a manifest variable is said
factor reflection to change a negative
to be "made up of" the factor being ex-
sign on the factor loadings for positive
amined.
items as well as to change a positive sign
factor method any means by which for sadness items. The resulting pattern
LATENT VARIABLES (factors) are extracted of loadings thus shows positively fo-
or identified in FACTOR ANALYSIS. Widely cused items as having a positive sign and
used factor methods include PRINCIPAL sadness-focused items as having a nega-
COMPONENTS ANALYSIS, whlch seeks tive sign, such that the entire scale could
to find a set of LINEAR COMBINATIONS accurately be said to measure the posi-
called components that help explain the tive state of happiness.
correlations among variables; and PRIN-
CIPAL-AXIS FACTOR ANALYSIS, in which f a c t o r r o t a t i o n in FACTOR ANALYSIS,
underlying dimensions or factors are the repositioning of LATENT VARIABLES
sought to explain the correlations (factors) to a new, more interpretable
among variables after separating out configuration by a set of mathematically
COMMONALITY and putting aside the specifiable TRANSFORMATIONS. Factors
ERROR VARIANCE In a Set of Variables. initially are extracted to meet a mathe-
matical criterion of maximal variance
f a c t o r pattern m a t r i x in FACTOR explanation, which often does not result
ANALYSIS, a matrix of regression-like in a scientifically meaningful represen-
WEIGHTS that indicate the composition tation of the data. Indeed, for any one

Item

1. 1 am emotionally stable. -.19 -.58

2. Others consider me to be emotionally stable. -.61 -.19
3. 1 have a positive self-evaluation. .19 .89
4. Others evaluate me positively. .90 .21
5. I have an intemal locus of control. .16 .55
6. Others describe me as having an intemal .61 .10
locus of control.
7. 1 am self-efficacious. .19 .85
8. Others describe me as self-efficacious. .88 .27
9. My self-esteem is high. .17 .85
10. Others believe my self-esteem is high. .85 .20

factor pattern matrix

 false negative

factor solution thatfitsthe data to a spe- show the number of additional studies
cific degree there will exist an infinite that would need to be added (e.g., from
number of equally good mathematical unpublished FILE-DRAWER ANALYSIS) in
solutions, each represented by a differ- order to make the effect size non-
ent FACTOR STRUCTURE MATRIX. ThuS, significant. The larger the value of fail-
rotation is required to obtain a solution safe JV, the higher the credibility of the
that is both mathematically viable and estimated meta-analytic effect size.
logically sound. See OBLIQUE ROTATION;
ORTHOGONAL ROTATION. f a i l u r e time see SURVIVAL TIME.

f a c t o r score an estimate of the quanti- fairness n. the equitable treatment of

tative value that an individual would test takers in order to eliminate system-
have on a LATENT VARIABLE were it pos- atic variance in outcome scores among
sible to measure this directly; the latent people with different racial or cultural
variable or factor is determined through experiences and other background influ-
FACTOR ANALYSIS. ences. Fundamentally a sociocultural
(rather than a technical) issue, fairness
f a c t o r structure in FACTOR ANALYSIS, is a broad area encompassing quality
the pattern of FACTOR LOADINGS that management in test design, administra-
shows how a set of items correlate with tion, and scoring; adequate coverage
underlying latent variables. See FACTOR of relevant content; sufficient work to
STRUCTURE MATRIX. establish CONSTRUCT VALIDITY; equal
f a c t o r structure m a t r i x in FACTOR learning opportunities and access to
ANALYSIS, a matrix of FACTOR LOADINGS
testing; and items measuring only the
showing the correlations between items skill or ability under investigation with-
and underlying LATENT VARIABLES (fac- out being unduly influenced by con-
tors). This is in distinction to a FACTOR struct-irrelevant variance introduced
PATTERN MATRIX, whlch gives the corre- through test-taker background factors.
lations with a factor while taking into See also CULTURE-FAIR TEST.
account other factors. When examining faithful-subiect role the behavior
ORTHOGONAL or uncorrelated factors, adopted by a participant in a research
the factor pattern matrix and factor study who tries to respond accurately
structure matrix are the same, whereas and does not try to provide what
they are different when examining cor- he or she perceives as the researcher's de-
related factors in an OBLIQUE SOLUTION. sired response. Compare APPREHENSIVE-
f a c t o r theory in CLASSICAL TEST THE- SUBJECT ROLE; GOOD-SUBJECT ROLE;
ORY, the supposition that a set of corre- NEGATIVISTIC-SUBJECT ROLE.
lated variables can be reduced to a
false negative a case that is incorrectly
smaller set of underlying dimensions or
excluded from a group by the test used
FACTORS. It follows that each person's
to determine inclusion. In diagnostics,
observed score can be represented by a
for example, a false negative is an indi-
FACTOR LOADING timcs a factor plus
vidual who, in reality, has a particular
ERROR VARIANCE. condition but whom the diagnostic in-
fail-safe N a value often used in META- strument indicates does not have the
ANALYSIS to indicate the number of condition. In INFERENTIAL STATISTICS, a
nonsignificant studies that would be false negative is also referred to as a TYPE
needed to render a significant EFFECT II ERROR; this is the error that occurs
SIZE no longer significant. For example, when a study's findings indicate there is
if a researcher conducted a meta-analysis not a significant treatment effect when
and estimated a significant effect size, in fact there is such an effect. Compare
the fail-safe N could be calculated to FALSE POSITIVE.

131
false positive

f a l s e p o s i t i v e a case that is incorrectly genes and disease by examining the pat-

included in a group by the test used to tern of alleles (i.e., alternate forms of
determine inclusion. In diagnostics, for genes on paired chromosomes) that are
example, a false positive is an individual transmitted and those that are not trans-
who, in reality, does not have a particu- mitted to a child from both parents. See
lar condition but whom the diagnostic also FAMILY STUDY.
instrument indicates does have the con-
f a m i l y s t u d y research conducted
dition. In INFERENTIAL STATISTICS, a
among siblings, parents, or children to
false positive is also called a TYPE I
assess evidence for genetic links for char-
ERROR; this is the error that occurs when
acteristics or outcomes, often related to
a research study is thought to have sig-
health or disease. For example, a family
nificant findings when in fact it does
study might be conducted to assess
not. Compare FALSE N E G A T I V E .
whether individuals from the same fam-
falsifiability n. the condition of admit- ily who share a similar genetic structure
ting falsification: the logical possibility also have similar responses to a health-
that an assertion, hypothesis, or theory promotion intervention such as diet, ex-
can be shown not to be true by an obser- ercise, or medication. See also FAMILY-
vation or experiment. The most impor- BASED ASSOCIATION STUDY.
tant properties that make a statement
falsifiable in this way are (a) that it f a m i l y - w i s e a l p h a l e v e l a set value of
the probability of making a TYPE I ERROR
makes a prediction about an outcome or
a universal claim of the type "All Xs have when carrying out a set of hypothesis
property Y" and (b) that what is pre- tests (e.g., when conducting MULTIPLE
dicted or claimed is observable. COMPARISONS within a data set). In
Falsifiability is an essential characteristic many studies, family-wise alpha is kept
of any genuinely scientific hypothesis. at .05, such that there is only a 5% total
chance of rejecting a N U L L HYPOTHESIS
See also FALSIFICATIONISM. falsi-
fiable adj. when it should be retained over a set of
tests. See also FAMILY-WISE ERROR RATE.
falsificationism n. the position that
f a m i l y - w i s e e r r o r r a t e the probabil-
(a) the disproving, rather than proving,
ity of making a TYPE I ERROR when con-
of hypotheses is the basic procedure of
ducting MULTIPLE COMPARISONS among
scientific investigation and the chief
groups within a data set. Although simi-
means by which scientific knowledge is
lar to the FAMILY-WISE A L P H A LEVEL,
advanced; and (b) FALSIFIABILITY is the
which is a predetermined probability
property that distinguishes scientific
value to keep the amount of FALSE-
claims from truth claims of other kinds,
POSITIVE errors at a manageable level,
such as those of metaphysics or polifical
the family-wise error rate is assessed after
ideology. ^falsificationist adj.
a set of hypothesis tests have been con-
f a m i l y n. a collection of mathematically ducted. Researchers often like to keep
or statistically related entities. For exam- this value at .05. See also TESTWISE
ple, a set of statistical tests conducted ERROR RATE.
when there are more than two groups
f a n - S p r e a d h y p o t h e s i s a prediction
for an I N D E P E N D E N T V A R I A B L E within an
that differences between groups in
ANALYSIS OF VARIANCE Constitutes a
QUASI-EXPERIMENTAL RESEARCH will be
family of tests. See also FAMILY-WISE
proportional to the means and STAN-
ALPHA LEVEL; FAMILY-WISE ERROR RATE.
DARD DEVIATIONS of the individual
familial adj.
groups over time. In other words, when
f a m i l y - b a s e d a s s o c i a t i o n s t u d y re- the gap between two groups' mean
search assessing the connection between scores increases or decreases over time.
 field research

the variation of individual scores within CRITICAL VALUE of the F RATIO needed to
each group also increases or decreases. reject the NULL HYPOTHESIS. See F TEST.
The fan-spread hypothesis recognizes
feasibility study see PILOT STUDY.
that changes over time may get smaller
as individuals reach an ultimate goal. f i d e l i t y n. the degree of accuracy of a
For example, students in a program to measuring instrument or STATISTICAL
increase learning would not be expected MODEL. For example, a representation
to keep a constant rate of improvement derived from STRUCTURAL EQUATION
but rather to improve faster than a com- MODELING that depicts a pattern of rela-
parison group of students. tionships between health attitudes and
farthest neighbor see COMPLETE- behaviors could be said to have fidelity if
LINKAGE CLUSTERING.
it accurately explains the VARIATION
and COVARIATION in the data.
fatigue effect a decline in performance
on a prolonged or demanding research f i e l d n. somewhere other than a labora-
task that is generally attributed to the tory, library, or academic setting in
participant becoming tired or bored which experimental or NONEXPERI-
with the task. The fatigue effect is an im- MENTAL work is carried out or data col-
portant consideration when administer- lected.
ing a lengthy survey or test in which f i e l d experiment a study that is con-
participants' performance may worsen ducted outside the laboratory in a "real-
simply due to the challenges of an ex- world" setting. Participants are exposed
tended task. to one of two or more levels of an INDE-
FDA abbreviation for FUNCTIONAL DATA PENDENT VARIABLE and observed for
ANALYSIS. their reactions; they are likely to be un-
aware of the research. Such research
F d i s t r i b u t i o n a theoretical PROBABIL- often is conducted without RANDOM
ITY DISTRIBUTION widely used in the SELECTION or RANDOM ASSIGNMENT of
ANALYSIS OF VARIANCE, MULTIPLE RE- participants to conditions and no delib-
GRESSION, and other statistical tests of erate experimental manipulation of the
hypotheses about population variances. independent variable by the researcher.
It is the ratio of the variances of two in- See FIELD RESEARCH; QUASI-EXPERI-
dependent random variables each di- MENTAL RESEARCH.
vided by its DEGREES OF FREEDOM. In an
analysis of variance, for example, the F f i e l d notes notes on observations made
distribution is used to test the hypothe- in natural settings (i.e., the field) rather
sis that the variance between groups is than in laboratories. Field notes com-
significantly greater than the variance prise the data for subsequent analysis
within groups, thus demonstrating evi- in FIELD EXPERIMENTS and FIELD RE-
dence of some differences among the SEARCH.
means. Also called Fisher distribu-
f i e l d research studies conducted out-
tion; Fisher's F distribution; Fisher-
side the laboratory, in a "real-world" set-
Snedecor distribution; Snedecor's F
ting, which typically involve observing
distribution. See F RATIO; F TEST.
or interacting with participants in their
F d i s t r i b u t i o n table a table of values typical environments over an extended
listing the DEGREES OF FREEDOM for the period of time. Field research has the ad-
BETWEEN-GROUPS SUM OF SQUARES along vantages of ECOLOGICAL VALIDITY and
one side and the degrees of freedom for the opportunity to understand how and
the WITHIN-GROUPS S U M OF SQUARES why behavior occurs in a natural social
along another right-angle side, with val- environment; it has the disadvantages of
ues in the middle corresponding to the loss of environmental control and abil-

133
field survey

ity to do precise experimental manipula- large proportion of all studies actually

tions. Thus, field research is often said to conducted are not available for review
have more EXTERNAL VALIDITY and less because they remain unpublished in
INTERNAL VALIDITY than laboratory- "file drawers," having failed to obtain
based research. See also FIELD EXPERI- positive results. Thus, the results of a
MENT. meta-analysis may not yield reliable EF-
FECT SIZE estimates since only studies
f i e l d survey an assessment that in- that have been published or otherwise
volves collecting information on a spe- are widely available to researchers can be
cific topic in a relevant group of included in the analysis. See FILE-
individuals or entities, usually in their DRAWER ANALYSIS.
natural environment. For example, a field
survey could be conducted on a sample f i l l e r m a t e r i a l information added to a
of studentsfromunderrepresented groups test, questionnaire, or survey that is not
to assess their attitudes, experience, and related to the actual research, in order to
performance regarding quantitative me- keep individuals from discerning the
thods before providing an intervention tme nature of the study. The filler mate-
to increase quantitative reasoning. See rial usually is not analyzed as part of the
also FIELD RESEARCH. statistical findings.
figure n. a graph, drawing, or other de- f i l t e r n. any analytical procedure used in
piction used to convey the essential TIME-SERIES ANALYSIS to remove fluctua-
findings from a research study. Com- tions from the data and separate out its
mon figures used in psychological re- trend and cyclical components. A MOV-
search include BAR GRAPHS, which show ING AVERAGE is an example of a filter.
the frequency of endorsement for sev- f i l t e r coefficient a value needed to es-
eral categories (e.g., the number of indi- timate a process, as is common in SIG-
viduals who have various diseases), and NAL DETECTION THEORY and APPLIED
VENN DIAGRAMS, which use overlapping STATISTICS. Processes that have fewer fil-
circles to show how much shared vari- ter coefficients generally are easier to es-
ance there is between two or more vari- timate.
ables.
f i l t e r question a preliminary inquiry
file-drawer analysis a statistical pro- to assess whether a specific condition ex-
cedure for addressing the FILE-DRAWER ists that would require further assess-
PROBLEM by computing the number of ment. For example, a health researcher
unretrieved studies, averaging an EFFECT could use a filter question to determine
SIZE of .00, that would have to exist in if participants have ever experienced a
file drawers before the overall results of specific health concern, such as high
a META-ANALYSIS would become non- blood pressure. If the participant an-
significant at p > .05, that is, would ex- swers yes then the researcher could ask a
ceed an acceptable probability level (.05) subsequent set of questions to elicit
of occurring by chance. A small com- more information, such as the weight,
puted value indicates a finding may diet, exercise levels, and family history
have SPURIOUS PRECISION, such that it relating to high blood pressure of these
may just be due to chance. Conversely, a participants.
large value suggests that the finding is
rather robust, as it would take a large f i n i t e m i x t u r e d i s t r i b u t i o n a data
number of nonsignificant findings to set that is believed to describe several
provide enough evidence to refute the fairly distinct and limited clusters or
results. subgroups of individuals, even if the ac-
tual clusters are not completely observ-
file-drawer p r o b l e m the fact that a able. For example, a health researcher
 first-order partial correlation

may collect data from a sample of people score is believed to be related to

to assess their frequency of unprotected the immediately previous score plus
sex, their number of sexual partners, and some RANDOM ERROR. See also AUTO-
the perceived risk of sexual partners. The REGRESSIVE MODEL.
ultimate goal of the research could be to
uncover, through a procedure such as first-order factor in FACTOR ANALY-
CLUSTER ANALYSIS Or LATENT CLASS SIS, any of the LATENT VARIABLES (factors)
ANALYSIS, several distinct clusters of in- that are derived from the CORRELATION
dividuals at varying risk for HIV, such as (or covariance) among the MANIFEST
a low-risk group, whose members have VARIABLES, as opposed to SECOND-
varying frequency of unprotected sex ORDER FACTORS, which are determined
and a monogamous sexual relationship from the correlation (or covariance)
of low perceived sexual risk; a moderate- among the factors. Also called primary
risk group, whose members have moder- factor.
ate frequency of unprotected sex and first-order i n t e r a c t i o n an effect in
several sexual partners of varying per- which the pattern of values on one vari-
ceived partner risk; and a high-risk able changes depending on the combi-
group, whose members have high values nation of values on two other variables.
for all three variables. From this analysis, For example, an analysis could reveal
the researcher may conclude that the that although gender and teaching style
sample contained a finite mixture distri- may each have some effect on perfor-
bution of three clusters of sexual risk, mance, the specific degree or amount of
where each cluster most likely has differ- the performance effect changes depend-
ent MEAN and VARIANCE values. ing on the particular combination of
gender and teaching method. Thus,
f i n i t e p o p u l a t i o n a limited universe
male students may show moderately
of individuals or entities from which a
high performance regardless of teaching
researcher may wish to sample for a
style, whereas female students may
study. For example, an educational psy-
show high performance with a hands-on
chologist may consider the finite popu-
teaching style and low to moderate per-
lation of elementary-school-age children
formance with a lecture teaching style.
in a limited geographical area (e.g., city,
See also HIGHER ORDER INTERACTION.
county, state) to recruit participants for
a study of learning and attention deficit first-order M a r k o v model a statisti-
disorders. cal model in which each value in a
f i n i t e p o p u l a t i o n correction an ad- MARKOV CHAIN Is described in terms of
justment made to a PARAMETER esti- its relation to the immediately previous
mate, often a VARIANCE, when sampling value. A second-order Markov model
from a limited subset of a larger defined would be required to describe the cur-
POPULATION. For example, using N - l rent value in the chain in relation to the
in the denominator of a variance calcu- two previous stages, and so on.
lation, instead of just N, could be seen as first-order p a r t i a l correlation the
a finite population correction when esti- association between two variables, x and
mating the variance of intelligence in a y, after controlling for or taking into ac-
sample of students drawn from a FINITE count the effect of one other variable, z.
POPULATION. In other words, a first-order partial corre-
f i r s t moment see MOMENT. lation is the relationship between the
RESIDUAL from X and the residual from y
first-order autoregressive model a after predicting each of x and y with an-
statistical process, often examined using other variable, z. The actual quantitative
TIME-SERIES ANALYSIS, In which each value describing the relationship is

135
first passage time

called the first-order partial correlation John H. Halton, British-born U.S. com-
coefficient. Also called first-order par- puter scientist]
tial. See also PARTIAL CORRELATION.
Fisher F test see F TEST. [Sir Ronald
f i r s t passage time the amount of time Aylmer Fisher]
that elapses until the occurrence of a Fisher g test a statistical procedure for
random process in an entity or individ- analyzing whether there is any differ-
ual. For example, a medical researcher ence between the patterns of observed
may be interested in the first passage and expected CATEGORICAL DATA. The
time for a specific disease in a specific pa- Fisher g test is a LIKELIHOOD-RATIO TEST
tient. that is approximately distributed as a
CHI-SQUARE and can be used in the same
f i r s t q u a r t i l e see QUARTILE. situations. [Sir Ronald Aylmer Fisher]
Fisher-Behrens problem see BEHRENS- Fisher-Hayter m u l t i p l e compari-
FISHER PROBLEM. son test a statistical procedure for as-
sessing whether pairs of means are
Fisher d i s t r i b u t i o n see F DISTRIBU-
significantly different that is used after a
TION. [Sir Ronald Aylmer Fisher (1890-
significant F TEST. The Fisher-Hayter
1962), British statistician and geneticist]
multiple comparison test is a modifica-
Fisher exact test a statistical procedure tion of the FISHER LEAST SIGNIFICANT
to determine whether two CATEGORICAL DIFFERENCE TEST that controls for FAM-
variables are related. Appropriate for ILY-WISE ERROR RATE; it is relatively easy
small samples (a CHI-SQUARE TEST FOR to calculate. See POST HOC COMPARISON.
INDEPENDENCE is used with large sam- [Sir Ronald Aylmer Fisher; Anthony J.
ples), it examines the associations be- Hayter, British-born U.S. statistician]
tween the rows and columns of data in a Fisher least significant difference
fourfold (2 X 2) CONTINGENCY TABLE. For test (Fisher LSD test; LSD test) a sta-
example, the Fisher exact test could be tistical procedure to compare pairs of
used to assess whether there is a relation- means, conducted after an F TEST has re-
ship between gender (male or female) vealed that at least one pair of means is
and pizza topping preference (cheese or significantly different. The test calcu-
multitopping) in a group of 20 individu- lates the smallest value that would be
als. The test yields an exact p value statistically different from chance when
rather than a range of p values (e.g., p < subtracting one mean from another
.05). Also called Fisher-Irwin test; mean. If the absolute value of the actual
Fisher-Yates test. [Sir Ronald Aylmer difference between a pair of means is
Fisher; Joseph Oscar Irwin (1898-1982), larger than this least significant differ-
British statistician; Frank Yates (1902- ence (LSD), a researcher can reject a
1994), British statistician] NULL HYPOTHESIS that the means are
equal and conclude that they are signifi-
Fisher-Freeman-Halton test an ex- cantly different. Also called protected
tension of the FISHER EXACT TEST used to t test. [Sir Ronald Aylmer Fisher]
assess whether two variables, each hav-
ing two or more categories, are related in Fisher scoring method a procedure
a relatively small data sample. For exam- for finding a MAXIMUM LIKELIHOOD so-
ple, the Fisher-Freeman-Halton test lution when estimating PARAMETERS,
could be used to assess whether gender such that the resulting estimates are se-
and pizza topping preference are related lected to make the sample data have the
when the latter has several choices (e.g., highest probability of being drawn from
cheese, vegetable, meat, everything). [Sir a population with the given estimates.
Ronald Aylmer Fisher; G. H. Freeman; [Sir Ronald Aylmer Fisher]

136
 fit index

Fisher's F d i s t r i b u t i o n see F DISTRI- SIS is tme (i.e., assuming the means are
BUTION. [Sir Ronald Aylmer Fisher] equal). Thus, the null hypothesis can be
rejected and a researcher can conclude
Fisher's i n f o r m a t i o n a variance or that there is evidence that the two
covariance, often of M A X I M U M LIKELI- means differ more than would occur by
HOOD estimates. When the variances chance. [Sir Ronald Aylmer Fisher]
and covariances are organized together
in a SQUARE MATRIX the resulting dis- Fisher's r to z t r a n s f o r m a t i o n a sta-
play is called Fisher's information ma- tistical procedure that converts a Pear-
trix. [Sir Ronald Aylmer Fisher] son PRODUCT-MOMENT CORRELATION
COEFFICIENT to a Standardized z SCORE
Fisher's least s i g n i f i c a n t d i f f e r - in order to assess whether the correla-
ence see FISHER LEAST SIGNIFICANT DIF- tion is statistically different from zero.
FERENCE TEST.
The test is useful in providing a nor-
Fisher's linear d i s c r i m i n a n t func- mally distributed statistic (called the
t i o n an additive combination of con- Fisher transformed value or Fisher's z)
tinuous variables that helps explain a that can be used in HYPOTHESIS TESTING
CATEGORICAL outcome. For example, an or in forming a CONFIDENCE INTERVAL.
organizational psychologist could ex- Also called Fisher transformation;
amine whether measures of perceived Fisher z transformation; z trans-
career influence, work respect, and insti- formation. [Sir Ronald Aylmer Fisher]
tutional climate distinguish among
three different departments within the Fisher-Yates test see FISHER EXACT
institution. A Fisher's linear discrimi- TEST. [Sir Ronald Aylmer Fisher; Frank
nant function analysis would reveal two Yates (1902-1994), British statistician]
linear discriminant functions in which f i s h i n g n. an informal term for search-
the discriminant coefficients associated ing unsystematically through a data set
with each predictor variable for each in order to look for any noticeable pat-
function indicate how strongly that terns of relationship among variables.
variable relates to the set of depart- More formal procedures called DATA
ments. See DISCRIMINANT FUNCTION. MINING have been developed using sta-
[Sir Ronald Aylmer Fisher] tistical and computer-science methods
to discern recognizable patterns in huge
Fisher-Snedecor d i s t r i b u t i o n see F
data sets. Fishing should not substitute
DISTRIBUTION. [Sir Ronald Aylmer Fisher;
for more rigorous methods based on
George W. Snedecor (1881-1974), U.S.
sound EXPERIMENTAL DESIGN. See also
statistician]
DATA DREDGING; DATA SNOOPING.
Fisher's randomization procedure
f i t n. the degree to which values pre-
a statistical test that compares the actual
dicted by a model correspond with em-
difference between sample means for
pirically observed values. For example,
two distinct groups, with a set of mean
in STRUCTURAL EQUATION MODELING, a
differences obtained by randomly split-
researcher may want to see how well
ting the combined data from the two
groups into all possible rearrangements his or her hypothesized model of the
of two arbitrary groups. The proportion relationships ambng a set of variables ac-
of mean differences from the various tuaUy fits the VARIATION and COVARI-
sets of randomly split data is viewed as a ATION in the data.
p VALUE, such that if the proportion is f i t index a quantitative measure of how
less than, say, .05, it indicates that there well a statistical model corresponds to
is a less than 5% chance of finding the VARIATION and COVARIATION in a
means as different or more so than the set of data. For example, in STRUCTURAL
two sample means if the NULL HYPOTHE- EQUATION MODELING, a well-recognized

137
fitted distribution

fit index is the Bentier comparative fit in LINEAR REGRESSION the fitted value is
index (CFI), which indicates how well a y', where y' = a + bx, in which a equals
proposed model fits the data compared the Y-INTERCEPT, b equals the unstan-
to a null model that just posits variances dardized REGRESSION COEFFICIENT, and
for the variables and no relationships A; is a predictor variable. Note that sub-
among the variables. An index value tracting a fitted value from an actual y
greater than .90 (or preferably .95) indi- score in linear regression will equal a FIT-
cates a model that explains the pattern TED RESIDUAL, which will be close to
of relationships in the data reasonably zero if the linear regression model is a
well. If the fit of a model is poor, then good match to the data.
the model needs to be respecified and
five-number s u m m a r y in EXPLOR-
then reanalyzed. See GOODNESS OF FIT.
ATORY DATA ANALYSIS, the Characteriza-
See also BENTLER-BONETT INDEX.
tion of a data set through the use of five
f i t t e d d i s t r i b u t i o n a known statistical summary statistics: the two extreme
pattern of data (e.g., a NORMAL DISTRI- scores, the upper and lower QUARTILES,
BUTION) that is compared to sample and the MEDIAN or middle quartile. For
data. To assess whether afitteddistribu- example, suppose the scores on a test are
tion is a good match for the data, one (a) 100, 93, 90, 82, 76, 72, 64, 61, 60, and
plots the sample data with a HISTOGRAM 47. The five-number summary from
or BAR GRAPH; (b) compares the sample highest to lowest numbers would be
graph with the known statistical pattern 100, 90, 74, 61, and 47that is, 100 is
(e.g., using a program designed for that the highest extreme score, 90 is the
purpose); and (c) assesses how closely highest quartile (the midpoint in the top
the sample data and statistical pattern half of the data), 74 is the middle point,
match using a GOODNESS-OF-FIT statistic 61 is the lowest quartile (the midpoint in
(e.g., CHI-SQUARE). If the fitted distribu- the bottom half of the data), and 47 is
tion describes the sample data fairly the lowest extreme score. These values
closely, the goodness-of-fit statistic provide the basic highlights of a set of
should be small and nonsignificant, in- data and can be used to create a BOX-
dicating there is little difference between AND-WHISKER PLOT.
the statistical pattern (e.g., normal dis- fixed-alternative question a test or
tribution) and the sample data pattern survey item in which several possible re-
(depicted in the initial histogram). sponses are given and participants
f i t t e d frequency see EXPECTED FRE- are asked to pick the correct response or
QUENCY. the one that most closely matches their
preference. An example of a fixed-
f i t t e d residual the value obtained alternative question is "Which of the
when subtracting an expected or pre- following most closely corresponds to
dicted value from the original data your age: 12 or younger, 13 to 19, 20 to
value. For example, in LINEAR REGRES- 39, 40 to 59, 60 to 79, or 80 or older?" A
SION the fitted residual equals y - y', fixed-alternative question is sometimes
where y' is the predicted value formed referred to as a closed question, al-
from filling actual x values into the re- though this can also refer to any inquiry
gression equation; and y' = a + bx, where requesting a short definite answer (e.g.,
a equals the Y-INTERCEPT, b equals the "How old are you?"). Also called fixed-
unstandardized REGRESSION COEFFI- choice question; forced-choice ques-
CIENT, and jit is a predictor variable. tion; multiple-choice question. Com-
pare FREE-RESPONSE QUESTION.
f i t t e d value the predicted value of a
particular score where a statistical model f i x e d effect an INDEPENDENT VARIABLE
is compared to actual data. For example, whose levels are specified by the re-

us
 flowchart

searcher rather than randomly chosen for study rather than randomly chosen
within some level of permissible values. from a wide range of possible values. For
For example, a health researcher who' example, a researcher may wish to inves-
specifically chose to examine the effect tigate the effects of the available dosages
on weight loss of no exercise, one hour of a certain drug on symptom allevia-
of exercise, or three hours of exercise per tion. Fixed-effects models generally are
week would be treating time spent exer- intended to make inferences solely
cising as a fixed effect. In other words, all about the specific levels of the inde-
levels of interest are included in the de- pendent variables actually used i n the
sign and thus anyone wanting to repli- experiment. Compare MIXED-EFFECTS
cate the study would have to use the MODEL; RANDOM-EFFECTS MODEL.
same levels of exercise as in the original.
f i x e d f a c t o r see FIXED EFFECT.
As a rule, one should not generalize re-
sults of fixed-effect studies beyond the f i x e d p a r a m e t e r a specific value as-
specific levels or conditions used in the signed (as opposed to estimated) by a re-
experiment (see FIXED-EFFECTS F A L L A C Y ) . searcher when testing a statistical
Also called f i x e d factor. See FIXED- model. For example, in STRUCTURAL
EFFECTS MODEL; RANDOM-EFFECTS MODEL. E Q U A T I O N M O D E L I N G , researchers may
Compare R A N D O M EFFECT. use a fixed parameter of 1.0 for one of
the F A C T O R L O A D I N G S or variances of
fixed-effects analysis o f variance a
each LATENT VARIABLE in a model. Com-
statistical procedure to determine
pare FREE PARAMETER.
whether means for an outcome or D E -
PENDENT VARIABLE differ across a spe- f i x e d p o p u l a t i o n a specific range of
cific set of conditions (i.e., levels of the individuals or entities that constitute
INDEPENDENT VARIABLE) that have been the larger base of interest to a researcher.
selected by a researcher. This is con- See POPULATION.
trasted with a RANDOM-EFFECTS ANALY-
SIS OF VARIANCE, i n whlch the f i x e d v a r i a b l e a variable whose value is
conditions have been randomly selected specified by a researcher or otherwise
from a wide range of possible choices be- predetermined and not the result of
fore any mean differences are examined. chance. Compare R A N D O M VARIABLE.

f i x e d - e f f e c t s f a l l a c y a situation in f l o o r e f f e c t the situation in which a

which researchers inappropriately gen-
eralize beyond the specific conditions
they tested in an analysis, treating a
study design as if it were based on ran-
domly selected conditions that would
more appropriately allow generaliza-
tion. For example, a fixed-effects fallacy whose items are too difficult for those
occurs when a researcher uses a FIXED-
EFFECTS ANALYSIS O F V A R I A N C E but
states his or her results as if he or she had
used a R A N D O M - E F F E C T S A N A L Y S I S O F
V A R I A N C E , in which the conditions ex-
amined were selected from a large popu-
lation of possible condition options.
f i x e d - e f f e c t s m o d e l any statistical
procedure or experimental design that
uses independent variables whose levels
are specifically selected by the researcher
large proportion of participants perform
very poorly on a task or other evaluative
measure, thus skewing the distribution
of scores and making it impossible to dif-
ferentiate among the many individuals
at that low level. For example, a test
taking it would show a floor effect be-
cause most people would obtain or be
close to the lowest possible score of 0.
Also called basement effect. Compare
C E I L I N G EFFECT.
f l o w c h a r t n. a diagram representing the
ordered steps to follow when conduct-
ing a procedure, such as constructing a
computer program or performing a spe-
ciflc research study. The illustrafion over-
leaf gives a basic example.

139
fluctuation

Stepi

Jl

flowchart

Flowcharts also are used in PATH ANAL- test; Hartley f^^x test for homoge-
YSIS and STRUCTURAL EQUATION MOD- neity of variance; Hartley test;
ELING to illustrate the causal links Hartley test for homogeneity of
among the variables being studied. variance.
f l u c t u a t i o n n. 1. variation in size or focused comparison in research, the
value. 2. in TIME-SERIES ANALYSES, an ir- practice of putting the same set of care-
regularity that is not related to any trend fully chosen, strictly relevant questions
and is thus often removed via SMOOTH- to each participant in an experiment (or
ING. to the same participant on multiple oc-
Fmax d i s t r i b u t i o n in an ANALYSIS OF casions in a WITHIN-SUBJECTS DESIGN).
VARIANCE, a pattern of values used for This allows for systematic comparison of
assessing whether variances are homo- the answers given and produces more
geneous across the groups being studied. generalizable and replicable findings
See F,^x STATISTIC. than does asking questions that vary
from case to case.
Fmax statistic in an ANALYSIS OF VARI-
ANCE, a value formed from the ratio of focus group a small set of people, typi-
the largest variance over the smallest cally 8 to 12 in number, who share com-
variance from the set of groups being as- mon characteristics (e.g., working
sessed. The statistic is assessed for parents with 5- to 8-year-old children)
significance by looking up the CRITICAL that are relevant to the research ques-
VALUE in a statistical table for the rele- tion and who are selected to discuss a
vant number of groups and DEGREES OF topic of which they have personal expe-
FREEDOM. If the calculated F^^,^ statistic rience (e.g., their children's reading abil-
is larger than the critical ratio, then ities and school performance). A leader
there is evidence of HETEROGENEITY OF conducts the discussion and keeps it on
VARIANCE, which Violates one of the as- target while also encouraging free-
sumptions of the analysis and renders its flowing, open-ended debate. Originally
results invaUd. Otherwise, a researcher used in marketing to determine con-
can assume that there is not enough evi- sumer response to particular products,
dence for heterogeneity and the assump- focus groups are now used for determin-
tion of HOMOGENEITY OF VARIANCE Can ing typical reactions, adaptations, and
be reasonably maintained. solutions to any number of issues,
events, or topics and are associated par-
Fmax test the procedure used to obtain ticulariy with QUALITATIVE RESEARCH.
an F^^Ax STATISTIC and evaluate it for sig-
nificance. Also called Hartley F^^^ foldover design a type of FRACTIONAL

140
 foreperiod

FACTORIAL DESIGN in which the re- f o l l o w - u p study a long-term research

searcher deliberately removes INTERAC- project designed to examine the degree
TION EFFECTS among the variables. This to which effects seen shortly after the
is achieved by carrying out a second mn imposition of an intervention persist
of tests in which the elements in the DE- over time. Follow-up studies are also
SIGN MATRIX are multiplied by -1. The used for the long-term study of partici-
results from the two runs of test are then pants in a laboratory experiment to ex-
combined to ehminate the interaction amine the degree to which effects of the
effects, such that the MAIN EFFECTS be- experimental conditions are lasting.
come the primary focus. A foldover de-
sign may be useful when there are f o l l o w - u p test see POST HOC TEST.
several variables that each have two or forced-choice question see FIXED-
more levels, such that it is difficult to an- ALTERNATIVE QUESTION.
alyze the full set of possible conditions.
forced d i s t r i b u t i o n a rating system in
follow-back study research that col- which raters must make a prescribed
lects earlier data in order to understand number of entries for each level of the
the causes of an event of interest. For ex- rating scale used. For example, in em-
ample, a researcher might conduct a fol- ployee evaluation a forced distribution
low-back study of deceased hospital might be used in which it is required
patients, gathering additional infor- that 5% of employees are categorized as
mation regarding cause of death so as to poor, 15% as below average, 60% as av-
understand the progress of different dis- erage, 15% as above average, and 5% as
eases. excellent.
f o l l o w - t h r o u g h n. an exercise in forecast n. a prediction about a future
which researchers provide summaries of event based on current status and
their main findings in order to provide past history. For example, TIME-SERIES
practical guidance to policymakers, ANALYSES oftcn are used to create statis-
funding agencies, or professionals in the tical forecast models that provide a best
relevant fields. guess about what to expect for variables
of interest.
f o l l o w - u p plot (FU plot) a type of
. graph, mainly used in LONGITUDINAL foreperiod n. in reaction-time experi-
DESIGNS, in which the specific results for ments, the pause or interval between the
specific individuals are displayed over "ready" signal and the presentation of
time. Below is a hypothetical example. the stimulus.
140

Participant 1
Participant 2
Participant 3

20 40
Time (months)
follow-up plot

141
forest plot

Author A, 1999

Author B, 2001

Authors C & D, 2001

Authors E & F, 2002

Author G, 2004

Author H, 2007

Authors I & J, 2008

Odds Ratio

forest plot

forest plot a graph, often used in META- various behaviors in mathematical

ANALYSES, in which the authors or titles terms. Formal theories often are devel-
of the studies are Usted along the left oped and studied in the field of political
side, and the corresponding EFFECT SIZE science and psychology. An example of
for each is given on the right. Consider a formal theory is the so-called "pris-
the example above, in which effect sizes oner's dilemma" model, which can be
are given as ODDS RATIOS. applied to various situations in which
The effect sizes for the individual studies participants have to decide between the
are usually indicated by squares in which costs and benefits of cooperating or
the size of the square is proportional to competing. The model uses a series of
the size of the effect for a specific study. rules to determine the outcomes of vari-
At the bottom the overall effect size is ous behaviors where two suspects sepa-
given, usually in a diamond shape whose rately have the option to confess or
width indicates the outer points of a refuse to speak about a potential crime.
CONFIDENCE INTERVAL for the Overall ef- If both refuse to speak, they both get a
fect across all of the studies listed. minor penalty (e.g., 1 month in jail). If
both confess, they both get a moderate
f o r m a l experimentation a strictly penalty (e.g., 3 months in jail). If one
controlled study in which the INDE- confesses and the other refuses to speak,
PENDENT VARIABLE is manipulated and the one who confesses goes free and the
one group of individuals or entities is other gets a severe penalty (e.g., 1 year in
randomly assigned to receive the TREAT- jail).
MENT and another set of individuals or
entities is randomly assigned to act as a f o r m a t i v e evaluation a process in-
CONTROL GROUP. Results from formal tended to improve or guide the develop-
experimentation can attribute a CAUSAL ment of a program in its early stages
INFERENCE to the independent variable through the use of qualitative or quanti-
if those in the treatment group perform tative research methodology. Ideally,
better than those in the control group. the individual performing a formative
evaluation wUl repeatedly interact, often
f o r m a l theory a model or set of rules informally, with the program personnel
and assumptions used to understand from the outset of the work to clarify

142
 fractional polynomial

goals, monitor implementation, and as- f o r w a r d selection a technique used in

sess staff and resource requirements. See creating MULTIPLE REGRESSION models
also SUMMATIVE EVALUATION. in which independent variables from a
large set of such variables are added to
f o r m u l a n. a precise statement about the REGRESSION EQUATION in the order
how two pieces of information, usually of their predictive power (i.e., largest to
mathematical, are related. The most smallest increase in the COEFFICIENT OF
common type of formula is a mathemat- MULTIPLE DETERMINATION) until a pre-
ical EQUATION in which one variable is a set criterion is reached and there is
weighted FUNCTION of another variable no further significant change in the
plus a constant. For example, the for- model's predictive power. Also called
mula for a straight line is y = mx + b, forward inclusion; f o r w a r d step-
where y is an outcome plotted on a verti- wise regression; stepup selection.
cal or y-axis, m is the slope of the line, x See also F-TO-ENTER; F-TO-REMOVE.
is the value of a variable along the hori-
zontal or A-axis, and b is where the f o u r f o l d p o i n t correlation coeffi-
straight line crosses the vertical axis. cient see PHI COEFFICIENT.
f o u r f o l d table see TWO-BY-TWO TABLE.
f o r m u l a scoring a method of scoring a
multiple-choice test that includes a COR- f o u r t h moment see MOMENT.
RECTION FOR GUESSING. Participants are
fractal n. a geometric shape that can be
encouraged to respond to an item if they
divided into parts that each resemble the
can rule out at least one response choice,
pattern of the whole shape. In other
but they are discouraged from respond-
words, a fractal is a shape possessing the
ing if they do not have any idea of the
quality of self-similarity.
correct answer. A participant's score is
determined by noting the number of f r a c t i l e n. see QUANTILE.
items he or she got right (R) and sub-
f r a c t i o n a l f a c t o r i a l design a type of
tracting the number of wrong items (W)
experimental design in which some con-
divided by the response choices per item
ditions are omitted, such that not all lev-
minus 1 ( C - 1). For example, if a partici-
els of one or more of the INDEPENDENT
pant had 40 right answers and 10 wrong
VARIABLES are combined with all other
answers on a 50-item test that has four
levels of the other variables. A LATIN
response choices per question (e.g., A, B,
SQUARE is an example. Fractional facto-
C, or D), that person's formula score (FS)
rial designs might be used because of a
would be
small overall sample available for study
FS = R - [W/(C- J)] = 40- 10/3 = 36.67. or because of difficulty obtaining partici-
pants or assigning them to some condi-
Note that without the correction, the tions. Additionally, there may simply be
score would have been 40. Thus, the for- too many combinations to study (e.g., in
mula score assumes that the participant a 3 x 4 x 2 x 5 design, there are 120 com-
guessed among three choices after ruling binations of levels). Fractional factorial
out one of them on each of the 10 items designs can be used to examine MAIN EF-
that were missed. This would mean that FECTS (and sometimes TWO-WAY INTER-
he or she would get 1/3 correct purely by ACTIONS) but cannot be used to assess
guessing, hence this amount (.333 of 10 any three-way or other HIGHER ORDER
or 3.33) is subtracted from the score of INTERACTIONS. Also called incomplete
40 to give the corrected formula score of factorial design. Compare COMPLETE
36.67.
FACTORIAL DESIGN.

f o r w a r d - l o o k i n g study see PROSPEC- f r a c t i o n a l p o l y n o m i a l a transforma-

TIVE RESEARCH. tion of a continuous variable to a qua-

143
frailty model

dratic, cubic, or further power when the study at any time without penalty. See
relationship is believed to be NONLIN- also FREEDOM FROM HARM; PARTICI-
EAR. For example, if a researcher expects PANTS' RIGHTS.
a relationship between achievement
motivation and performance that re- Freeman-Tukey test a GOODNESS-OF-
sembles an upside-down U (CURVI- FiT TEST used with data that have multi-
LINEAR) shape, a fractional polynomial
ple categories of counts or frequencies.
could be formed by squaring the [M. F. Freeman, U.S. statistician; John
achievement motivation score for each Wilder Tukey (1915-2000), U.S. statisti-
individual in order to predict achieve- cian]
ment more accurately as a nonlinear free parameter a value that is esti-
pattern. mated from data, usually in a modeling
procedure such as STRUCTURAL EQUA-
f r a i l t y model a form of SURVIVAL
TION MODELING. A free parameter con-
ANALYSIS that allows for HETEROGENE-
trasts with a FIXED PARAMETER, whlch is
ITY in the sample. It incorporates a RAN-
kept at a known, specific value. For ex-
DOM EFFECT to account for COVARIATES
ample, in CONFIRMATORY FACTOR ANAL-
across different subsets of participants in
YSIS, one FACTOR LOADING for each
the sample.
LATENT VARIABLE Is usually a fixed pa-
F ratio (symbol: F) in an ANALYSIS OF rameter of 1.0 and the other loadings are
VARIANCE or a MULTIVARIATE ANALYSIS free parameters that are estimated.
OF VARIANCE, the amount of EXPLAINED
free-response question a test or sur-
VARIANCE divided by the amount of
vey item that allows the respondent to
ERROR VARIANCE; that is, the ratio of be-
respond entirely as he or she pleases, as
tween-groups variance to within-group
opposed to a FIXED-ALTERNATIVE QUES-
variance. Its value determines whether
TION, in which the respondent must
or not to accept the NULL HYPOTHESIS
choose from several provided responses.
stating that there is no difference be-
tween the treatment and control condi- frequency (symbol: f ) n. the number
tions, with a large value indicating the of occurrences of a particular phenome-
presence of a significant effect. Ideally, a non, particularly a CATEGORICAL VARI-
researcher prefers to have rather small ABLE such as gender. For example, it is
variation within each group and maxi- often of interest to find the frequencies
mal variation between the groups in or counts of the men and women who
order to demonstrate significant group are participating in a research study.
differences. Also called F statistic; F Also called absolute frequency. See
value. also RELATIVE FREQUENCY.
freedom f r o m h a r m one of the basic frequency data information that rep-
rights of research participants that is en- resents the counts or number of occur-
sured by an INSTITUTIONAL REVIEW rences of particular response classes,
BOARD. Freedom from harm states that a usually for a CATEGORICAL VARIABLE.
research participant should not incur For example, a political psychologist
undue risk as a result of taking part in a may want to record frequency data on
study. See also FREEDOM TO WITHDRAW; the number of men and women, the
PARTICIPANTS' RIGHTS. number of different ethnic groups, and
the number of individuals who endorse
freedom to w i t h d r a w one of the a Democratic, Republican, or other po-
basic rights of research participants that litical party affiliation.
is ensured by an INSTITUTIONAL REVIEW
BOARD. Freedom to withdraw allows a frequency density the number of oc-
research participant to drop out of a currences of an event divided by the

144
 frequentist inference

class size for a particular CATEGORICAL midpoints of the bars of a HISTOGRAM is

VARIABLE of interest. a frequency polygon, as in the hypothet-
frequency f l i s t r i b u t i o n a tabular rep- ical example below.
resentation of the number of times a frequency test a statistical procedure
specific value or datum point occurs. for assessing data that contain counts or
The left column lists the different cate- the numbers of occurrences of various
gories of a CATEGORICAL VARIABLE or categories or classes. A common exam-
scores df a CONTINUOUS VARIABLE, and ple of a frequency test is a CHI-SQUARE
the right column lists the number of oc- TEST that compares the pattern of ob-
currences of each. For example, one served counts or frequencies to those
could construct a frequency distribution that are expected to occur.
of the variable gender for a sample of 40
women and 60 men, as depicted below. frequency-within-interval record-
i n g the process of tracking the number
Gender Frequency ( f )
of occurrences of an event or behavior
across various time periods. It is a com-
monly used tool when working with
Female 40
clinical populations (e.g., those with au-
Male 60 tism) on behavioral change.
frequentist n. a researcher who ap-
proaches issues of PROBABILITY in terms
When a frequency distribution is plot- of the frequency (number of occur-
ted on a graph, it is often called a fre- rences) for a particular parameter over a
quency curve, frequency diagram, or period of time. This approach is con-
FREQUENCY POLYGON. When represented trasted with that of a BAYESIAN, who ex-
mathematically via an equation, it is amines the data and assesses whether
called a frequency function. Also called particular values are more credible or be-
frequency table. See also CUMULA- lievable than others.
TIVE FREQUENCY DISTRIBUTION; RELA-
TIVE FREQUENCY DISTRIBUTION. frequentist inference an approach to
drawing conclusions from statistical
frequency polygon a graph depicting samples that is based on the number of
a statistical distribution, made up of times an event is expected to occur in
lines connecting the peaks of adjacent the long mn if the conditions for observ-
intervals. A LINE GRAPH connecting the ing the event are held, constant. It con-

o o
o o o o o o o o o o o o o o o
o o o o o o
o oo_oo oo oooooooo_o o_
ooooo o
o o
o o o
ooo oo oo oo o
t - Cv[ CO lO CD h-" co" Oi O T-" CN CO Tt If) (D r^" (30" 05 O

Debt (dollars)
frequency polygon
145
Friedman two-way analysis of variance

siders any research study to be one of a F-to-enter for any of the remaining vari-
very large possible number of replica- ables was at least 4.0 and if so would add
tions. Also called classical inference. that variable into the analysis as well.
Compare BAYESIAN INFERENCE. When the researchers found an F-to-
enter for a variable less than 4.0, they
Friedman two-way analysis of would stop the process and examine
variance a NONPARAMETRIC TEST of only those variables that had been
whether two categorical INDEPENDENT added to that point. Compare F-TO-
VARIABLES have a consistent relation-
REMOVE.
ship with a DEPENDENT VARIABLE that
involves ORDINAL DATA. For example, a F-to-remove n. in model-building pro-
researcher might use the Friedman two- cedures such as BACKWARD ELIMINATION,
way analysis of variance to determine STEPWISE REGRESSION, Or backward STEP-
whether three different tests produce WISE DISCRIMINANT ANALYSIS, the Spe-
consistent outcomes when they are used cific ratio of variances needed to justify
to rank the performance of students un- keeping a variable as a predictor. Gen-
dergoing three different training pro- erally, an F-to-remove value around 4.0
grams. Also called Friedman rank is sufficient to allow a variable to be re-
test; Friedman test. [Milton Fried- tained in an analysis. For example, if re-
man] searchers want to assess which of a set of
six variables are the most important in
F statistic see F RATIO. predicting an outcome, they could con-
F test any of a class of statistical proce- duct a backward stepwise multiple re-
gression. All six variables initially would
dures, such as ANALYSIS OF VARIANCE Or
be included in the regression analysis,
MULTIPLE REGRESSION, that rely on the
providing an index of the amount of
assumption that the calculated statis-
variance explained by the set of six vari-
ticthe F RATIOfollows the F DISTRI-
ables. Then, the variable that had the
BUTION when the null hypothesis is
smallest F-to-remove value (i.e., less
true. F tests are tests of hypotheses about
than 4.0) would be removed from the
population variances or of whether RE-
analysis. The researchers would con-
GRESSION COEFFICIENTS are zero. Also
tinue by evaluating each of the remain-
called Fisher's F test; variance ratio
ing five variables and dropping any
test.
that had an F-to-remove less than 4.0,
F-to-enter n. in model-building proce- such that only the retained variables ul-
dures such as FORWARD SELECTION and timately are examined in the analysis.
forward STEPWISE DISCRIMINANT ANALY- Compare F-TO-ENTER.
SIS, the specific ratio of variances needed
f u g i t i v e literature see GRAY LITERA-
to justify adding a variable as a predic-
tor. Generally, an F-to-enter value TURE.
around 4.0 is sufficient to allow a vari- f u l l model a statistical representation
able to be included in an analysis. For that includes all of the variables of inter-
example, if researchers want to assess est, or all of the main PARAMETERS (e.g.,
which of a set of six variables are the REGRESSION COEFFICIENTS), among a set
most important in predicfing an out- of variables. For example, in STRUC-
come, they could conduct a STEPWISE TURAL EQUATION MODELING, a full
REGRESSION. The regression analysis model might include estimated regres-
would begin with no variables. Then, sion parameters between a set of INDE-
the variable with the largest F-to-enter PENDENT VARIABLES, MEDIATORS, and
that met the initial criterion (e.g., 4.0) DEPENDENT VARIABLES. Where there are
would be added as a predictor. The re- as many parameters estimated in a full
searchers would then assess whether the model as there are DEGREES OF FREEDOM.
 funnel plot

it is said to be a SATURATED MODEL. Also ical equation that describes how values
called unrestricted model. of one variable are dependent on an-
other variable. The formula for a straight
f u l l - n o r m a l plot see NORMAL PROBA-
line, y = mx + b, describes a functional re-
BILITY PLOT.
lationship: The value y can be deter-
f u l l r a n k the quality of a MATRIX in mined by knowing the SLOPE, m, of
which each of the rows (or columns) are a line multiplied by a value, x, plus
independent of each other, such that the point where the line crosses the Y-
they do not provide redundant informa- AXIS, b.
tion. For example, a CORRELATION MA- f u n c t i o n a l variable a variable with
TRIX of three variables would be of fuU quantitative values that form a curve
rank if each of the three variables was when plotted on a graph.
unique, such that no variable com-
pletely overlapped with another vari- f u n c t i o n coefficient a quantitative
able. If, however, the three variables in value that multiplies a variable and that
the matrix were verbal intelligence, non- can change depending on other vari-
verbal intelligence, and full scale intelli- ables or COVARIATES. A function coeffi-
gence (i.e., some combination of verbal cient differs from other coefficients (e.g.,
and nonverbal intelligence), the matrix a REGRESSION COEFFICIENT) in that
would not be of full rank. it can vary whereas the others are con-
stant over all entities or participants. A
f u l l y recursive model a statistical function coefficient is often used in sta-
representation in which one variable tistical methods such as REGRESSION
predicts a second variable and the sec- ANALYSIS and TIME-SERIES ANALYSIS,
ond variable predicts the first variable. particularly when the data change over
In other words, all the variables in a fully time or space.
recursive model are reciprocally linked.
f u n d a m e n t a l science see BASIC SCI-
f u n c t i o n n. (symbol: f) a mathematical ENCE.
procedure that relates or transforms one
number, quantity, or entity to another f u n n e l plot a graph of EFFECT SIZES
according to a defined rule. For example, along the horizontal X-AXIS and some
if y = 2;t -I-1, y is said to be a function of x. measure of sample size along the vertical
This is often written y = f(x). Here y is the Y-AXIS. It is used to check for the possi-
dependent variable and x is the inde- bility of bias in META-ANALYSIS studies.
pendent variable. A funnel plot that has a pyramidal, up-
side-down cone shapeas in the follow-
f u n c t i o n a l data analysis (FDA) an ing exampleis thought to indicate low
area of statistics in which mathemati- levels of bias.
cal FUNCTIONS are used to study how The funnel shape reflects the fact that
things change across time. It usually in- most of the studies found approximately
volves the use of derivatives to form
curves that model the pattern of change
in some phenomenon (e.g., health,
achievement).
f u n c t i o n a l p r i n c i p a l components
analysis the reducfion of data to a
smaller set of dimensions or components,
where the data are FUNCTIONS or curves
rather than simple variables. See PRINCI-
PAL COMPONENTS ANALYSIS. -0.4 -0.2 0 0.2 0.4 0.6 0.8

f u n c t i o n a l relationship a mathemat- Effect Size

147
funnel sequence

medium-sized effects and used some- and gradually narrows the focus to more
what large sample sizes, with a few small specific items.
or large effects found with smaller sam-
FU plot abbreviation for FOLLOW-UP
ple sizes. Consider, however, a graph of
PLOT.
effect sizes and sample sizes that depicts
a growing, slanted mountain shape in- f u z z y set theory an approach to SET
creasing from lower left to upper right. THEORY that allows gradations of mem-
This positive slanted pattern would indi- bership in a set, instead of only assign-
cate a direct and positive relationship ing a 0 or a 1 that would indicate
(i.e., bias) between effect size and sample nonmembership or absolute member-
sizes. Conversely, if the graph revealed a ship, respectively. Some phenomena are
left-slanted mountain shape, from upper more readily modeled with fuzzy set the-
left to lower right, this would indicate a ory than conventional binary set theory,
negative relationship (and bias) between particularly when there is incomplete
effect size and sample size. This ne- information or varying degrees of the
gative, inverse pattern would reveal entity being studied. For example, in
that smaller effects could only be no- designation of a mental illness fuzzy set
ticed with large sample sizes and larger theory may be employed to allow
effects could be seen with smaller sam- for cases in which individuals could be
ple sizes. classified somewhere between an ab-
sence or a full diagnosis of a particular
f u n n e l sequence a method of stmctur- condition.
ing the order of questions in surveys and
interviews that starts with general items F value see F RATIO

148
 Gg
G symbol for LIKELIHOOD-RATIO CHI- then the first player gets the entire prize,
SQUARE. (b) If both players decide to share the
prize, then the prize is split, (c) If both
test see LIKELIHOOD-RATIO TEST. players decide to keep the entire prize,
gain score see DIFFERENCE SCORE. then both leave empty-handed. The
players must therefore base their deci-
G a l b r a i t h plot see RADIAL PLOT. [Rex sions on what they think the other will
F. Galbraith, British statistician] choose to do.
G A M abbreviation for GENERALIZED AD- G A M M abbreviation for GENERALIZED
DITIVE MODEL. ADDITIVE MIXED MODEL.
gambler's f a l l a c y a failure to recog- g a m m a (symbol: y) n. any of several dif-
nize the independence of chance events, ferent statistical indices, most com-
leading to the mistaken belief that one monly GOODMAN-KRUSKAL'S GAMMA.
can predict the outcome of a chance
event on the basis of the outcomes of g a m m a d i s t r i b u t i o n a pattern of data
past chance events. For example, a per- that can be depicted on vertical and hor-
son might think that the more often a izontal axes where information is
tossed coin comes up heads, the more known about two parameters, scale and
likely it is to come up tails in subsequent shape, both of which are values equal to
tosses, although each coin toss is inde- zero or higher. For example, a standard
pendent of the others, and the true prob- gamma distribution is one in which the
ability of the outcome of any toss is still shape and location parameters of the
just .5. continuous probability function are 0
and 1, respectively. Gamma distribu-
game theory a branch of mathematics tions are common in Bayesian statistics
concerned with the analysis of the be- and often used to describe waiting times
havior of decision makers (called players) and reaction times.
whose choices affect one another. Game
garbage i n , garbage out (GIGO) an
theory is often used in both theoretical
expression indicating that if the data
modeling and empirical studies of con-
used in analyses are not reliable or co-
flict, cooperation, and competition, and
herent, the results will not prove useful.
it has helped to stmcture interactive de-
The phrase was first used in computer
cision-making situations in numerous
science, cautioning users against tmst-
disciplines, including economics, politi-
ing computer output derived from unre-
cal science, social psychology, and eth-
liable input. Although the term is not
ics. A simple game theory example could
used much currently, it remains a rele-
be a scenario from a reality television
vant reminder to researchers to pay at-
show in which the final two players are
tention to how studies are designed and
each asked privately to make a decision
how data are collected, and not to de-
to share or keep the entire game prize.
pend on analyses to produce meaningful
The players are told that the outcome
results where the input data are not
will be decided by the following rules:
themselves meaningful.
(a) If one player decides to keep the en-
tire prize and the other decides to share, G A R C H abbreviation for GENERALIZED

149
Gaussian distribution

AUTOREGRESSIVE CONDITIONAL HETERO- needed to understand performance,

SCEDASTICITY. such as cultural norms and expectations.
Gaussian d i s t r i b u t i o n see NORMAL general effect an overall scientific
DISTRIBUTION. [Kari Friedrich Gauss finding about a relationship between
(1777-1855), German mathematician] phenomena. For example, a researcher
Gaussian k e r n e l a formula or function examining the relationship between
used in SMOOTHING out the rectangular various teaching styles and academic
bar pattern of data in a HISTOGRAM. It is performance may conclude the exis-
a particular type of KERNEL DENSITY ESTI- tence of a general effect in which better
MATOR often applied to estimate the student performance is obtained with an
shape of the underlying PROBABILITY appUed, hands-on style. This conclusion
DENSITY FUNCTION, whlch Is believed to might be made on the basis of a single
have a more continuous pattern than EXPERIMENTAL DESIGN in which stu-
that found in histograms and other dis- dents were randomly assigned to teach-
crete plots. [Karl Friedrich Gauss] ing conditions, or more likely it may be
based on multiple studies with varied
Gauss-Markov theorem a funda- conditions (e.g., male or female teacher,
mental theorem of mathematical statis- elementary or secondary education) and
tics that deals with the generation of course content (e.g., math, science, lan-
linear unbiased ESTIMATORS with mini- guage). See also INTERACTION EFFECT;
mum variance in the GENERAL LINEAR MAIN EFFECT.
MODEL. [Karl Friedrich Gauss; Andrei
Markov (1856-1922), Russian mathema- generalizability n. the extent to
tician] which results or findings obtained from
a sample are applicable to a broader pop-
GCR abbreviation for GREATEST CHARAC- ulation. For example, a theoretical
TERISTIC ROOT. model of change would be said to have
GEE abbreviation for GENERALIZED ESTI- high generalizability if it applied to nu-
MATING EQUATION. merous behaviors (e.g., smoking, diet,
substance use, exercise) and varying
Gehan's generalized W i l c o x o n populations (e.g., young children, teen-
test see BRESLOW TEST. [Edmund A. agers, middle-age and older adults). A
Gehan (1929- ), U.S. biostatistician; finding that has greater generalizability
Frank Wilcoxon (1892-1965), Irish-born also is said to have greater EXTERNAL VA-
U.S. statistician] LIDITY, in that conclusions pertain to
gender research the study of issues re- situations beyond the original study.
lated to femininity, masculinity, sexual- generalizability coefficient a nu-
ity, and gay, lesbian, and transsexual merical value that indicates how likely a
concerns. Gender research may involve finding is to occur across different con-
QUALITATIVE Studies that use focus ditions, people, and situations. A gen-
groups or interviews to understand be- eralizability coefficient is similar to the
havior as well as QUANTITATIVE analyses RELIABILITY COEFFICIENT used in CLASSI-
that examine potential gender group CAL TEST THEORY and can take on values
differences or prediction models. For ex- ranging from zero to one. See GENERAL-
ample, traditional research has assessed IZABILITY THEORY.
whether there are significant group dif-
ferences between men and women on generalizability model in GENERALIZ-
math and science performance. In more ABILITY THEORY, a statistical model used
recent times, more complex prediction to predict and test the broader applicabil-
models are sometimes examined to ity of a specific study's results for a par-
demonstrate that multiple factors are ticular phenomenon. For example, a

150
 generalized inverse

generalizability model could be hypoth- generalized autoregressive condi-

esized and assessed to determine whether tional heteroscedasticity (GARCH)
certain findings are relevant not only to a condition sometimes present in TIME-
the young adults initially evaluated but SERIES data and particularly in AUTO-
also to older adults, males and females, REGRESSIVE MOVING-AVERAGE MODELS,
and individuals from different ethnic in which one observation depends on
backgrounds. a previous observation (see AUTO-
CORRELATION). In other words, the data
generalizability theory a framework show differing VARIANCE of one variable
of principles and assumptions about at different levels of another variable
how to determine the RELIABILITY of a (i.e., heteroscedasticity). GARCH models
set of data. Researchers investigate the are important in econometrics and fi-
various FACETS of a study (items, raters, nancial modeling.
settings, etc.) to understand specific
sources of error and to determine the generalized distance see MAHALA-
conditions under which observations NOBIS DISTANCE.
will be consistent and applicable across generalized estimating equation
different contexts (e.g., age groups, geo- (GEE) any of a set of statistical proce-
graphic regions, socioeconomic status). dures used to analyze correlated data
that do not require all of the ASSUMP-
generalized additive mixed model TIONS of the GENERAL LINEAR MODEL,
(GAMM) a type of statistical procedure such that normality and independence
used to analyze data with more than one are not needed.
subgroup (e.g., different classrooms of
students) when the data within sub- generalized eta squared (symbol:
groups are correlated; The generalized T|G^) in analysis of variance, a measure of
additive mixed model is an extension of effect that takes into account specific
the GENERAL LINEAR MODEL that takes features in the design that could alter
such dependency within subgroups into the size of the effect. Generalized eta
account. It is also a specialized form of a squared ranges from 0 to 1 in value, with
MIXED-EFFECTS MODEL, as it incorpo- larger values indicating that a specific
rates both FIXED EFFECTS (e.g.. Specific factor explains more of the VARIANCE in
choice of teaching strategy) and RAN- the DEPENDENT VARIABLE. See also ETA
DOM EFFECTS (e.g., a set of student study SQUARED.
times randomly chosen from a set of all generalized g a m m a d i s t r i b u t i o n a
possible times). PROBABILITY DISTRIBUTION that extends
generalized additive model (GAM) the GAMMA DISTRIBUTION to include a
a statistical procedure for analyzing third parameter of rate or location as
data that may not strictly conform to well as the two parameters of scale and
a LINEAR MODEL. A generalized additive shape. It often is used in SURVIVAL ANAL-
YSIS to model the time until a specific
model still assumes that there is an AD-
DITIVE EFFECT of the INDEPENDENT VARI-
event, such as death or onset of an ill-
ABLES on the DEPENDENT VARIABLE, ness. Various special cases of the general-
ized gamma distribution include the
such that no effect depends on any other
and the individual effects add up to the EXPONENTIAL DISTRIBUTION and LOG-
total explained effect. It is an extension NORMAL DISTRIBUTION.
of the GENERAL LINEAR MODEL that al- generalized inverse (g inverse) a
lows a researcher to posit and assess vari- MATRIX that serves as a divisor to an-
ous nonlinear patterns in a data set and other matrix when there is not a com-
discern which pattern most closely ap- mon denominator matrix or INVERSE
proximates the data. See ADDITIVE MODEL. MATRIX readily available to provide a di-

151
generalized least squares regression

rect solution. A generalized inverse also generalized m a x i m u m - l i k e l i h o o d

allows the division of matrices that are r a t i o test a N O N P A R A M E T R I C statistical
not square (i.e., do not have the same analysis used with data that follow a
number of rows as columns). For exam- CHI-SQUARE DISTRIBUTION but dO nOt
ple, i n a MULTIVARIATE ANALYSIS OF meet conventional ASSUMPTIONS such
VARIANCE, one obtains a ratio by divid- as NORMALITY and H O M O G E N E I T Y OF
ing the matrix of between-groups vari- VARIANCE. It can be used in more cir-
ance by the matrix of within-groups cumstances than the traditional M A X I -
variance. If it is difficult to find the M U M LIKELIHOOD ratlo test.
within-groups inverse matrix that would
allow this calculation, a generalized in- generalized mixed model (GMM)
verse could be found instead. Also called see GENERALIZED LINEAR MIXED MODEL.
pseudoinverse. generalized multilevel model a
more specific extension of the GENERAL-
g e n e r a l i z e d l e a s t s q u a r e s regres-
s i o n a procedure used to estimate RE- IZED LINEAR MODEL that incorporates pro-
GRESSION COEFFICIENTS and other PA- cedures for addressing multiple groups of
RAMETERS when ERROR V A R I A N C E S are data, such as classrooms of students.
correlated or show HETEROGENEITY O F g e n e r a l i z e d o d d s r a t i o an extension
VARIANCE, such that a conventional or- of the proportional hazards model (see
dinary LEAST SQUARES REGRESSION pro- c o x REGRESSION ANALYSIS) that is used
cedure will produce biased results. to assess relationships between ORDINAL
generalized linear mixed model VARIABLES having more than two cate-
(GLMM) a model that takes into ac- gories.
count bpth specifically selected FIXED generalized Poisson distribution
EFFECTS and a broader range of R A N D O M an extension of the traditional POISSON
EFFECTS in the data. A generalized linear DISTRIBUTION in which the variance
mixed model is an extension of the more and mean are not required to be equal.
common G E N E R A L I Z E D LINEAR M O D E L
For example, the number of deaths from
(which only involves fixed effects) and automobile crashes may be a constant
provides greater E X T E R N A L VALIDITY
low rate in a rural area with a temperate
when generalizing beyond a specific climate. In an urban area having differ-
study. Also called g e n e r a l i z e d m i x e d ent weather and traffic patterns through-
m o d e l (GMM). See also MIXED-EFFECTS out the year, however, the number of
MODEL. automobile-related deaths may vary
g e n e r a l i z e d l i n e a r m o d e l (GLM) a widely such that a generalized Poisson
broad class of statistical procedures that distribution may provide a more accu-
allow variables to be related in a predic- rate representation of the process over
tion or REGRESSION ANALYSIS by taking time.
into account the variance of each. The
g e n e r a l i z e d v a r i a n c e see DETERMI-
generalized linear model extends ordi- NANT.
nary LEAST SQUARES REGRESSION to In-
clude other proceduressuch as LOGISTIC g e n e r a l l i n e a r m o d e l (GLM) a large
REGRESSION and POISSON REGRESSION class of statistical techniques, including
that allow for the use of categorical de- REGRESSION ANALYSIS, ANALYSIS O F
pendent variables having very lopsided VARIANCE, and C O R R E L A T I O N ANALYSIS,
distributions, in which most people en- that describe the relationship between a
dorse a particular category (e.g., no her- D E P E N D E N T V A R I A B L E and one or more
oin use) or fail to endorse a particular explanatory or I N D E P E N D E N T VARIABLES.
category (e.g., heroin use). See also G E N - It is a more specific type of GENERALIZED
ERALIZED MULTILEVEL MODEL. LINEAR MODEL. Most Statistical tech-

152
 Gini coefficient

niques employed in the behavioral sci- MEAN except that the latter involves the
ences can be subsumed under the sum of the quantities divided by the
general linear model. number of quantities, whereas the for-
mer is the product of the quantities with
general n o r m a standard that is widely
the product raised to the power of 1 di-
accepted, such as a regularly occurring vided by the number of quantities. Thus,
level of performance on an intelligence for the numbers 1, 2, and 3 the arithme-
test or an expected pattern of alcohol tic mean is 6/3 = 2, and the geometric
use on college campuses. mean is the cube root of (1)(2)(3) = 6^'^
general systems theory an interdisci- 1.817. See also HARMONIC MEAN.
plinary conceptualframeworkfocusing
geometric progression a series of
on wholeness, pattern, relationship, hi-
numbers in which each successive num-
erarchical order, integration, and orga-
ber is the product of the previous num-
nization. It was designed to move
ber times a constant. For example, the
beyond the reductionistic and mecha-
geometric progression 5, 10, 20, and 40
nistic tradition in science (see REDUC-
would have 80 as the next number and
TIONISM) and integrate the fragmented
the constant, called the common ratio,
approaches and different classes of phe-
is 2 (i.e., 2 X 5 = 10; 2 X 10 = 20; 2 X 20 =
nomena studied by contemporary sci-
40; 2 X 40 = 80; etc.). Also called geo-
ence into an organized whole. In this
metric sequence; geometric series.
framework, an entity or phenomenon
should be viewed holistically as part of a geostatistics n. a set of methods for an-
set of elements interacting with one an- alyzing data related to the physical envi-
other; the goal of general systems theory ronment. In psychology, for example,
is to identify and understand the princi- geostatistics could be used to under-
ples applicable to all entities in the set. stand which factors in the environment
The impact of each element depends on can bring about or exacerbate behavioral
the role played by other elements in- conditions, such as substance abuse, or
volved and order arises from interaction spread diseases, such as AIDS.
among these elements. Also called sys-
tems theory. Gibbs s a m p l i n g a statistical method of
randomly generating sets of data that in-
generating f u n c t i o n a mathematical corporate the patterns of values or PROB-
method of finding the number of occur- ABILITY DISTRIBUTIONS from two or
rences in a large sequence. It is formed more variables. Also called Gibbs sam-
by a power series where the k^^ variable is pler, [losiah W. Gibbs (1839-1903),
raised to the k^^ power. For example, a U.S. mathematician and physicist]
generating function may take the form
GIGO abbreviation for GARBAGE IN, GAR-
UQX^ + AjX' -I- a2X^ + a^X^ where a is
BAGE OUT.
a weight and the generating function is a
weighted sum. See also MOMENT GENER- G i n i coefficient a measure of variance
ATING FUNCTION; PROBABILITY GENER- in a FREQUENCY DISTRIBUTION, provid-
ATING FUNCTION. ing a numerical summary of a LORENZ
CURVE. It ranges in value from 0, indi-
geometric d i s t r i b u t i o n the PROBA- cating no variance such that all entities
BILITY DISTRIBUTION of the number of
are equal, to 1, indicating maximum
failed trials before the first success in a variance such that every entity is differ-
series of BERNOULLI TRIALS. ent. Also called coefficient of con-
geometric mean a measure of CEN- centration; G i n i concentration;
TRAL TENDENCY Calculated for k num- Gini index; G i n i ratio; Gini statis-
bers Xi...Xi. as (x-i xx2'x X2...xi^f'^. A geo- tic. [Corrado Gini (1884-1965), Italian
metric mean is similar to an arithmetic statistician and sociologist]

153
g inverse

g inverse abbreviation for GENERALIZED and ultimately very low sales as the
INVERSE. fad dies out. Gompertz curves often are
Glass's d an EFFECT SIZE measure that used in SURVIVAL ANALYSIS. Also called
represents the standardized difference Gompertz distribution; Gompertz
between means (i.e., the difference in function. [Benjamin Gompertz (1779-
average values for two samples divided 1865), British mathematician]
by the STANDARD DEVIATION of the sec- Goodman-Kruskal's g a m m a a mea-
ond sample). It is often used in META- sure of association between two CATE-
ANALYSIS and other research in which it GORICAL VARIABLES, where 0 represents
is important to determine whether an ef- no relationship and 1 represents perfect
fect persists across studies in order to association. For example, a marketing
consolidate a result. Also called Glass's researcher may want to assess the rela-
delta. See also COHEN'S D; HEDGES'S G. tionship between the location of a store
[Gene V. Glass (1940- ), U.S. statisti- and the type of clothes purchased. A
cian] Goodman-Kruskal's gamma of .30 or
G L M 1. abbreviation for GENERAL LINEAR higher would indicate a medium-sized
MODEL. 2. abbreviation for GENERAL- correlation such that clothes distribu-
IZED LINEAR MODEL. tors desiring to maximize sales should
provide specific kinds of clothes at spe-
G L M M abbreviation for GENERALIZED cific locations. Also called Goodman-
LINEAR MIXED MODEL. Kruskal's tau. [Leo A. Goodman
global m a x i m u m the largest possible (1928- ) and William Henry Kruskal
value in a set of numbers. For example, (1919-2005), U.S. statisticians]
the global maximum for a CORRELATION
goodness of f i t the degree to which
COEFFICIENT is 1, whereas the global values predicted by a model agree with
maximum for a STANDARD DEVIATION empirically observed values. For exam-
or VARIANCE is not finite (i.e., infinity). ple, a researcher may wish to assess
global m i n i m u m the smallest possible whether a pattern of frequencies from
value in a set of numbers. For example, a study is the same as theoretically ex-
the global minimum for a CORRELATION pected, whether two CATEGORICAL
COEFFICIENT is -1 and for a STANDARD VARIABLES are independent, or whether
DEVIATION or VARIANCE it is 0. a REGRESSION EQUATION correctly pre-
dicts obtained data. A small, non-
GLS abbreviation for generalized least significant value from a GOODNESS-OF-
squares. See GENERALIZED LEAST SQUARES
FiT TEST indicates a well-fitting model.
REGRESSION.
G M M abbreviation for generalized mixed goodness-of-fit test any of various sta-
model. See GENERALIZED LINEAR MIXED tistical procedures that assess how
MODEL. closely data obtained from a sample
match an expected population pattern
Gompertz curve a function or curve or model. The CHI-SQUARE GOODNESS-
used to describe a pattern of growth that OF-FIT TEST is a commonly used exam-
is slow at the beginning and end but ple.
rather rapid or steep in the middle, thus
forming a shape resembling a leaning good-subject role the behavior dis-
letter S. For example, the adoption of played by a participant in a research
a fashion fad may follow a Gompertz study who is eager to perform as he
curve: Few people purchase the product or she thinks is desired or wanted.
initially but there is a subsequent period The adoption of such a role could
of highly increased sales that precedes bias study results as the true effect of
a lessening of interest in the product the INDEPENDENT VARIABLE cannot

154
 graphic rating scale

be adequately determined. See DEMAND formed from the inner products of a

CHARACTERISTICS. Compare APPREHEN- set of VECTORS. Gramian matrices are
SIVE-SUBJECT ROLE; FAITHFUL-SUBJECT important in STRUCTURAL EQUATION
ROLE; NEGATIVISTIC-SUBJECT ROLE. MODELING or when working with CO-
VARIANCE structures. Also called Gram
Gower s i m i l a r i t y coefficient a mea- matrix. [Jorgen Pedersen Gram (1850-
sure of proximity used in CLUSTER ANAL- 1916), Danish mathematician]
YSIS to indicate the similarity of different
entities or participants. It may assume g r a n d mean a numerical average
values from 0 to 1, with the former indi- (MEAN) of a group of averages. For exam-
cating items that are very distinct and ple, if the average test score for one class-
the latter indicating items that are very room is 75 and the average score for
similar or alike. [John C. Gower] another classroom is 73, the grand mean
grade-of-membership model a sta- of the two classrooms is (75 -i- 73)/2 = 74.
tistical representation or procedure for Similariy, if an ANALYSIS OF VARIANCE
describing the underlying stmcture of reveals average values of 3, 10, and 20
continuous data. It is an extension of LA- on a response or DEPENDENT VARIABLE
TENT CLASS ANALYSIS, which is used with
for three groups of study participants,
discrete CATEGORICAL DATA, to CONTIN-
the grand mean of scoresthat is, the
UOUS VARIABLES by allowing individuals
average of all participants' responses re-
gardless of the condition of the INDEPEN-
or entities to simultaneously be partial
DENT VARIABLESis (3 -I-10 -F 20)/3 = 11.
members of several classes or categories.
Graeco-Latin square (Greco-Latin Granger causality the condition that
square) an experimental design that su- arises when a TIME SERIES, X, is found to
perimposes one LATIN SQUARE upon an- predict values of another time series, y.
other. In this type of FRACTIONAL Granger causality involves a process of
FACTORIAL DESIGN, two scts of elements HYPOTHESIS TESTING to assess whether a
are arranged in the same set of cells value of X at one time point is related to
in such a way that every row and every one or more future values of y. [Clive
column contains each element of both Granger (1934-2009), British-born U.S.
sets once and once only, and each cell economist]
contains a different ordered pair. The g r a p h n. a visual representation of the
following is a hypothetical example for re- relationship between numbers or quan-
search comparing three methods of praise tities, which are plotted on a drawing
(1,2, and 3) administered for three types with reference to axes at right angles (see
of task (A, B, and C) under three condi- X-AXIS; Y-AXIS) and linked by lines, dots,
tions of supervision (a, p, and y). or the like, BAR GRAPHS, HISTOGRAMS,
and FREQUENCY POLYGONS are com-
Treatment monly used examples.
1 2 3
graphic model a graph depicting the
Participant 1 Aa BP Cy relations of CONDITIONAL INDEPENDENCE
Participant 2 By Ca Ap between three or more RANDOM VARI-
ABLES. Graphic models are often used in
Participant 3 Cp Ay Ba
BAYESIAN NETWORKS and MARKOV
CHAINS, in which they take the form of
Graeco-Latin squares are used in re- DIRECTED ACYCLIC GRAPHS and UNDI-
search to minimize or eliminate the in- RECTED GRAPHS. Also called graphical
fluence of extraneous variables and to
model.
balance ORDER EFFECTS.
graphic r a t i n g scale a series of an-
G r a m i a n m a t r i x a SQUARE MATRIX chored points (usually from low to high)

155
gray literature

Employee overall performance

during past year

1 2 3 4 5
poor below average above excellent
average average

graphic rating scale

on a continuum, often used to evaluate to examine the greatest characterisfic root

the performance or behavior of individ- of the between-groups matrix and com-
uals in a work or learning environment. pare it to the greatest characteristic root
For example, a graphic rating scale for of the within-groups matrix. Also called
assessing an employee's overall perfor- largest characteristic root; Roy's
mance during the past year might take greatest characteristic root; Roy's
the form of a line with the response op- greatest root; Roy's largest root.
tions of 1 (poor), 2 (below average), 3 (av-
greatest-characteristic-root test
erage), 4 (above average), and 5 (excellent)
one of several criteria used to assess the
listed, as shown above. The supervisor
significance of between-groups and
would mark the line to indicate the em-
within-groups differences in a MULTI-
ployee's performance level.
VARIATE ANALYSIS OF VARIANCE. It in-
gray literature research findings that volves finding the largest EIGENVALUE of
are not readily available because they the between-subjects matrix over the
have not been published in archival within-subjects matrix. Also called larg-
sources. Examples include dissertations, est-root test; Roy's greatest-root
papers presented at meetings, papers ei- test; Roy's greatest-characteristic-
ther not submitted or rejected for publi- root test; Roy's largest-root test;
cation, and technical reports. Gray Roy's root test. See GREATEST CHARAC-
literature contributes to the FILE- TERISTIC ROOT.
DRAWER PROBLEM in META-ANALYSIS,
Greco-Latin square see GRAECO-
the situation in which analytic results
LATIN SQUARE.
are biased due to researchers not having
access to nonsignificant findings that g r i d s a m p l i n g the process of dividing
are not easily found. Also called fugi- an area into equal sections, randomly
tive literature. checking each section, and making im-
provements and changes to some pro-
greatest characteristic root (GCR)
cess based on the findings. Grid
the largest EIGENVALUE of a matrix that
sampling is often used in agricultural re-
describes the variance of the largest LIN-
search and was conducted by Sir Ronald
EAR COMBINATION of variables in that
Fisher when he devised the often-used
data array, summarizing the association
ANALYSIS OF VARIANCE and its accompa-
between a set of independent variables
and a set of dependent variables. GCRs nying F TEST.
are used in analyses involving several grounded theory a set of procedures
variables, such as CANONICAL CORRELA- for the systematic analysis of unstruc-
TION ANALYSIS, MULTIVARIATE ANALYSIS tured QUALITATIVE DATA SO as to derive
OF VARIANCE, Or PRINCIPAL COMPO- by INDUCTION a supposition that ex-
NENTS ANALYSIS. In the multivariate plains the observed phenomena. A re-
analysis of variance, for example, one of searcher typically observes the entity of
the methods for assessing the results is interest in a natural setting before draw-

156
 grouped frequency distribution

ing conclusions about its nature and ex- ual level. For example, educational re-
istence. searchers may be interested in providing
a group dimension by depicting results
group n. a collection of participants in a
at the classroom level rather than by in-
research study whose responses are to be
dividual students.
compared to the responses of one or
more other collections of research par- group f l i s t r i b u t i o n the scores for a
ticipants. Participants in a particular group of individuals. For example, re-
group all experience the same experi- searchers may be interested in providing
mental conditions or receive the same the group distribution performance
treatment, which differs from the exper- scores for separate classrooms or
imental conditions or treatments partic- schools.
ipants in other groups experience or
receive. group-divisible design a research de-
sign in which separate groups each con-
group-average clustering see AVER- tain a unique set of BLOCKS whose
AGE-LINKAGE CLUSTERING. members are assigned a unique pair of
group class 1. in statistical methods treatments. The group-divisible design
such as ANALYSIS OF VARIANCE, the des- provides a balanced structure for assign-
ignation of the response options for a ing multiple treatments across blocks
CATEGORICAL VARIABLE as the levels of and across two or more groups.
an INDEPENDENT VARIABLE. For exam- grouped data information that is
ple, if a researcher is studying religious grouped into one or more sets in order to
affiliation, he or she may include gender analyze, describe, or compare outcomes
as a variable in the analysis and use its at a combined level rather than at an in-
two distinct options (i.e., female and dividual level. For example, data from a
male) as levels for comparison. 2. the re- FREQUENCY DISTRIBUTION may be ar-
sults derived from such groups. ranged into CLASS INTERVALS. See also
group-comparison design a type of GROUP-COMPARISON DESIGN; GROUP-
research approach that investigates po- DIVISIBLE DESIGN.
tential differences across sets of individ-
grouped frequency d i s t r i b u t i o n a
uals who are often randomly assigned to
a CONTROL CONDITION or to One or description of how often a set of specific
more specific EXPERIMENTAL CONDI- responses, organized into equal-sized
TIONS. Data from a group-comparison
subsets of possible responses, occur in a
design are often analyzed with such sta- sample. It is a type of FREQUENCY DISTRI-
tistical methods as a T TEST or ANALYSIS BUTION that is particularly useful when
OF VARIANCE. Also called group- there are a large number of response
difference design. choices (e.g., 10-20 or more) and re-
searchers wish to present the informa-
group design any of a variety of proce- tion more concisely. For example,
dures intended to compare two or more suppose that a researcher wants to sum-
collections of individuals or entities. marize the individuals in a sample in
which age ranges from 18 to 85 years. In-
group difference any observed varia- stead of listing all 68 different ages, the
tion between groups of participants in researcher could combine the data into
an experiment when considering each 5-year intervals, which would produce
group as a single entity. See GROUP- 14 subsets of ages. More specifically, the
COMPARISON DESIGN. researcher would create two columns,
group dimension the perspective listing in the left-hand one the 14 age
gained by viewing results at a larger, subsets (e.g., 16-20, 21-25, 26-30, ...
combined level instead of at an individ- 81-85). In the right column, the re-

is?
group effect

searcher would list the frequencies of being used to evaluate the data. It is sim-
each age subset; that is, how often each ilar to a RANDOMIZED CLINICAL TRIAL
age subset was endorsed by the individu- but distinct in assigning intact clusters
als in the sample, thus reducing a large of participants to the specific conditions
number of scores into smaller, more rather than individual participants. Also
manageable groups. called cluster-randomized trial.
group effect a research finding specific group sequential design a research
to the group of individuals to which a design in which a treatment is randomly
participant belongs. A group effect could assigned to different groups at varying
appear in an assigned subset, such as a times, such that every group receives the
treatment or intervention, or in a natu- treatment eventually. A group sequen-
rally occurring subset, such as age level tial design is useful when it is important
or classroom. For example, a researcher for each participant to have an opportu-
might be interested in a group effect of a nity to benefit from the treatment but a
specific reading intervention, or in a comparison group nonetheless is re-
group effect of book reading for students quired: The portion of the sample that
in the current decade who may be read- receives the treatment later acts as a
ing less than previous sets of students CONTROL GROUP initially.
owing to more common use of the com-
puter and television. g r o w t h curve a graphic representation
of progress over time on a specific vari-
group experiment an experiment in able within a specific group (e.g., the de-
which subsets of individuals serve as the velopment of an organism, as in the
unit of analysis. For example, a re- hypothetical example opposite).
searcher could compare the perfor- Data in a growth chart often are evalu-
mance of different classrooms rather ated with TIME-SERIES ANALYSIS, LATENT
than of specific individuals. GROWTH CURVE ANALYSIS, or Other lon-
group i n t e r v i e w an interview in gitudinal methods.
which one or more questioners elicit in- g r o w t h curve analysis a statistical
formation from two or more respon- procedure for assessing the level and rate
dents in an experimental or real-life (SLOPE) of change over time on a specific
situation. The participation and interac- variable in a sample of individuals. See
tion of a number of people, particularly also LATENT GROWTH CURVE ANALYSIS.
if they are acquainted with each other as
members of a club or similar group, is g r o w t h f u n c t i o n the relationship be-
believed to yield more informative re- tween a DEPENDENT VARIABLE and Sev-
sponses than are typically obtained by eral levels of an INDEPENDENT VARIABLE
interviewing individuals separately. defined in units of time (e.g., days,
weeks, months, years).
group m a t c h i n g see MATCHING.
G statistic 1. a value used to assess the
group-randomized t r i a l a research presence of spatial clustering or
design in which groups of individuals AUTOCORRELATION in a data set. A large
are randomly assigned to TREATMENT
G statistic with a small accompanying p
conditions or CONTROL CONDITIONS. In
VALUE (e.g., < .05) suggests there are sub-
educational or organizational settings,
sets of dependence or clustering. 2. see
for example, a group-randomized trial
G^
could be used to study classrooms or de-
partments, with HIERARCHICAL LINEAR G study generalizability study: research
MODELING and similar analyses that about the widespread applicability of a
take the larger group membership (e.g., particular measure or test. A G study as-
classroom or department) into account sesses several facets or factors (e.g., in-

158

growth
strument type, administration occasion)
to determine under which conditions
the measurement provides reliable in-
formation. See also D STUDY.
G test see LIKELIHOOD-RATIO TEST.
G u m b e l d i s t r i b u t i o n a statistical pat-
tern often used in SURVIVAL ANALYSIS to
describe data having either very large or
very small values. It is a type of EXTREME-
VALUE DISTRIBUTION. [Emil Julius
Gumbel (1891-1966), German mathe-
matician]
G u t t m a n scale a type of attitude scale
that consists of multiple verbal state-
ments ordered to reflect increasing lev-
els of positive evaluation. Endorsement
of a particular statement impUes en-
159
Guttman scale
curve
dorsement of all statements less extreme
than that statement. For example. Item
1 could state "I believe that education is
valuable," Item 2 could state "I believe
that people who are educated are more
productive," and Item 3 could state "I
believe that I would be more productive
if I had more education." A person who
agreed with the third statement would
also agree with the first and second
statements. Although generally used to
measure attitudes, Guttman scales can
also be used to assess other properties of
a target of judgment. Also called cumu-
lative scale; scalogram. [first de-
scribed in 1944 by Louis Guttman
(1916-1987), U.S. experimental psy-
chologist]
 Hh
h symbol for COMMONALITY COEFFI- portant, whereas those that are further
CIENT. removed from zero (e.g., two or more
standard deviations) are considered sig-
HQ symbol for NULL HYPOTHESIS. nificant. Alternatively, when plotting
RESIDUALS, points far from zero reveal
H j symbol for ALTERNATIVE HYPOTHESIS.
the existence of OUTLIERS that indicate a
Ha symbol for ALTERNATIVE HYPOTHESIS. poorly fitting regression model. The re-
lated full-normal plot (see NORMAL
h a b i t u a t i o n n. the weakening of a re- PROBABILITY PLOT) shows both the origi-
sponse to a stimulus, or the diminished nal positive and negative variable val-
effectiveness of a stimulus, following re- ues.
peated exposure to the stimulus. Com-
pare DISHABITUATION. halo effect a rating bias in which a gen-
h a l f - n o r m a l plot a graph in which the eral evaluation (usually positive) of a
ABSOLUTE VALUES Obtained for a vari- person, or an evaluation of a person on a
able are given along the horizontal x- specific dimension, influences judg-
axis and their corresponding probabili- ments of that person on other specific
ties of occurrence under a STANDARD- dimensions. For example, a person who
IZED DISTRIBUTION are given along the is generally liked might be judged as
vertical y-axis. For example, a half- more intelligent, competent, and honest
normal plot could be used to assess the than he or she actually is.
importance of observed experimental ef- h a n g i n g rootogram a type of HISTO-
fects: Points that fall close to the mean GRAM in which the horizontal X-AXIS is
of zero are considered minor or unim- slightly elevated and the vertical Y-AXIS

^ 140
>
U
>
C 120
3
O- 100
0)
t. eoi
I 6(
ra

V)

-H 1I 1h- -HI 1-
o o o o o o o
o o o o o o o
o o o o o o o

Debt (dollars)
hanging rootogram
160
 hazard analysis

depicts the square root of frequencies TENDENCY. It Is Computed for n scores by

using bars that "hang" downward from a dividing the scores by the sum of their
curve showing the expected NORMAL RECIPROCALS; that is, n divided by the
DISTRIBUTION. Consider the hypotheti- sum of 1/xi -I- I/X2 + l/jf. See also GEO-
cal example on the previous page. METRIC MEAN; MEAN.
A hanging rootogram allows research- Hartley test see F^,,^ TEST.
ers to see how well their OBSERVED DIS-
[Hermann Otto Hartley (1912-1980),
TRIBUTION (e.g., of debt for a group of
German-born U.S. statistician]
middle-class families) fits an expected
THEORETICAL DISTRIBUTION. If the Ob- Hartley test f o r homogeneity o f
served data differ from the expected pat- variance see F^,^^ TEST. [Hermann
tern, the hanging histogram bars will Otto Hartley]
drop below the elevated horizontal axis,
hat m a t r i x (abbreviation: H) in statisti-
thereby indicating places of poor match
cal procedures such as LINEAR REGRES-
between the observed and theoretical
SION, a MATRIX used to compare an
distributions. Also called suspended
OBSERVED DISTRIBUTION Of ValueS tO FIT-
rootogram. See ROOTOGRAM.
TED VALUES. When there are large dis-
h a p h a z a r d s a m p l i n g any method of crepancies between the observed and
selecting research participants that is expected (fitted) values, there is evi-
neither random nor systematic and dence for OUTLIERS, indicating that the
hence likely to be biased, CONVENIENCE statistical model may not be appropriate
SAMPLING is an example. Thus, if re- for the data.
searchers decided to solicit participants H a w t h o r n e effect the effect on the
from an Internet advertisement the pro- behavior of individuals of knowing that
cess would be biased toward individuals they are being observed or are taking
who own a computer and who are more part in research: Participants typically
likely to respond to corhputer-based improve their performance simply as a
communications. Depending on the result of knowing that an interest is
prevalence and likelihood of computer being shown in them, rather than as a
use in various facets of the population, result of any experimental manipula-
the sample could be biased by gender, tion. The Hawthorne effect is named
ethnicity, geographical location, eco- after the Western Electric Company's
nomic status, or age, among other possi- Hawthorne Works plant in Cicero, Illi-
ble variables. nois, where the phenomenon was first
h a r d data information that is QUANTI- observed during a series of studies on
TATIVE and specific, usually obtained worker productivity conducted from
from rigorous EXPERIMENTAL RESEARCH, 1924 to 1932. These Hawthorne Studies
systematic measurement, and STATISTI- began as an investigation of the effects
CAL ANALYSIS. Hard data are sometimes of particular job characteristicssuch as
contrasted with SOFT DATA, which may illumination conditions, monetary in-
be more QUALITATIVE and involve anec- centives, and rest breakson productiv-
dotal evidence that is not obtained sys- ity, but it evolved into a much wider
tematically. consideration of the role of worker atti-
tudes, supervisory style, and group dy-
h a r m o n i c analysis a form of THEO- namics when initial results showed that
RETICAL STATISTICS used in SIGNAL DE- performance increased regardless of the
TECTION THEORY and neuroscience, levels of the job characteristics em-
which translates signals into frequen- ployed by the company.
cies.
hazard analysis an analysis used to
h a r m o n i c mean a measure of CENTRAL identify the level of risk associated with,

161
hazard function

say, an engineering process or a new Hedges's g an EFFECT SIZE measure that

medical or food product. It involves represents the standardized difference
identifying potential hazards, determin- between means. Hedges's g differs from
ing their probability, and evaluating the the more widely used COHEN'S D in that,
likely severity of their effects. Also called whereas both use a pooled STANDARD
hazard modeling. DEVIATION in the denominator, the for-
mer uses n - l and the latter uses n when
h a z a r d f u n c t i o n a mathematical for- calculating the standard deviations.
mula that describes the relationship be- Thus, Hedges's g tends to be considered
tween the risk of a particular event as having less bias than Cohen's d. See
occurring and time. It is one element of also GLASS'S D. [Larry V. Hedges, U.S.
SURVIVAL ANALYSIS. statistician]
h a z a r d plot in HAZARD ANALYSIS, a Hellinger distance a value that de-
graph used to indicate the rate of failures scribes how dissimilar two PROBABILITY
over successes across specific points in DISTRIBUTIONS are. A Hellinger distance
time. For example, in a study of a partic- of 1 indicates that the two distributions
ular disease, a hazard plot would show are completely different, whereas a
the number of individuals who died Hellinger distance of 0 indicates the two
from the disease divided by the number distributions are probabilistically the
who survived at various time points dur- same. [Ernst David Hellinger (1883-
ing the course of the study. 1950), German-born U.S. mathemati-
cian]
h a z a r d rate in HAZARD ANALYSIS, a
value used to indicate the immediate po- heterogeneity n. the quality of having
tential or risk of an event's occurrence at very different characteristics or values.
a particular instance in time. It often is For example, HETEROGENEITY OF VARI-
used in health studies and will vary de- ANCE is present in an ANALYSIS OF VARI-
pending on the time point of interest. ANCE when the average squared distance
For example, a hazard rate may be used of each score from the mean differs for
to determine whether an individual who each group in the study (e.g., control
is abstinent from alcohol or drugs at six group vs. treatment group). Compare
months is likely to relapse in the near fu- HOMOGENEITY.
ture.
heterogeneity o f regression the sit-
h a z a r d ratio in HAZARD ANALYSIS, a uation in which there are different val-
value used to indicate the risk of an ues for the SLOPE between a COVARIATE
event's occurrence in a reference group and DEPENDENT VARIABLE across differ-
versus that in a comparison group. For ent groups in an ANALYSIS OF COVARI-
example, a researcher studying rehabili- ANCE. Heterogeneity of regression violates
tation among older adults who have one of the basic assumptions of such an
broken a bone may compute a hazard analysis, as it indicates that the differ-
ratio for completion of rehabilitation ences among groups are too substantial
treatment according to cognitive status, for a researcher to average REGRESSION
with high functioning patients consid- COEFFICIENTS in order to estimate the ef-
ered the reference group and low func- fect of a covariate on a dependent vari-
tioning patients the comparison group. able. Compare HOMOGENEITY OF
REGRESSION.
h a z a r d regression a statistical proce-
dure used in SURVIVAL ANALYSIS to ana- heterogeneity o f variance the situa-
lyze which variables are useful in tion in which the variance of a random
predicting the occurrence of an event, variable is different at each level or value
such as disease or death. of another variable. Wai:(y\x) is not the
 hierarchical agglomerative clustering

same for all values of x; that is, the vari- estimate having a value that is impossi-
ance in y is a function of the variable x. ble or very rare (e.g., a negative ERROR
Heterogeneity of variance violates one VARIANCE estimate). Heywood cases
of the basic assumptions of REGRESSION may indicate any of the following: a
ANALYSIS and other stafisfical prodedures. sample that is too small to adequately es-
Also called heteroscedasticity. Com- timate the parameters; data that do not
pare HOMOGENEITY OF VARIANCE. have a NORMAL DISTRIBUTION or that
contain OUTLIERS; a misspecified model
heteromethod b l o c k in a MULTI- that is not appropriate for the data; or a
TRAIT-MULTIMETHOD MATRIX, the block parameter whose true value is so close to
of values representing the correlations a boundary (e.g., 1 or 0) in the popula-
between traits (e.g., depression, self- tion that its estimate exceeded this limit
esteem) assessed with different proce- due to sampling fluctuation.
dures (e.g., paper-and-pencil survey,
clinical interview, physiological mea- h i d d e n M a r k o v model a statistical
sure). If the correlations are similar re- representation of a series of time-
gardless of the type of procedure used to ordered events in which (a) each event is
assess the traits, a researcher may con- related to the immediately prior event
clude he or she is assessing the true value but not to earlier events and (b) the pro-
of the traits rather than obtaining scores cess that is driving the series is not di-
that are merely ARTIFACTS of the mea- rectly observable (i.e., is a LATENT
surement procedure used. See HETERO- VARIABLE). For example, consider a
TRAIT-HETEROMETHOD COEFFICIENT. study of the number of cigarettes a day
heteroscedasticity n. see HETEROGENE- that are smoked by individuals in a
ITY OF VARIANCE. hcteroscedastic adj.smoking cessation program. If a hidden
Markov model were used to represent
heterotrait-heteromethod coeffi- the process, this would indicate the pres-
cient in a MULTITRAIT-MULTIMETHOD ence of unobservable states or events
MODEL, a number that indicates the (e.g., peer pressure to smoke; stress or en-
strength of the association between dif- vironmental cues to smoke, as in the so-
ferent characteristics or variables mea- cial context of a bar) that are linked to
sured with different procedures. See also the number of cigarettes smoked, and
HETEROMETHOD BLOCK. that the best prediction of the number of
cigarettes that will be smoked the fol-
heterotrait-monomethod coeffi- lowing day is the number smoked on the
cient in a MULTITRAIT-MULTIMETHOD current day rather than the number
MODEL, a number that indicates the smoked on any prior days. See MARKOV
strength of the association between dif-
CHAIN.
ferent characteristics or variables mea-
sured with the same procedures. See also h i d d e n variable an undiscovered
HETEROMETHOD BLOCK. causative variable. When a relationship
heuristic n. in the social sciences, a con- is found between variables x and y, vari-
ceptual device, such as a model or work- able X may erroneously be thought to be
ing hypothesis, that is intended to the cause of y. However, the cause of y
explore or limit the possibilities of a may be a hidden variable z that is corre-
question rather than to provide an ex- lated with variable x. Also called lurk-
planation of the facts. See also CON- ing variable; t h i r d variable. See
STRUCT. THIRD-VARIABLE PROBLEM.

Heywood case any CORRELATION CO- h i e r a r c h i c a l agglomerative clus-

EFFICIENT, REGRESSION COEFFICIENT, t e r i n g see AGGLOMERATIVE CLUSTER-
FACTOR LOADING, Or similar PARAMETER ING.

163
hierarchical clustering

hierarchical clustering a multistage hierarchically nested design any

procedure in which entities are grouped research design that involves several lev-
into ever larger and more heterogeneous els of sampling, such that the entities on
clusters or separated into ever smaUer the lower levels of the design are nested
and more homogeneous clusters (see or subsumed within higher order
CLUSTER ANALYSIS). The most common groups. For example, a hierarchically
form of hierarchical clustering is AG- nested design for an educational study
GLOMERATIVE CLUSTERING, in which might involve students (A), who are
the individual entities are first paired, nested within classrooms (B), which are
then these pairs are paired, and so on nested within schools (C), which are fur-
until ultimately all entities form a single ther nested within school districts (D).
large group or cluster. A second form of Several statistical procedures (e.g., ANAL-
hierarchical clustering is DIVISIVE CLUS- YSIS OF VARIANCE, REGRESSION ANALY-
TERING, in which all entities initially are SIS, HIERARCHICAL LINEAR MODELS) alsO
part of one large group which is split incorporate a hierarchical aspect. Also
into progressively smaller groups ac- called hierarchical design; nested
cording to dissimilarity until ultimately design.
each entity is in its own group or cluster.
The outcome of both types of hierarchi- hierarchical model a statistical proce-
cal clustering may be represented graph- dure that takes into account situations
ically as a DENDROGRAM. in which lower level variables or entities
are part of a larger set or sets. For exam-
hierarchical correlation a pattern in ple, a hierarchical FACTOR ANALYSIS
which variables or entities that are simi- model posits that relationships among a
lar or proximal to one another show a subset of lower level primary variables
higher degree of association than vari- (e.g., verbal, mathematical, and social
ables or entities that are further away. intelligence) can be explained by a
Hierarchical correlations may be exam- higher order or general factor (e.g., gen-
ined via CLUSTER ANALYSIS. For exam- eral intelligence). Similarly, a HIERAR-
ple, individuals belonging to the same CHICAL LINEAR MODEL is a particular
cluster may have similar scores com- type of hierarchical model that includes
pared to individuals who are part of several levels of actual variables as com-
more distant clusters or groups. ponents of a progressively larger overall
h i e r a r c h i c a l design see HIERARCHI- set. Also called nested model.
CALLY NESTED DESIGN. hierarchical regression a statistical
procedure in which hypothesized pre-
h i e r a r c h i c a l linear model (HLM) a
statistical model that acknowledges dif- dictors of a DEPENDENT VARIABLE are in-
cluded in an analysis in several steps
ferent levels in the data, such that indi-
that illuminate the contribution of each
viduals or entities within each level have
correlated scores. The model enables set of variables. For example, a re-
a researcher to test hypotheses about searcher interested in predicting career
cross-level effects and partition the satisfaction could use hierarchical re-
VARIANCE and COVARIANCE compo- gression to assess the contribution of in-
nents among levels. For example, hierar- dividual-level variables (e.g., career
chical linear models often are used in influence), institutional-level variables
educational research since they can ac- (e.g., work climate), and interactional-
count for the fact that students within a level variables (e.g., work respect). In
classroom will behave similarly, as will this form of MULTIPLE REGRESSION, the
classrooms within the same school, and investigator would enter the variables
so on up the hierarchy of levels. Also into the analysis in a prespecified order,
called multilevel model. usually one suggested by temporal fac-

164
 hinge

tors or by a particular theory. After sepa- lem solving, algebra, relationships,

rately adding each of the three variables, communication).
hierarchical regression would enable an
higher order interaction in an
assessment of whether and how much
ANALYSIS OF VARIANCE, the joint effect
the COEFFICIENT OK MULTIPLE DETERMI-
of three or more INDEPENDENT VARI-
NATION increased after each addition.
ABLES on the DEPENDENT VARIABLE. For
Also called hierarchical multiple re-
example, a researcher could conduct a
gression; hierarchical regression
study to assess the effect of a particular
analysis; sequential regression.
treatment (e.g., treatment vs. no treat-
Compare SIMULTANEOUS REGRESSION.
ment) as well as the effect of age and
h i e r a r c h i c a l sum o f squares see socioeconomic status on cognitive func-
TYPE II SUM OF SQUARES. tioning. This design would allow an
examination of the MAIN EFFECTS for
higher order correlation see PAR- treatment, age, and socioeconomic sta-
TIAL CORRELATION. tus individually; the TWO-WAY INTERAC-
TIONS between treatment and age, age
higher order design any research de- and socioeconomic status, and treat-
sign that attempts to understand a phe- ment and socioeconomic status; and the
nomenon by assessing the separate and higher order THREE-WAY INTERACTION
joint effects of several factors at once. between treatment, age, and socioeco-
For example, an experiment could be nomic status.
conducted to assess the effect of a TREAT-
MENT coridition versus a CONTROL CON-
higher order partial correlation
DITION on cognitive functioning while
see PARTIAL CORRELATION.
simultaneously assessing the influence highest posterior density region
of gender; this would allow an assess- the area in a CREDIBLE INTERVAL that has
ment of the MAIN EFFECTS for cach factor the greatest probability of occurrence for
(treatment condition and gender) as a given score.
well as the interaction between them.
Data from such a design involving only high-risk design a research design
one outcome variable could be assessed used to study individuals or groups in
using a higher order ANALYSIS OF VARI- which there is thought to be a high
ANCE. If one or more COVARIATES were probability of some disorder. For exam-
included in the study as additional INDE- ple, a high-risk design could be used to
PENDENT VARIABLES, a higher order study families in which one or more in-
ANALYSIS OF COVARIANCE would be used dividuals have bipolar disorder so as to
for data analysis, and if several measures gauge the likelihood of additional fam-
of cognitive functioning were assessed a ily members subsequently experiencing
higher order MULTIVARIATE ANALYSIS OF the disorder.
VARIANCE would be adopted. h i l l - c l i m b i n g a l g o r i t h m a mathe-
matical procedure for arriving at an ac-
higher order f a c t o r a dimension pr ceptable, although not necessarily ideal,
variable that incorporates other dimen- solution to an optimization problem. A
sions or variables that are at a more basic researcher can use a hill-climbing algo-
or primary level. For example, a higher rithm to make small improvements (i.e.,
order factor of general intelligence could as in inching up a hill) until gradually a
be posited to explain several primary better solution is obtained.
factors of intelligence (e.g., verbal,
mathematical, social), which in turn hinge n. in EXPLORATORY DATA ANALY-
could explain several sets of variables SIS, either of the scores in a data set that
(e.g., vocabulary, comprehension, prob- divide the lower 25% of cases (the lower

165
histogram

20.000
18,000
16.000
14 000
12,000
10:000 mmi
8.000
6,000
A 000
rl".
2. 000

o o
o o o o o o o o
o ooo_o
oo_o
oo_o
o oo
o oo
o oo o
o o o o o
o_ oo_oo_oo_o o_
o o_
o o_ o o. o
o oo oo_oo_ o_ o_
CN co" Tj-" irT (D CO Ci O" CM" CO "C" in to" f^" co" O) o "

Debt (dollars)

histogram

hinge) and the upper 25% of cases (the searcher could examine the health re-
upper hinge) from the remainder of the cords of, say, smokers and nonsmokers
cases. to follow the path of an illness from be-
fore its manifestation through to its di-
histogram n. a graphical depiction of agnosis and treatment. It is distinct from
continuous data using bars of varying PROSPECTIVE RESEARCH generally, which
height, similar to a BAR GRAPH but with begins with individuals who are appar-
blocks on the A;-axis adjoining one an- ently healthy in the present and moves
other so as to denote their continuous forward to investigate whether a specific
nature. disorder will occur over time.
Consider the example above, showing
the average credit card debt of indivi- h i s t o r i g r a m n. a graph of quantitative
duals. Bars along the A-axis represent TIME-SERIES data, with the value of
amount of debt and are connected to the variable being examined given on
one another, while the heights of the the vertical Y-AXIS and time given on
bars represent the number or frequency the horizontal X-AXIS. For example, a
of individuals with each debt amount. historigram for a weight reduction pro-
gram might depict daily weight along
historical analysis research that ex- the y-axis over a six-month time period
amines past events to understand cur- along the x-axis (see overleaf).
rent or future events. For example,
If the weight reduction program was
researchers could perform a historical
effective then the historigram would be
analysis of an individual's or a family's
expected to slope downward from a high
substance use experiences to understand
point at the upper left.
the present substance use behavior of
that person or group. history effect the influence of events
or circumstances outside an experiment
h i s t o r i c a l c o n t r o l group a CONTROL
on an outcome variable of interest.
GROUP whose participants are selected
QUASI-EXPERIMENTAL RESEARCH Often
to be similar to those in the treatment
attempts to take history effects into ac-
group on the basis of data collected
count in order to rule out potential CON-
some time in the past.
FOUNDS to a posited or apparent link
historical prospective study a re- between two variables. An example
search project that examines LONGITU- would be the occurrence of an actual
DINAL DATA obtained in the past to track earthquake during a field study of the ef-
the incidence of a particular disorder fects of training in earthquake prepared-
over time and its association with vari- ness; this would likely increase news
ous risk factors. For example, a re- coverage of earthquakes, thereby result-
166
 honestly significant difference test

ing in greater knowledge dissemination SIS OF VARIANCE if the average values of

outside of the training program. the population groups being investi-
history t a k i n g the process of compil- gated were the same. Compare HETERO-
GENEITY.
ing background information about a
patient or research participant. Informa- homogeneity of covariance the
tion can be obtained from the individual condition in which multiple groups in
directly and from other sources, such as an experimental design have the same
the patient's family, hospitals or clinics, COVARIANCE MATRIX. A baslc assump-
psychiatrists or psychologists, neurolo- tion in a MULTIVARIATE ANALYSIS OF
gists, social workers, and others who VARIANCE, it is the multivariate analog
have direct knowledge ofthe individual. of HOMOGENEITY OF VARIANCE.
HLM abbreviation for HIERARCHICAL homogeneity o f regression the con-
LINEAR MODEL. dition in which the regression slopes be-
Hodges-Lehmann estimator an ES- tween the dependent variable of interest
TIMATOR used to approximate the me- and the covariates are equal for all popu-
dian value of a population, when lation groups. It is a basic assumption in
conventional PARAMETRIC STATISTICS an ANALYSIS OF COVARIANCE.
are not appropriate as the data do not
meet statistical ASSUMPTIONS. It is an ex- homogeneity o f variance the statis-
ample of a ROBUST ESTIMATOR. [Joseph tical ASSUMPTION of equal variance,
L. Hodges (1922-2000) and Erich L. Leh- meaning that the average squared dis-
mann (1917-2009), U.S. statisficians] tance of a score from the mean is the
same across all groups sampled in a
holdout sample see CROSS-VALIDA- study. This condition must be fulfilled
TION SAMPLE. in statistical methods that use a single
holdover effect see CARRYOVER EF- term to represent how widely scores vary
FECT.
across groups, as with ANALYSIS OF VARI-
ANCE, MULTIPLE REGRESSION analysis,
HoUand-Rosenbaum test o f u n i - and other procedures. Also called equal-
d i m e n s i o n a l i t y see ROSENBAUM'S ity of variance; homoscedasticity.
TEST OF UNIDIMENSIONALITY. Compare HETEROGENEITY OF VARIANCE.
homogeneity n. equality or near equal- homoscedasticity n. see HOMOGENEITY
ity between two statistical quantities of
OF VARIANCE. homoscedastic adj.
interest. The term most often is used in
connection with different populations. honestly s i g n i f i c a n t difference
For example, homogeneity of means test see TUKEY'S HONESTLY SIGNIFICANT
would be present in a ONE-WAY ANALY- DIFFERENCE TEST.

250

2 3 4

Time (months)
historigram

167
horizontal axis

h o r i z o n t a l axis see X-AXIS. streak goes on. This is in defiance of the

laws of probability. A gambler's belief
hot-deck i m p u t a t i o n one of several that his or her winning streak on a ca-
methods of inserting values for missing sino slot machine will continue is an ex-
data (see IMPUTATION) in which missing ample of the hot-hand hypothesis. Also
observations or data points are replaced called streaky hypothesis.
by values from similar responses in the
sample at hand. Suppose that a patient household i n t e r v i e w survey any se-
survey was conducted in two hospitals, ries of door-to-door or telephone inter-
A and B, and that five people from Hos- views with household residents in which
pital A failed to respond to an item of the samples are based on geographic bound-
survey. Sampling five values from the re- aries, such as cities and counties. The na-
spondents who did respond to that item tional census conducted every 10 years
at Hospital A and substituting these val- by the United States Census Bureau is an
ues for the missing observations is an example of a household interview sur-
example of a hot-deck imputation. vey.
Compare COLD-DECK IMPUTATION. HSD test abbreviation for honestly sig-
Hotelling's trace a statistic used in a nificant difference test. See TUKEY'S HON-
MULTIVARIATE ANALYSIS OF VARIANCE tO ESTLY SIGNIFICANT DIFFERENCE TEST.
evaluate the significance of any MAIN EF- H spread see INTERQUARTILE RANGE.
FECTS and INTERACTION EFFECTS that
have been identified. It is the pooled hyperbolic d i s t r i b u t i o n a continu-
ratio of the amount of variance in an ef- ous PROBABILITY DISTRIBUTION the loga-
fect that is attributed to the independent rithm of which is a hyperbola (see
variable (i.e., the EXPLAINED VARIANCE) example opposite).
to the amount of variance that remains The distribution is similar in shape to
unexplained (i.e., the ERROR VARIANCE). the NORMAL DISTRIBUTION but de-
Also called Hotelling-Lawley trace. creases more slowly and is generally
See also PILLAI-BARTLETT TRACE; WILKS'S asymmetric around its MODE.
LAMBDA. [Harold Hotelling (1895-1973), hyperbolic l o g a r i t h m see LOGA-
U.S. mathematician; D. N. Lawley] RITHM.
Hotelling's test a MULTIVARIATE hyperexponential d i s t r i b u t i o n a
ANALYSIS OF VARIANCE used to test the continuous distribution in which the
significance of the mean difference be- probability of obtaining a particular
tween two groups in their scores or out- value for a RANDOM VARIABLE, X, de-
comes on multiple dependent variables. pends on, or is a FUNCTION of, the pro-
An extension of the univariate T TEST, it portionally increasing or decreasing
is applicable to only two groups, rather values of another random variable, y.
than three or more groups. Also called
Hotelling's Hotelling's T- hypergeometric d i s t r i b u t i o n a dis-
crete PROBABILITY DISTRIBUTION Of the
squared test; multivariate t test; T-
number of successes (or failures) in a se-
squared test. [Harold Hotelling]
quence of draws that involve SAMPLING
hot-hand hypothesis the widespread WITHOUT REPLACEMENT; in this situa-
but groundless belief that a string of suc- tion each trial has two possible out-
cesses will breed further success. Accord- comes and the outcome of each trial
ing to this belief, which is common in alters the probability of the outcome of
professional sports, financial specula- the next and succeeding trials. If the
tion, and the like, the probability that a sampling is done with replacement a BI-
mn of successes will continue becomes NOMIAL DISTRIBUTION Is Obtained in-
greater the longer that the winning stead.

168
 hypothetico-deductive method

0.8

o
0.6

g 0.4

Q
.d
P 0.2

Logx

hyperbolic distribution

hyperparameter n. in BAYESIAN statis- design, the extent to which any results

tics, a numerical constant describing a obtained from a study are likely to pro-
probability within a population that is vide evidence for or against the hypoth-
itself based upon a distribution of prior esis under investigation. If the hypoth-
probabilities. For example, if p is the esis validity of a study is low, even clear
probability of success in each of a series results are unlikely to provide either sup-
of BERNOULLI TRIALS and a BETA DISTRI- port for, or refutation of, theoretically
BUTION with parameters a and b (a, b> derived predictions in a decisive and un-
0) is chosen as the PRIOR DISTRIBUTION ambiguous manner.
for the parameter p, then a and b are
h y p o t h e t i c a l construct see CON-
hyperparameters.
STRUCT.
hypothesis n. (pl. hypotheses) an em-
pirically testable proposition about some hypothetico-deductive method a
fact, behavior, relationship, or the like, method of scientific inquiry in which
usually based on theory, that states an the credibility or explanatory power of a
expected outcome resulting from spe- falsifiable hypothesis is tested by mak-
cific conditions or assumptions. ing predictions on the basis of this hy-
pothesis and determining whether or
hypothesis testing a statistical infer- not these predictions are consistent with
ence procedure for determining whether empirical observations. It is one of the
or not a given proposition about a popu- most widely used SCIENTIFIC METHODS
lation PARAMETER should be rejected on for disproving hypotheses and building
the basis of observed sample data. See corroboration for those that remain.
also SIGNIFICANCE TESTING. Also called mathematico-deductive
hypothesis v a l i d i t y in experimental method.

169
 Ii
I symbol for WILLIAMS'S AGREEMENT this case, an infinite number of combi-
MEASURE. nations of their values may yield the
same model prediction, such that one
ICC 1. abbreviation for INTRACLASS COR- cannot determine from a given set of ob-
RELATION coefficient. 2. abbreviation for
servations which values of the parame-
ITEM CHARACTERISTIC CURVE. ters actually generated the data.
IDA abbreviation for INITIAL DATA ANAL- identity m a t r i x (symbol: I) a SQUARE
YSIS. MATRIX with values of one along its
idealized experiment see THOUGHT main diagonal and zeros elsewhere. The
EXPERIMENT. following is an example.
i d e n t i f i c a t i o n n. in CONFIRMATORY
1 0 0
FACTOR ANALYSIS and STRUCTURAL EQUA-
0 1 0
TION MODELING, a situafion in which the
model contains a sufficient number of 0 0 1
both fixed and free PARAMETERS to result
in unique estimates from the observed Identity matrices are used in FACTOR
data. A model is said to be identified or ANALYSIS solutions and matrix algebra.
identifiable if a unique set of its parame- Also called unit matrix.
ter values can be determined from observa- i d i o g r a p h i c adj. relating to the descrip-
fions. OVERIDENTIFICATION occurs when tion and understanding of an individual
there are more knowns than free param- case, as opposed to the formulation of
eters, and UNDERIDENTIFICATION OCCUrS general laws describing the average case.
when it is not possible to estimate all An idiographic approach involves the
of the model's parameters. See also IDEN- thorough, intensive study of a single
TIFICATION PROBLEM; OVERIDENTIFIED person or case in order to obtain an in-
MODEL; UNDERIDENTIFIED MODEL. depth understanding of that person or
i d e n t i f i c a t i o n p r o b l e m in CONFIR- case, as contrasted with a study of the
MATORY FACTOR ANALYSIS and STRUC- universal aspects of groups of people or
TURAL EQUATION MODELING, the problem cases. In those areas of psychology in
as to whether the PARAMETER estimates which the individual person is the unit
obtained are stable given the number of of analysis (e.g., in personality, develop-
subjects and variables in the observed mental, or cUnical psychology), the
sample data. The problem arises from idiographic approach has appeal be-
the fact that more than one set of param- cause it seeks to characterize a particular
eter values may be associated with the individual, emphasizing that individ-
same PROBABILITY DISTRIBUTION of a ual's characteristic traits and the unique-
statistical model. When the identifica- ness of the individual's behavior and
tion problem arises, a model is said to be adjustment, rather than to produce a
unidentifiable and its parameter values universal set of psychological constructs
cannot be interpreted. For example, as- that might be applicable to a popula-
sume that the two parameters b and c in tion. Compare NOMOTHETIC
the linear regression model y = ax + b + c IID independent and identically distri-
-I- N(0, o^) are fully interchangeable. In buted: denoting a collection of random

170
 incidental stimulus

variables that have the same PROBABIL- reotypes, and emotions in social cogni-
ITY DISTRIBUTION but are statisfically un- tion research. Typically, an implicit
related to one another. For example, a measure is assessed as a response out-
sequence of random tosses of a fair coin come of an experimental procedure in
is IID, as is a sequence of colored balls which the participant is engaged in a
randomly drawn from an urn with re- cognitive task. For example, a word-stem
placement after each draw. IID is a key completion task might be employed to
assumption in the CENTRAL LIMIT THEO- assess emotion implicitly, such that jo_
REM and also a standard assumption in could be completed to form a positive
probability and statistical theory. emotional word (e.g., joy) or a neutral
word (e.g., jog).
ill-conditioned adj. denoting a
SQUARE MATRIX in which the ratio of importance s a m p l i n g a MONTE
the largest EIGENVALUE to the smallest CARLO RESEARCH method for computing
eigenvalue is much larger than 1, such expected values of a RANDOM VARIABLE
that small changes to the entries in the in which samples are generated from a
matrix result in a significantly large distribution similar to the distribution
change in its INVERSE MATRIX. An ill- of interest, which cannot be directly
conditioned matrix thus is very sensitive sampled. The method is particularly use-
to inconsequential changes (e.g., round- ful for high-dimensional models with
ing errors in computer calculations), many variables or parameters and is fre-
thereby making the calculations it yields quently used in BAYESIAN statistics.
difficult to interpret. See also COLLIN-
imposed etics the imposition of one's
EARITY.
own cultural perspective and bias when
illusory correlation 1. the appear- studying another culture. See ETICS.
ance of a relationship that in reality does i m p r o p e r p r i o r in BAYESIAN statistics,
not exist. 2. an overestimation of the de- a PRIOR DISTRIBUTION that does not in-
gree of relationship (i.e., correlation) be- tegrate to one. Although an improper
tween two variables. For example, if an prior in itself is not necessarily problem-
unusual action occurred at the same atic, especially when the corresponding
time that an adolescent was present, the POSTERIOR DISTRIBUTION is proper, it
assumption that the action was carried can sometimes lead to an uninterpret-
out by the adolescent would be an illu- able and misleading estimate of the
sory correlation. BAYES FACTOR.
impact analysis a quantitative ana- i m p u t a t i o n n. a procedure for filling in
lytic procedure used to assess the net missing values in a data set before ana-
success or failure of a program, usually lyzing the resultant completed data set.
through controlled experimentation. It There are several methods of imputa-
is appropriate only if the program's ob- tion, including HOT-DECK IMPUTATION,
jectives are specifiable and measurable, in which missing valiies are replaced
the program is well implemented for its with values drawn randomly from a sim-
intended participants, and the outcome ilar sample in the same data set; and re-
measures are reliable and valid. Also gression imputation, in which missing
called impact assessment. See also values are replaced with predicted values
SUMMATIVE EVALUATION. estimated from a statistical model of the
nonmissing values in the data. See also
i m p l i c i t measure a measurement of a
COLD-DECK IMPUTATION; MULTIPLE IM-
psychological constmct that is obtained
PUTATION.
while the individual being assessed is
unaware that the measurement is taking i n c i d e n t a l stimulus an unintentional
place, often used to assess attitudes, ste- or coincidental stimulus that may occur

171
inclusion-exclusion criteria

during an experiment or study, which COMPLETE COUNTERBALANCING. See

may elicit an unplanned response from also LATIN SQUARE.
the participants or result in the distor-
incomplete f a c t o r i a l design see
tion of research findings.
FRACTIONAL FACTORIAL DESIGN.
inclusion-exclusion c r i t e r i a in incremental v a l i d i t y the improve-
clinical research, criteria used to deter- ment obtained by adding a particular
mine which individuals are eligible to procedure or method to an existing
participate in a particular study. Inclu- combination of assessment methods. In
sion criteria might specify, for example, other words, incremental validity re-
age range, whereas exclusion criteria flects the value of each measure or piece
might specify, for example, the exis- of information to the process and out-
tence of more than one illness or psy- come of assessment. The standards for
chological disorder. evaluating incremental validity depend
incomplete b l o c k design an experi- on the goal of the assessment, such as
mental design in which treatments are whether one wishes to gather unique in-
grouped into sets or "blocks," not all of formation, predict a criterion, make a di-
which include every treatment, and agnosis, or choose a treatment. For
each block is administered to a different example, teacher observations of the
group of participants. Incomplete block daily frequency of a child's vocal tic may
designs often are employed to avoid ad- not add critical information in diagnos-
ministering too many treatment condi- ing a tic disorder but may contribute sig-
tions to the same group of participants, nificantly when the purpose of the
which may contribute to undesirable assessment is to monitor treatment
error. For instance, the following ar- progress.
rangement of four treatments (A, B, C,
independence n. 1. the condition of
D) is an incomplete block design: Block being unrelated to or free from the influ-
1 = A-B-C; Block 2 = A-B-D; Block 3 = A- ence of something else. For example, a
C-D; and Block 4 = B-C-D. In this partic- standard assumption in ANALYSIS OF
ular design, each block contains three VARIANCE, MULTIPLE REGRESSION, and
treatments, each treatment occurs three many other statistical analyses is inde-
times across all blocks, and each pair of pendence of observations, or the fact
treatments occurs together two times that the occurrence of one observation
across all blocks. An incomplete block does not influence the occurrence of any
design with such properties is called a others. Similarly, variables exhibiting in-
balanced incomplete block design. dependence have a complete lack of re-
Compare COMPLETE BLOCK DESIGN. See lationship, such that none is influenced
BLOCK DESIGN. by any other and that changes in one
have no implication for changes in any
incomplete counterbalancing an other. See also INDEPENDENT EVENTS.
experimental design that controls for 2. see STATISTICAL INDEPENDENCE.
ORDER EFFECTS by using a limited num- independent adj., n.
ber of possible sequences of treatments
independence model any statistical
administered in such a way that each
treatment appears equally often in each representation of a concept or process in
position. For instance, the following ar- which explanatory variables are as-
rangement of sequences of three treat- sumed not to interact and instead to be
ments (A, B, C), each assigned to a unaffected by one another.
different subgroup of participants, dem- independent c o n t r i b u t i o n in MUL-
onstrates incomplete counterbalancing: TIPLE REGRESSION, the portion of the
A-B-C to Subgroup 1, B-C-A to Subgroup variance that can be explained uniquely
2, and C-A-B to Subgroup 3. Compare by a given explanatory variable after

172
 independent variable

PARTIALING out Ot holding Constant the these groups. Also called independent
effects of all other explanatory variables. groups. Compare DEPENDENT SAMPLES.
See also COLLINEARITY.
independent-samples analysis o f
independent diagonals model see variance see BETWEEN-SUBJECTS ANAL-
DIAGONALS MODEL. YSIS OF VARIANCE.
independent events the situation in independent-samples design see BE-
which observing one event does not pro-
TWEEN-SUBJECTS DESIGN.
vide any additional information about
the occurrence or outcome of another independent-samples t test a T TEST
event. For example, the outcome of a used to analyze data from a BETWEEN-
coin flip and the Dow Jones Industrial SUBJECTS DESIGN, in which the different
Average are independent events, whereas groups of individuals or other entities
the temperatures of two consecutive measured are not associated with one
days are not necessarily independent. another. In the independent-samples t
Formally, two events, A and B, are inde- test, one calculates the value of the TEST
pendent if the probability of event A is STATISTIC t from the means, standard
the same as the conditional probability deviations, and sizes of the two groups
of event A given event B, that is, P(A) = of interest, and compares its value to a
P(A\B), or equivalently, P(B) = P(B\A). T DISTRIBUTION for a given DEGREE OF
Compare DEPENDENT EVENTS. FREEDOM under the assumptions of nor-
independent groups see INDEPEN- mality and INDEPENDENCE of Obser-
DENT SAMPLES. vations, among others. Also called inde-
pendent-measures t test; unrelated
independent-groups analysis o f t test. Compare DEPENDENT-SAMPLES T
variance see BETWEEN-SUBJECTS ANAL- TEST.
YSIS OF VARIANCE.
independent s a m p l i n g a process for
independent-groups design see BE-
selecting a sample of study participants
TWEEN-SUBJECTS DESIGN.
from a larger potential group of individ-
independent-measures analysis o f uals such that the probability of each
variance see BETWEEN-SUBJECTS ANAL- person being selected for inclusion is not
YSIS OF VARIANCE. influenced by which people have been
independent-measures design see chosen already. The resulting samples
BETWEEN-SUBJECTS DESIGN. will be INDEPENDENT SAMPLES.
independent r a n d o m s a m p l i n g see independent variable (IV) the vari-
SIMPLE RANDOM SAMPLING. able in an experiment that is specifically
manipulated or is observed to occur be-
independent random variables fore the occurrence of the dependent, or
RANDOM VARIABLES that exhibit a com- outcome, variable, in order to assess its
plete lack of relationship, such that no effect or influence. Independent vari-
information about one variable, x, con- ables may or may not be causally related
veys any information about another to the DEPENDENT VARIABLE. In Statisti-
variable, y. Any events related to these cal analysessuch as PATH ANALYSIS,
variables are INDEPENDENT EVENTS.
REGRESSION ANALYSIS, and STRUCTURAL
independent samples groups of indi- EQUATION MODELINGan independent
viduals or sets of data that are unrelated variable is likely to be referred to as a
to one other. For example, experimental causal variable, explanatory variable,
groups consisting of different and unre- regressor variable, or PREDICTOR VARI-
lated participants are independent sam- ABLE. See also EXOGENOUS VARIABLE;
ples, as are the data sets obtained from TREATMENT.

173
indeterminacy

indeterminacy n. 1. the inability to index of v a r i a b i l i t y see DISPERSION

uniquely determine the form or magni- MEASURE.
tude of a relationship. 2. the inability to
i n d i c a t o r variable 1. see DUMMY
arrive at a unique solution to a problem
VARIABLE. 2. see MANIFEST VARIABLE.
or mathematical form. 3. in FACTOR
ANALYSIS, the inability to form a unique indifference p o i n t the intermediate
representation of the factor structure. region between experiential opposites.
For example, on the pleasure-pain di-
index n. 1. a reference point, standard,
mension, it is the degree of stimulation
or indicator. 2. a variable that is em-
that provokes an indifferent or neutral
ployed to indicate the presence of an-
response.
other phenomenon or event. 3. a
number formed from a combination of i n d i f f e r e n t stimulus any stimulus
other measures to represent another, that has not yet elicited the reaction
more general entity. For example, an being studied.
index of a person's graduate school po-
indirect correlation see NEGATIVE
tential might be formed from his or her
CORRELATION.
undergraduate grade point average and
score on an admissions test. indirect measurement a method in
which a researcher gathers data about
index o f determination see COEFFI- one variable (or a combination of vari-
CIENT OF DETERMINATION.
ables) as a means of representing a sec-
index of discrimination the degree ond variable of interest, which cannot
to which a test or test item differentiates be assessed in a more straightforward
between individuals of different perfor- manner. See PROXY VARIABLE.
mance levels, often given as the percent-
indirect relationship in STRUCTURAL
age difference between high-performing
EQUATION MODELING, a Correlation be-
and low-performing individuals who an- tween two variables that involves an in-
swer a target item correctly. Also called tervening variable or MEDIATOR. For
discrimination index. example, age may affect rate of pay,
index o f dispersion see DISPERSION which in turn may affect job satisfac-
MEASURE. tion: This being so, the correlation be-
tween age and job satisfaction would be
index o f forecasting efficiency in an indirect relationship. The term is
LINEAR REGRESSION, a measure of the often used incorrectly to denote a NEGA-
amount of reduction in prediction errors TIVE RELATIONSHIP. Compare DIRECT
relative to the baseline prediction using RELATIONSHIP.
the mean of the dependent variable,
when prediction errors are measured by i n d i v i d u a l comparison see COMPAR-
the STANDARD ERROR OF ESTIMATE. ISON.
Formally, the index of forecasting effi- i n d i v i d u a l differences traits or other
ciency is defined as 1 - V (1 - r^), where characteristics by which individuals may
is the COEFFICIENT OF DETERMINATION.
be distinguished from one another. See
An index value of 0 indicates no reduc- also SUBJECT VARIABLE.
tion in errors beyond the mean predic-
tion, whereas a value of 1 indicates the inflividual-differences scaling (IND-
full, maximum possible reduction. SCAL) a method of studying how per-
sonal attributes influence judgments
index of r e l i a b i l i t y see RELIABILITY about the similarity among members or
COEFFICIENT. items of a category. A form of MULTIDI-
index of v a l i d i t y see VALIDITY COEFFI- MENSIONAL SCALING, it assumes that
CIENT. people may differ on both the character-

174
 informative prior

istics they use to define a category and i n f e r e n t i a l statistics a broad class of

the importance they place on each char- statistical techniques that allow INFER-
acteristic. The method not only yields ENCES about characteristics of a popula-
an overall PROXIMITY MATRIX for a tion to be drawn from a sample of data
group of individuals as a whole but also from that population while controlling
provides a unique proximity matrix for (at least partially) the extent to which er-
each individual in the group. Also called rors of inference may be made. These
weighted multidimensional scal- techniques include approaches for test-
ing. ing hypotheses, estimating the value of
parameters, and selecting among a set of
individual-difference variable see
competing models. Also called induc-
SUBJECT V A R I A B L E .
tive statistics; interpretive statis-
INDSCAL abbreviation for INDIVIDUAL- tics. Compare DESCRIPTIVE STATISTICS.
DIFFERENCES SCALING.
i n f e r e n t i a l test any statistical proce-
induction n. 1. a general conclusion, dure used to evaluate hypotheses about
principle, or explanation derived by differences between sample and popula-
reasoning from particular instances or tion distributions. Examples include the
observations. Compare DEDUCTION. 2. CHI-SQUARE GOODNESS-OF-FIT TEST, the
the process of inductive reasoning itself. F TEST, and the T TEST. Inferential tests
inductive adj. more commonly are known as signifi-
cance tests (see SIGNIFICANCE TESTING).
inductive statistics see INFERENTIAL
STATISTICS. i n f e r e n t i a l v a l i d i t y the extent to
which causal inferences made in a labo-
inequality constraint in parameter ratory setting are applicable to the real-
ESTIMATION and other optimization life experiences they are meant to repre-
problems, a condition requiring that sent. See also EXTERNAL VALIDITY.
particular items do not have the same
value. For example, one may wish to i n f i n i t e sequence an ordered list of
find the values of a and b that minimize numbers or items that continues with-
a function f(a, b) = 3a* - Sa^b -i- 2b^ + 10 out end. Examples of an infinite se-
under the following inequality con- quence are
straints: 0 < a < 10 and b <0. Compare 5= |1, 1/2, 1/4, 1/8, 1/16, ...]
EQUALITY CONSTRAINT.
and
inference n. 1. a conclusion deduced
from an earlier premise or premises ac- V={a, ab, ab^, ab^, ab^, ...).
cording to valid mles or procedures for The sum of the terms of an infinite se-
drawing such a conclusion. Some hold quence is called an infinite series.
that an inference, as contrasted with a
mere conclusion, requires that the per- influence analysis a set of techniques
son making it actually believes that the that allows one to determine the degree
inference and the premises from which to which specific data points affect the
it is drawn are true. Also called logical overall result of a statistical procedure.
inference. 2. in statistical analysis, a For example, in REGRESSION ANALYSIS it
conclusion about a population based on is used to examine which observations
logical reasoning from data gathered have a disproportionate influence on
about a smaller sample (see INFERENTIAL the proposed REGRESSION EQUATION.
STATISTICS). The most common exam-
ple of this type of inference is statistical i n f o r m a t i o n m a t r i x see FISHER'S IN-
hypothesis testing (see INFERENTIAL FORMATION.
TEST). inferential adj. i n f o r m a t i v e p r i o r in BAYESIAN statis-

175
informed consent

tics, a type of PRIOR DISTRIBUTION based dents, finances, technology, campus cli-
on large amounts of previously accumu- mate, and other characteristics so as to
lated knowledge about the phenome- facilitate better decision making.
non of interest. An informative prior
i n s t i t u t i o n a l review b o a r d (IRB) a
tends to provide more specific informa-
committee named by an agency or insti-
tion and have a stronger influence upon
tution to review research proposals orig-
calculations of an associated POSTERIOR
inating within that agency for ethical
DISTRIBUTION than a NONINFORMATIVE
acceptability and compliance with the
PRIOR.
organization's codes of conduct. IRBs
i n f o r m e d consent a person's volun- help protect research participants and
tary agreement to participate in a proce- are mandatory at any U.S. institution re-
dure on the basis of his or her ceiving federal funds for research.
understanding of its nature, its potential i n s t i t u t i o n a l survey in INSTITU-
benefits and possible risks, and available TIONAL RESEARCH, a questionnaire or in-
alternatives. Informed consent is a fun- terview used to collect information
damental requirement of research with about a university and its operations.
humans and typically involves having
participants sign documents prior to the i n s t r u c t i o n a l variable 1. any INDE-
start of a study describing specifically PENDENT VARIABLE that is studied in
what their involvement would entail researching the outcomes and effective-
and noting that they are free to decline ness of learning programs. Examples of
participation or to withdraw from the re- instmctional variables include the spe-
search at any time. See also INSTITU- cific activities used, instructor quality,
TIONAL REVIEW BOARD. learner workload, and learner motiva-
tion. 2. more generally, any set of direc-
i n i t i a l data analysis (IDA) an exami- tions manipulated in an experiment. For
nation of the quality of a data set that is example, a researcher investigating goal-
undertaken before beginning more com- setting may wish to assess how the per-
plex statistical tests of hypotheses. It formance of participants who are in-
often involves calculating DESCRIPTIVE structed simply to do their best differs
STATISTICS such as the mean and STAN- from that of participants told to try to
DARD DEVIATION, identifying minimum achieve a specified target level.
and maximum values, and graphing the
data points to determine whether their instrument n. any tool, device, or other
arrangement follows a known or ex- means by which researchers assess or
pected distribution. IDA is important for gather data about study participants. Ex-
ensuring that data conform to assump- amples include tests, INTERVIEWS, QUES-
tions of the statistical models that will TIONNAIRES, SURVEYS, RATING SCALES,
be used for subsequent formal analysis. and reaction-time apparatus. See also AS-
SESSMENT INSTRUMENT.
i n l i e r n. an incorrect value in a data set.
Inliers may arisefromSYSTEMATIC ERROR, i n s t r u m e n t a l i s m n. 1. in the philoso-
respondent error, or processing error. phy ofscience, the position that theories
For example, a researcher may inadver- are not to be considered as either true or
tently report a value in the wrong unit, false but as instmments of explanation
such as including a RAW SCORE in a dis- that allow observations of the world to
tribution of STANDARDIZED SCORES.
be meaningfully ordered. 2. a theory of
knowledge that emphasizes the prag-
i n s t i t u t i o n a l research a study con- matic value, rather than the truth value,
ducted to obtain information about an of ideas. In this view, the value of an
academic setting. For example, a univer- idea, concept, or judgment lies in its
sity may research its faculty, staff, stu- ability to explain, predict, and control
 interaction contrast

one's concrete functional interactions students in a school. Intact groups often

with the experienced world. This view is are used in studies using a NONEQUIVA-
related to PRAGMATISM. instrumen- LENT-GROUPS DESIGN.
talist adj., n.
integral n. 1. a whole number (e.g., 2, 9,
i n s t r u m e n t a l variable in REGRES- 11), as opposed to numbers that are frac-
SION ANALYSIS, STRUCTURAL EQUATION tions (e.g., 1/4, 1.5). 2. the area under-
MODELING, or PATH ANALYSIS, 3 new neath a curve plotted on an ;if-axis and y-
variable substituted for an INDEPENDENT axis between two points, a and b. It is
VARIABLE when the latter is correlated one of two main conceptsthe other
with the ERROR TERM. Often a MEDIA- being the DERIVATIVEapplied within
TOR, an instrumental variable must be calculus to the study of change and area.
associated with the independent vari- Also called antiderivative.
able it is replacing.
intention-to-treat analysis in RAN-
i n s t r u m e n t a t i o n n. the creation and DOMIZED CLINICAL TRIALS, a Strategy for
use of equipment and devices (including minimizing BIAS in which a researcher
psychological tests) for the measure- examines data for all patients who met
ment of some attribute or the control of the criteria for initial entry into the trial
experiments (e.g., automation of stimu- and were to receive the experimental
lus presentation and data collection). It treatment, regardless of whether or not
is important that a measurement instm- they actually complete the trial or expe-
ment not only be trustworthy and reli- rience the treatment. For example,
able but also properly maintained and many people enrolled in a weight-loss
consistently administered so as to en- study for whom the new diet being in-
sure INTERNAL VALIDITY. vestigated is ineffective may drop out.
i n s t r u m e n t a t i o n effect any change Since the remaining participants primar-
in the DEPENDENT VARIABLE in a study ily are those actually losing weight, the
that arises from changes in the measur- new diet may seem to be more beneficial
ing instrument used. For example, an than it actually is if the researcher were
experimenter may subtly but uninten- to examine data only for the individuals
tionally alter his or her testing method who finished the trial.
across sessions, such that a pretest and a interaction analysis the identification
posttest are administered differently. In- and examination of an INTERACTION EF-
stmmentation effects pose a threat to IN- FECT among variables using INTERAC-
TERNAL VALIDITY. See also INSTRUMENT
TION CONTRASTS and other techniques.
DRIFT.
interaction contrast the statistical
instrument d r i f t changes in an IN-
comparison of specific levels of the vari-
STRUMENT, usually gradual and often
ous INDEPENDENT VARIABLES Studied in
predictable, that can threaten the valid-
a FACTORIAL DESIGN. For example, sup-
ity of conclusions drawn from the data
obtained with that instmment. An ex- pose a researcher obtains a significant F
ample would be the stretching of spring RATIO in an ANALYSIS OF VARIANCE and
scales. Periodic recalibration of the mea- discovers that three methods of instruc-
suring device helps minimize or rectify tion (Aj, A2, A3) and two types of con-
instrument drift. Also called instru- tent (BJ, B2) jointly influence student
ment decay. See also OBSERVER DRIFT. academic performance. He or she could
then use several different interaction
intact group a naturally occurring col- contrasts to examine which specific
lection of individuals, rather than one to combinations of methods and content
which participants have been randomly have the greatest influence. Thus, the re-
assigned. An example is a classroom of searcher may evaluate how Aj and A2 in-

177
interaction effect

teract with Bj and B2, how Aj and A3 i n t e r a c t i o n sum o f squares the SUM
interact with B, and B2, and so forth. In- OF SQUARES associated with the INTER-
teraction contrasts may be undertaken ACTION EFFECT between two or more
as suggested by ongoing evaluation of variables. In an ANALYSIS OF VARIANCE
study data (see POST HOC COMPARISON) or MULTIPLE REGRESSION, the total sum
or may be decided upon in advance (see of squares can be divided into that for
A PRIORI COMPARISON). Also called in- each variable individually, that for the
teraction comparison. interaction of the variables, and that due
to error.
interaction effect in a FACTORIAL DE-
SIGN, the joint effect of two or more in- i n t e r a c t i o n variance in a FACTORIAL
dependent variables on a dependent DESIGN, the amount of variance in the
variable above and beyond the sum of dependent variable that is explained by
their individual effects: The indepen- an INTERACTION EFFECT between two or
dent variables combine to have a differ- more independent variables.
ent (and multiplicative) effect, such that
intercept n. the point at which either
the value of one is contingent upon the
axis of a graph is intersected by a line
value of another. This indicates that the
plotted on the graph. For example, it is
relationship between the independent
the value of y when ;if = 0 in an equation
variables changes as their values change.
of the form y = a + bx, or the value of x
For example, if a researcher is studying
where a REGRESSION LINE crosses the y-
how gender (female vs. male) and diet-
axis, as shown below. See also X-INTER-
ing (Diet A vs. Diet B) influence weight
CEPT; Y-INTERCEPT.
loss, an interaction effect would occur if
women using Diet A lost more weight interclass correlation the degree of
than men using Diet A. Interaction ef- bivariate relationship between two vari-
fects contrast withand may obscure ables from different measurement classes
MAIN EFFECTS. See also HIGHER ORDER (i.e., the variables have distinct metrics
INTERACTION. Compare ADDITIVE EFFECT. and variances). For example, one may

intercept

178
 interquartile range

use an interclass correlation to deter- homogeneity. See also INTERITEM COR-

mine the relation of IQ points (a class of RELATION.
measurement representing aptitude) to
grade point averages (a class of measure- i n t e r n a l v a l i d i t y the degree to which
ment representing achievement), or the a study or experiment is freefromflaws
relation of length (e.g., inches) to weight in its internal structure and its results
(e.g., pounds). Currently, the only inter- can therefore be taken to represent the
class correlation in common use is the true nature of the phenomenon. In
Pearson PRODUCT-MOMENT CORRELA- other words, internal validity pertains to
TION COEFFICIENT. Compare INTRACLASS the soundness of results obtained within
CORRELATION. the controlled conditions of a particular
study, specifically with respect to
interclass variance a rare synonym of whether one can draw reasonable con-
between-groups variance. See BETWEEN- clusions about cause-and-effect relation-
GROUPS MEAN SQUARE. ships among variables. Compare EXTER-
intercoder reliability see INTERRATER NAL VALIDITY.
RELIABILITY.
interobserver r e l i a b i l i t y see INTER-
intercorrelation n. the CORRELATION RATER RELIABILITY.
between each variable and every other
variable in a group of variables. i n t e r p o l a t i o n n. a strategy for deter-
mining an unknown value given knowl-
interdecile range the difference be- edge of surrounding data points. For
tween the ninth and first DECILES of a example, if the average score on a test is
distribution (i.e., the number obtained 70 for a group of beginner students and
by subtracting the 10th PERCENTILE the average score on the same test is 90
score from the 90th percentile score). for an advanced group of students, one
The interdecile range provides a measure might use interpolation to estimate an
of DISPERSION in a data set that is mini- average score of 80 on the test for an in-
mally affected by OUTLIERS (extreme termediate group of students.
values). See also INTERQUARTILE RANGE.
i n t e r i t e m correlation the degree of interpretive phenomenological
correlation between each test item in a analysis (IPA) a QUALITATIVE approach
set, used as a measure of the INTERNAL in which a researcher explores how a
CONSISTENCY of.a test and thus of its RE-
participant understands his or her per-
LIABILITY. For example, a participant's sonal and social worlds and gives mean-
responses to each item on a 10-item test ing to particular experiences, events, and
could be correlated with his or her re- states. For example, the stiategy might be
sponses to every other item on the test used to understand how a person comes
to determine the extent to which the to terms with the death of a spouse. IPA
items reflect the same construct. uses STRUCTURED INTERVIEWS to gather
verbal and nonverbal information, which
intermediary variable see MEDIA- is then analyzed to uncover and describe
TOR. underlying themes.
internal consistency the degree of in- interpretive statistics see INFEREN-
terrelationship or homogeneity among TIAL STATISTICS.
the items on a test, such that they are
consistent with one another and mea- interquartile range (IQR) an index
suring the same thing. Internal consis- of the DISPERSION within a data set: the
tency is an index of the RELIABILITY of a difference between the 75th and 25th
test. Also called intemal consistency PERCENTILE scores (also known as the
reliability; intemal rehability; scale upper and lower HINGES) within a distri-

179
interrater reliability

bution. Also called H spread; mid- in a study of asthmatic symptoms

spread. among mine workers, the onset of a par-
ticular symptom in an individual might
interrater reliability the extent to only be known to have occurred be-
which independent evaluators produce tween two health examinations. See
similar ratings in judging the same abili- CENSORED DATA.
ties or characteristics in the same target
person or object. It often is expressed as i n t e r v a l data numerical values that in-
a CORRELATION COEFFICIENT. If Consis- dicate magnitude but lack a "natural"
tency is high, a researcher can be confi- zero point. Interval data represent exact
dent that similarly trained individuals quantities of the variables under consid-
would produce similar scores on targets eration, and when arranged consecu-
of the same kind. If consistency is low, tively they have equal differences among
there is little confidence the obtained adjacent values (regardless of the spe-
scores could be reproduced with a differ- ciflc values selected) that correspond to
ent set of raters. Also called intercoder genuine differences between the physi-
reliability; interjudge reliability; cal quantities being measured. Tempera-
interobserver reliability; inter- ture is an example of interval data: The
scorer reliability. difference between 50 F and 49 F is the
same as the difference between 40 F and
interrupted time-series design a 39 F, but a temperature of 0 F does not
QUASI-EXPERIMENTAL DESIGN in which indicate that there is no temperature.
the effects of an intervention are evalu- Compare RATIO DATA. See also INTERVAL
ated by comparing outcome measures SCALE.
obtained at several time intervals before
and several time intervals after the inter- i n t e r v a l estimate an estimated range
vention was introduced. Unlike tradi- of likely values for a given population
tional TIME-SERIES DESIGNS, which make PARAMETER. For example, a researcher
use of a continuous predictor variable, might use data from a sample to deter-
an interrupted time-series design uses a mine that the average score on a particu-
categorical predictorthe absence or lar variable in the larger population
presence of an intervention. falls between 20 and 25. Compare POINT
ESTIMATE.
interscorer r e l i a b i l i t y see INTER-
RATER RELIABILITY. i n t e r v a l recording a process in which
a researcher divides a specific observa-
i n t e r t r i a l i n t e r v a l (ITl) the time tion period into equally sized smaller
between successive presentations of time periods (intervals) and indicates
the stimulus in a series of experimental whether or not a target behavior oc-
trials. curred during each interval. Either PAR-
TIAL-INTERVAL RECORDING or WHOLE-
i n t e r v a l n. a range of scores or values,
INTERVAL RECORDING may be used. See
such as a CLASS INTERVAL or CONFI-
also TIME SAMPLING.
DENCE INTERVAL.
i n t e r v a l s a m p l i n g a form of RANDOM
interval-censored observation an
SAMPLING in which participants at uni-
event of interest whose exact value or
formly separated points are selected for
timing is not known because it occurred
study and the starting point for selection
during an interval of time when record-
is arbitrary. An example is choosing
ing did not take place. Interval-censored
every 10th name from a list of candi-
observations are common in clinical tri-
als, during which patients are seen only dates.
at prescheduled times rather than being i n t e r v a l scale a scale marked in equal
continuously monitored. For example. intervals so that the difference between

180
 intraclass contingency table

any two consecutive values on the scale interview n. a directed conversation in

is equivalent regardless of the two values which a researcher, therapist, clinician,
selected. Interval scales lack a true, employer, or the like (the interviewer)
meaningful zero point, which is what intends to elicit specific information
distinguishes them from RATIO SCALES. from an individual (the interviewee) for
For example, Fahrenheit temperature purposes of research, diagnosis, treat-
uses an interval scale: The difference be- ment, or employment. Conducted face
tween 50 F and 49 F is the same as the to face or by telephone, interviews may
difference between 40 F and 39 F, but a be either standardized, including set
temperature of 0 F does not indicate questions, or open ended, varying with
that there is no temperature. See also IN- material introduced in responses by the
TERVAL DATA. interviewee. Their RELIABILITY is of par-
ticular concern, and interviewers must
i n t e r v a l variable a variable that is be careful to minimize or eliminate per-
measured using an INTERVAL SCALE. Be- sonal judgment and biases in evaluating
cause values on such a scale are equally responses. See STRUCTURED INTERVIEW;
spaced, the differences between values UNSTRUCTURED INTERVIEW.
of an interval variable are meaningful.
Compare ORDINAL VARIABLE. interviewer effect the influence of
the characteristics of an interviewer
i n t e r v a l w i d t h the difference between upon the responses provided by an in-
the upper REAL LIMIT and the lower real terviewee. The interviewer's age, gender,
limit of the CLASS INTERVALS used in a and level of experience may affect the
FREQUENCY DISTRIBUTION. For example, manner in which the interviewee re-
one might construct a distribution sponds, as may his or her general de-
showing the number of students with meanor and nonverbal cues. For example,
test scores between 1 and 10 points, 11 a person might discuss sensitive topics,
and 20 points, 21 and 30 points, and so such as sexual or drinking behavior,
forth. The corresponding real limits of more openly and truthfully with an in-
each class interval are 0.5 and 10.5, 10.5 terviewer who is of the same gender.
and 20.5, 20.5 and 30.5, and so on. Sub-
tracting the real lower limit from the real intraclass contingency table a table
upper limit (e.g., 10.5 - 0.5) yields an in- that presents the pooledfrequencycounts
terval width of 10. on certain CATEGORICAL VARIABLES for
pairs of entities where there is no mean-
i n t e r v e n i n g variable see MEDIATOR. ingful distinction between the pair.
intervention n. in research design, an Thus, it is an amalgamation of the sepa-
EXPERIMENTAL MANIPULATION. rate CONTINGENCY TABLES for the pair
members individually. For example,
intervention research strategies and suppose a researcher wishes to examine
processes designed to measure the the occurrence of cancer and heart dis-
change in a situation or individual after ease across sibling pairs and genetic
a systematic modification (i.e., diet, twins. Since it is irrelevant which person
therapeutic technique, etc.) has been in each pair is affected by a disease (i.e.,
imposed or to measure the effects of one what matters is simply that the disease is
type of intervention program as com- present in the pair), the researcher might
pared to those of another program. Ex- study the relationship between the dis-
perimentation is the most common type eases by creating an intraclass contin-
of intervention research but CLINICAL gency table that shows the overall
TRIALS and qualitative studies may also number of sibling pairs with each dis-
be used. Also called intervention ease and the overall number of twin
study. pairs with each disease.

181
intraclass correlation

intraclass correlation the degree of will assume or below which it will

homogeneity among unordered mem- fall. Also called percent point func-
bers (people, items, etc.) of a group. For tion.
example, it may reflect the level of agree- inverse f a c t o r analysis see Q-
ment among different judges rating a TECHNIQUE FACTOR ANALYSIS.
specific attribute. It is indexed by
the intraclass correlation coefficient inverse Fisher t r a n s f o r m a t i o n a
(ICC), which ranges from 0 to 1 in value. statistical procedure applied to convert
A larger ICC indicates more homogene- the PROBABILITY DENSITY FUNCTION Of a
ity, and thus that a correspondingly data set to one that will yield a shape
smaller proportion of the total variance that more closely resembles a NORMAL
in a dependent variable is attributable to DISTRIBUTION. See FISHER'S R TO Z
INDIVIDUAL DIFFERENCES Within the TRANSFORMATION. [Sir Ronald Aylmer
group. Compare INTERCLASS CORRELA- Fisher (1890-1962), British statistician
TION. and geneticist]

intraclass variance within-groups inverse f u n c t i o n a mathematical

variance: the variability among entities FUNCTION that reverses the input and
in a group. See WITHIN-GROUPS MEAN output of another function; that is, it
uses the results from the original func-
SQUARE.
tion to calculate the initial values of that
i n t r a i n d i v i d u a l differences the function. For example, if y = x^, then its
variations between two or more traits, inverse function is x = Vy.
behaviors, or characteristics of a single
inverse Gaussian d i s t r i b u t i o n a
person. For example, certain aptitude
distribution having positive SKEWNESS
tests measure a testee's strengths in that is used to model nonnegative data.
mathematical, verbal, and analytic abili- Also called inverse normal distribu-
ties; differences among the three stan- tion; Wald distribution. [Karl
dardized scores represent intraindividual Friedrich Gauss (1777-1855), German
differences. Compare INDIVIDUAL DIF- mathematician]
FERENCES.
inversely p r o p o r t i o n a l see DIRECTLY
invariance n. the property of being un- PROPORTIONAL.
changed by a TRANSFORMATION. For ex-
ample, after adding a constant value to inverse m a t r i x a SQUARE MATRIX that
each of a set of scores on a test, the de- is the RECIPROCAL of another such ma-
grees of difference among the scores trix, as in the following example.
will demonstrate invariance, being the
same for the original and transformed A (inverse)
set. invariant adj.
inventory n. a list of items, often in
11 41
question form, used in describing and
studying behavior, interests, and atti- The product of a square matrix and its
tudes. inverse matrix is an IDENTITY MATRIX.
Inverse matrices are commonly used in
inverse correlation see NEGATIVE FACTOR ANALYSIS and Similar multi-
CORRELATION. variate statistical procedures.
inverse d i s t r i b u t i o n f u n c t i o n an in- inverse n o r m a l d i s t r i b u t i o n see IN-
verted CUMULATIVE DISTRIBUTION FUNC- VERSE GAUSSIAN DISTRIBUTION.
TION: for any given probability value
from 0 to 1, a formula that yields the inverse p r e d i c t i o n see REGRESSION OF
specific value that a random variable X ON Y.

182
 item discriminability

inverse relationship see NEGATIVE IRB abbreviation for INSTITUTIONAL RE-

RELATIONSHIP. VIEW BOARD.
inverse sine t r a n s f o r m a t i o n see IRT abbreviation for ITEM RESPONSE THE-
ARC SINE TRANSFORMATION. ORY.

inverse t r a n s f o r m a t i o n see RECIP- isosensitivity f u n c t i o n see RECEIVER-

ROCAL TRANSFORMATION. OPERATING CHARACTERISTIC CURVE.

inverted U-shaped d i s t r i b u t i o n see isotonic regression a type of LEAST

U-SHAPED DISTRIBUTION. SQUARES REGRESSION in whlch the vari-
ables in the regression equation are as-
invertible m a t r i x see NONSINGULAR sumed to be MONOTONIC (i.e., the
MATRIX. variables all trend in the same direction
IPA abbreviation for INTERPRETIVE with no NEGATIVE RELATIONSHIPS be-
PHENOMENOLOGICAL ANALYSIS. tween variables).

ipsative method a type of research i t e m analysis a set of procedures used

procedure in which a person's responses to evaluate the statistical merits of indi-
(e.g., scores) are compared only to other vidual items comprising a psychological
responses of that person rather than to measure or test. These procedures may
the responses of other people. It is thus be used to select items for a test from a
an IDIOGRAPHIC approach rather than a larger pool of initial items or to evaluate
NORMATIVE OnC. items on an estabhshed test. A variety of
different statistics may be computed in
ipsative scale a scale in which the item analyses, including DIFFICULTY
points distributed to the various differ- VALUES, INDEXES OF DISCRIMINATION,
ent items must sum to a specific total. In INTERITEM CORRELATIONS, and ITEM
such a scale, all participants will have DISCRIMINABILITIES.
the same total score but the distribution
of the points among the various items item characteristic curve (ICC) a
will differ for each individual. For exam- plot of the probability that a test item is
ple, a supervisor using an ipsative scale answered correctly against the ex-
to indicate an employee's strength in aminee's underlying ability on the trait
different areas initially might assign 20 being measured, as in the example
points for communication, 30 for timeli- overleaf.
ness, and 50 for work quality but a few The item characteristic curve is the
months later assign 30 points for com- basic building block of ITEM RESPONSE
munication, 30 for timeliness, and 40 THEORY: The curve is bounded between
for quality of work. The total number of 0 and 1, is monotonically increasing,
points distributed in each case, however, and is commonly assumed to take the
is the same (100). Ipsative scales also shape of a LOGISTIC FUNCTION. Each
may involve ranks: Respondents use the item in a test has its own item character-
same numbers for ranking but may as- istic curve.
sign them differently. For example, two i t e m d i f f i c u l t y see DIFFICULTY VALUE.
individuals indicating their preferences
for 10 different restaurants will both use i t e m d i s c r i m i n a b i l i t y a statistical
the ranks 1 through 10 but the restau- measure of how well an item on a test
rant chosen as #1 will not be the same differentiates among subgroups of test
for each person, the restaurant chosen as takers, typically those individuals who
#2 will not be the same, and so on. possess a high degree of some ability ver-
sus those who possess a low degree of the
IQR abbreviation for INTERQUARTILE ability. Item discriminabilities typically
RANGE. are given as CORRELATION COEFFICIENTS

183
item distractor

ranging from -1 to -i-l in value, with the Item response theory models differ in
latter indicating a perfect discriminator. terms of the number of parameters in-
INTERITEM CORRELATIONS also may be cluded in the model. For example, the
used in determining item discrim- RASCH MODEL Is based on the single pa-
inability. See ITEM ANALYSIS. rameter of item difficulty, whereas other
i t e m distractor one of the alternatives models additionally examine ITEM
to the correct answer provided for each DISCRIMINABILITY and the chances of
item on a MULTIPLE-CHOICE TEST. Well- successful guessing. Compare CLASSICAL
chosen distractors can differentiate TEST THEORY.
among the different classes of test takers: i t e m scaling the assignment of a test
Those who know the material should item to a scale position on some dimen-
select the correct alternative despite sion, often that of item difficulty.
the presence of plausible distractors, i t e m selection the selection of test
whereas those with limited knowledge items for inclusion in a test battery based
are likely to guess randomly from upon the final psychometric properties
among the available distractors. of the test battery, the ITEM RESPONSE
i t e m homogeneity see INTERNAL CON- THEORY parameters of the individual
SISTENCY. items, and the clarity and fairness of the
individual items.
item nonresponse see NONRESPONSE.
i t e m - t o t a l correlation the degree to
i t e m r e l i a b i l i t y see INTERNAL CONSIS- which scores on a single item from a test
TENCY.
are related to the total score for the test
i t e m response theory (IRT) a psy- as a whole, used in determining the IN-
chometric theory of measurement based TERNAL CONSISTENCY of the test. Low
on the concept that the probability that item-total correlations (i.e., less than
an item will be correctly answered is a 0.2) indicate poor associations of an
function of an underlying trait or ability item with others on the test, such that it
that is not directly observable, that is, a is not measuring the same construct and
latent trait (see LATENT TRAIT THEORY). should be discarded.

1.0

0.9

10 0 . 8 -
c
o
a. 0.7-
in
0)
te. 0.6-

u
u 0.5-
i-

o 0.4
o
o 0.3-
.>
O
k. 0.2-
0.
0.1 _

"T" 1 \ T
-3.5 -2.5 -1.5 0.5 0.5 2.5

Ability
item characteristic curve
 IV

item v a l i d i t y the extent to which an iterative a l g o r i t h m a mathematical

individual item in a test or experiment process of successive approximation in
measures what it purports to measure. which a sequence of increasingly accu-
rate procedures for solving a class of
item w e i g h t i n g a numerical value as-
problems is generated. The outcome of
signed to a test item that expresses its
new ALGORITHMS is compared to that of
percentage of the total score of the test.
the prior algorithms from which they
For example, an essay question may be
are derived until certain criteria have
assigned a value of 40, representing 40
been satisfied or a certain condition ob-
out of 100 possible points. See WEIGHTING.
tained.
iterated bootstrapping a BOOT-
iterative p r o p o r t i o n a l f i t t i n g a
STRAPPING procedure in which samples
mathematical procedure in which a se-
are randomly drawn (with replacement)
ries of calculations is repeatedly applied
from an observed data set, values for dif-
to adjust the values in one data set so
ferent population characteristics are esti-
that they gradually conform to the con-
mated, and the estimates averaged
straints in another associated data set.
across the set of samples. Then, the sam-
For example, a researcher may have both
ples created during this initial procedure
CELL FREQUENCIES and MARGINAL FRE-
are themselves sampled and their PA-
QUENCIES for sales of a product in the
RAMETERS estimated and averaged. The
previous year but only estimates of mar-
resampling and recalculating process
ginal frequencies for the current year. He
continues until a stable parameter esti-
or she might use iterative proportional
mate is obtained. Iterated bootstrapping
fitting to adjust the cell frequencies from
usually is less prone to error than using a
the previous year to fit the estimated
single set of bootstrapping samples.
marginals of the current year.
iteration n. the repetition of a certain
ITI abbreviation for INTERTRIAL INTER-
computational step until further repeti-
VAL.
tion no longer changes the outcome or
until the repetition meets some other IV abbreviation for INDEPENDENT VARI-
predefined criterion. ABLE.

185
 Ji
j a c k k n i f e n. a statistical procedure used
to estimate the variability of a PARAME-
TER associated with a set of data, such as
the STANDARD ERROR or CONFIDENCE
INTERVAL. It is particularly appropriate
when the VARIANCE or underlying distri-
bution is not known. A number of sam-
ples are obtained from the original data
by eliminating one or more observations
at a time, the parameter in question is
calculated for each sample, and the indi-
vidual parameters are combined to pro-
vide an estimate of the overall parameter
for the entire data set. A jackknife is sim-
ilar to BOOTSTRAPPING, except that
bootstrapping involves replacing obser-
vations after they have been sampled,
such that after each observation is cho-
sen it is reinserted into the data for possi-
ble selection again.
James-Stein estimator a formula used
to approximate a population PARAMETER
from sample data. It is an improved ver-
sion of the ordinary LEAST SQUARES ESTI-
MATOR, providing a smaller and more
accurate measure of the MEAN SQUARED
ERROR between actual and predicted
scores. [Willard James, U.S. mathemati-
cian; Charies M . Stein (1920- ), U.S.
statistician]
J curve a pattern on a graph that resem-
bles the letter J: There is an initial drop
in value for a short period of time fol-
lowed by a continuous increase. For ex-
ample, a high achieving person training
for a difficult job might show a J curve
for performance (as shown in the dia-
gram below). His or her performance ini-
tially might drop below that of other
employees while he or she learns diffi-
cult skills but rapidly increase to excel-
lence after fully attaining the required
job skills. Also called J-shaped flistri-
bution.
Jeffreys's p r i o r in BAYESIAN statistics,
a method used to estimate the PROBA-
BILITY DISTRIBUTION for a PARAMETER of
interest, such as the population mean.
Jeffreys's prior is based on the square
root of the DETERMINANT of FISHER'S IN-
FORMATION, which provides an approxi-
mation of the degree of variability that

4 (high)

c 3 -
ra
ce.
0) 2 -
u
c
n
E
4) 1 (iow) -
0.

I r
2 3

Month
/ curve
186
 joint probability

Not Jittered Jittered

og oO O ^<D O
- 08

oocoooooDCO oo ooccoococco

T 1 1 r T 1 1 1 r
Variable x Variable x

jittering

could be expected for the parameter. See rate the points and make it easier to dis-
also PRIOR DISTRIBUTION. [Sir Harold cern a pattern or relationship.
Jeffreys (1891-1989), British mathema-
tician and geophysicist] joint d i s t r i b u t i o n the pattern of val-
ues obtained when estimating the prob-
Jensen's i n e q u a l i t y a mathematical ability of occurrence of two or more
FUNCTION stating that a convex TRANS- RANDOM VARIABLES. For example, the
FORMATION of a mean value will be probability values for drawing a heart
equal to or smaUer than the mean of a and a jack in a 'deck of cards would form
set of convex transformed values. In the following joint distribution: (13/52)
other words, a transformation that cre- X (48/52) to obtain a heart but not a jack;
ates a set of smaller values that follow a (39/52) X (48/52) to obtain a nonheart
convex pattern will yield a smaller over- and a nonjack; (39/52) x (4/52) to obtain
all value than will the average of a set of a nonheart and a jack; and (13/52) x
values that have already undergone a (4/52) to obtain a heart that is also a
convex transformation. [Johan Jensen jack. This is illustrated below.
(1859-1925), Danish mathematician] A joint distribution for two variables is
referred to more specifically as a BI-
j i t t e r i n g n. deliberately adding a small VARIATE DISTRIBUTION, whereas a joint
amount of random noise to each obser- distribution for more than two variables
vation in a SCATTERPLOT where vari- is called a MULTIVARIATE DISTRIBUTION.
ables take on relatively few values,
resulting in identical data points that joint p r o b a b i l i t y the chance that
stack on top of each other and thus are two events will occur simultaneously.
hard to interpret. Consider the generic For example, the joint probability of
example above. Primarily a data visual- drawing a heart that is a jack from a deck
ization technique, jittering helps sepa- of 52 cards is equal to (13/52) x (4/52), or

Card suit Jack Nonjack

Heart 1/52 = .02 = 2% 12/52 = .23 =23%
Nonheart 3/52 = .06 = 6% 36/52 = .69 = 69%

joint distribution
187
joint probability density function

1/52. See also JOINT DISTRIBUTION; of personal opinion as to what is repre-

JOINT P R O B A B I L I T Y DENSITY FUNCTION. sentative of the population under study.
For example, judgment sampling could
joint probability density function involve choosing a number of well-
the statistical formula for calculating the informed people from different social
likelihood of the simultaneous occur- groups and asking them who they be-
rence of particular values on two or Ueve wiU be the next U.S. president. The
more continuous variables (e.g., age and outcome from such a study may well not
income). The joint probability density be consistent with the result of asking a
function thus underlies a JOINT DISTRI- RANDOM SAMPLE from the population.
BUTION where the variables have a con-
This means that although judgment
tinuous, rather than a discrete or sampling may be useful in a PILOT
categorical, form. See also PROBABILITY STUDY, it generally does not result in
DENSITY FUNCTION.
warranted inference to any population
Jonckheere k-sample test a NON- of interest. See DELIBERATE SAMPLING.
PARAMETRIC method for analyzing just-identified model a model in
whether two or more ordered samples are which the number of population PA-
significantly different. The Jonckheere k- RAMETERS estimated is the same as the
sample test is a MANN-WHITNEY U TEST number of data points in the sample
on multiple samples and is similar to data set (i.e., the DEGREES OF FREEDOM
a KRUSKAL-WALLIS ONE-WAY ANALYSIS of the model are equal to zero). In
OF VARIANCE. Also Called Jonckheere- nonstatistical terms, a just-identified
Terpstra k-sample test; Jonckheere- model is like having funding (e.g., data
Terpstra test. [A. R. Jonckheere (1920- points) to pay for a major intervention
2005), French-bom British statistician (e.g., to estimate parameters) without
and psychologist; T. J. Terpstra, Dutch having any left (e.g., degrees of freedom)
mathematician] to evaluate whether the intervention
was successful (e.g., whether the model
J-shaped d i s t r i b u t i o n see j CURVE.
fit or not). Also called exact identifica-
judgment s a m p l i n g selecting a group tion. Compare OVERIDENTIFIED MODEL;
from which to gather data on the basis UNDERIDENTIFIED MODEL.

188
 Kk
k 1. symbol for COEFFICIENT OF ALIEN- who are expected to survive a specific ill-
ATION. 2. symbol for the number of in- ness over varying numbers of years; the
dividuals, groups, or other units in a graph of values from a Kaplan-Meier es-
statistical analysis. timator presumably would show a de-
cline over time, with fewer and fewer
Kaiser-Meyer-Olkin test of sam- surviving patients. Also called product-
p l i n g adequacy (KMO test) a measure l i m i t estimator. [Edward L. Kaplan,
of whether the PARTIAL CORRELATIONS
U.S. mathematician; Paul Meier (1924-
among a set of variables are sufficient to 2011), U.S. biostatistician]
conduct a FACTOR ANALYSIS. KMO test
values may range from 0 to 1, with val- k a p p a n. see COHEN'S KAPPA.
ues of .50 or higher indicating there is a
sufficient degree of relationship among K e n d a l l robust line-fit method see
the variables that they reasonably could THEiL-SEN ESTIMATOR. [Sir Mauticc
be analyzed to reveal one or more under- Kendall (1907-1983), British statistician]
lying factors. [Henry F. Kaiser (1927- Kendall's coefficient of concor-
1992), U.S. psychometrician and statisti- dance see COEFFICIENT OF CONCOR-
cian; Michael Meyer, U.S. statistician; DANCE. [Sir Maurice Kendall]
Ingram Olkin (1924- ), U.S. stafistician]
Kendall's tau (symbol: x) a nonpara-
K a l m a n f i l t e r a statistical method used metric measure of the degree of associa-
in TIME-SERIES ANALYSES and other lon- tion between two ordinal variables (i.e.,
gitudinal procedures that combines esti- rank-ordered data). For example, a re-
mates from several sources (e.g., actual searcher could calculate Kendall's tau to
measures, predicted values) to obtain a assess how much relationship there is
more precise indication of a set of values between the rankings of students' per-
over time. It is similar to a HIDDEN formance provided by two observers
MARKOV MODEL but based on CONTINU- (e.g., a teacher and a teaching assistant).
OUS rather than DISCRETE measures. For Also called Kendall's rank correla-
example, a Kalman filter could be ap- tion coefficient. [Sir Maurice Kendall]
plied to assess the amount of drug use
across time for one or more individuals, Kendall's W see COEFFICIENT OF CON-
whereas a hidden Markov model would CORDANCE. [Sir Maurice Kendall]
be used to assess the distinct stages of
kernel density estimator a method
substance use across time. [Rudolf E.
for approximating a smooth, bell-shaped
Kalman (1930- ), Hungarian-born U.S.
distribution from the choppy representa-
electrical engineer and mathematician]
tion in a bar graph or histogram. For ex-
Kaplan-Meier estimator a statistical ample, one might apply a kernel density
method used in SURVIVAL ANALYSIS with estimator to the possible values when
CENSORED DATA for estimating the per- throwing two dice (e.g., one frequency
centage of individuals (or other entities) of 2: a 1 and a 1; two frequencies of 3: a 1
that will remain after a set amount of and a 2 or a 2 and a 1; three frequencies
time. For example, the Kaplan-Meier es- of 4: a 3 and a 1, a 1 and a 3, and a 2 and a
timator could be used in health studies 2; etc.), as they yield a boxy histogram
to estimate the percentage of patients that peaks in the middle at 7 and has low

189
kernel function

frequencies of one to the far left and far happy individuals, and high scores for
right for 2 and 12, respectively. Also individuals with depressed affect who
called kernel estimator; Parzen win- demonstrated little activity or motiva-
dow estimator. tion. It is a form of CONSTRUCT VALID-
ITY.
k e r n e l f u n c t i o n a NONPARAMETRIC
method used to estimate the distribu- Kolmogorov-Smirnov goodness-
tion for a RANDOM VARIABLE. For exam- o f - f i t test a NONPARAMETRIC method
ple, a kernel function may estimate a for comparing the distribution from
rectangular-shaped UNIFORM DISTRIBU- a sample data set to an expected distri-
TION, indicating approximately the bution or the known distribution of a
same number of low, medium, and high given population. If the test yields a dis-
scores. Such a kernel function would be crepancy (D) larger than the CRITICAL
in contrast to the bell-shaped pattern of VALUE, then the sample data are consid-
a NORMAL DISTRIBUTION, in whlch most ered to be significantly different
scores pile up in the middle and gradu- from the reference distribution. Also
ally decline on either side. called Kolmogorov-Smirnov D test;
Kolmogorov-Smirnov one-sample
kernel regression a NONPARAMETRIC test. Compare KOLMOGOROV-SMIRNOV
method of forming a smooth pattern TWO-SAMPLE TEST. [Andrei Nikolaevich
when estimating the relationship be- Kolmogorov (1903-1987) and Nikolai
tween two variables. For example, kernel Vasilevich Smirnov (1900-1966), Soviet
regression could be used to find a rela- mathematicians]
tively smooth (i.e., not choppy) pattern
in a SCATTERPLOT depicting the age and K o l m o g o r o v - S m i r n o v two-sample
height of each chOd in a classroom. Also test a NONPARAMETRIC method for
called kernel regression smoothing; comparing the distributions from two
kernel smoother. samples to see if they are similar. A signif-
icant result from a Kolmogorov-Smirnov
k-means clustering a type of CLUSTER two-sample test indicates that the two
ANALYSIS that separates a large set of en- samples are derived from different distri-
tities into a smaller number of k subsets butions. In contrast, the KOLMOGOROV-
(i.e., clusters) by assigning each entity to
SMIRNOV GOODNESS-OF-FIT TEST IS USed
the subset closest in mean value. For ex-
when only one sample is available
ample,fc-meansclustering might take a
but the larger population distribution is
group of individuals and separate them known. [Andrei Kolmogorov and Nikolai
into three subsets or clusters depending Smirnov]
on the closeness of their IQ scores to the
cluster means of, say, 70, 100, and 130, K-R 20 abbreviation for Kuder-Richard-
for low, medium, and high IQ, respec- son formula 20. See KUDER-RICHARD-
tively. soN FORMULAS.

K M O test abbreviation for KAISER- K-R 21 abbreviation for Kuder-Richard-

MEYER-OLKIN TEST OF SAMPLING ADE- son formula 21. See KUDER-RICHARD-
QUACY. SON FORMULAS.
known-group v a l i d i t y the demon- k r i g i n g n. a statistical method that uses
stration that an instrument can distin- INTERPOLATION to estimate an un-
guish between several groups known to known value between two known values
vary on the dimension being measured. whUe minimizing prediction error. For
For example, a newly created measure of example, a simple form of kriging could
depression would be said to have estimate that the unknown point half-
known-group validity if it resulted in way between two known points of 1 and
low scores for high-functioning and 9 would be a value of 5. [Daniel
 kurtosis

Gerhardus Krige (1919- ), South Afri- equal. It is an extension of the M A N N -

can geostatistician] WHITNEY u TEST, which is conducted
Kronecker product o f matrices the when there are only two independent
SQUARE MATRIX that results from multi- samples. Also called Kruskal-Wallis
plying each of the values in one matrix test. [WiUiam Henry Kruskal (1919-
by each of the values in another matrix: 2005) and WUson Allen Wallis (1912-
It consists of four blocks corresponding 1998), U.S. statisticians]
to each of the row values in the first ma- Kuder-Richardson f o r m u l a s two
trix by each of the column values in the methods, Kuder-Richardson formula 20
second matrix. For example, imagine a (K-R 20) and Kuder-Richardson formula
matrix A with two rows of values 1, 3 21 (K-R 21), for assessing the INTERNAL
and 2, 4, respectively, and a matrix B CONSISTENCY reliability of a test or
with two rows of values 6, 8 and 5, 7, re- subtest made up of binary variables (i.e.,
spectively. DICHOTOMOUS ITEMS). The K-R 21 re-
quires that aU of the dichotomous items
Kronecker product be equally difficult, whereas the K-R 20
of AxB does not. For example, a teacher could
A B [ 6 8 18 24 use the K-R 21 formula to calculate inter-
r l 31 [6 81 5 7 15 21 nal consistency reliability for a test of 20
^2 4J 15 7I 12 16 24 32 equaUy difficult items, but would use
.10 14 20 28 the K-R 20 formula for test items differ-
ing in difficulty level. [Frederic Kuder
(1903-2000) and Marion Webster Rich-
As shown, the Kronecker product of ma-
ardson (1896-1965), U.S. psychologists]
trices A and B is a 4 x 4 matrix with val-
ues of 6, 8, 18, and 24 in the first row; 5, K u l l b a c k - L e i b l e r i n f o r m a t i o n the
7, 15, and 21 in the second row; 12, 16, amount of extra data that is needed to
24, and 32 in the third row; and 10, 14, approximate an expected or theoretical
20, and 28 in the fourth row. [Leopold distribution from an observed distribu-
Kronecker (1823-1891), German mathe- tion. The Kullback-Leibler information
matician] is sometimes referred to as the Kullback-
Leibler deviance to indicate the discrep-
K r u s k a l stress a measure of the distance ancy or deviance between the expected
between predicted and actual dissimilari- and actual disfributions. [Solomon Kull-
ties in a MULTIDIMENSIONAL SCALING back (1907-1994) and Richard Leibler
analysis of the perceived similarities be- (1914-2003), U.S. mathematicians]
tween stimuli. See also DISSIMILARITY
COEFFICIENT; SIMILARITY COEFFICIENT. kurtosis n. the fourth CENTRAL MOMENT
[Joseph B. Kruskal (1928-2010), U.S. of a PROBABILITY DISTRIBUTION. It iS 3
statistician and computer scientist] Statistical description of the degree of
peakedness of that distribution. For ex-
K r u s k a l - W a l l i s one-way analysis ample, the ages of a sample of college
o f variance a NONPARAMETRIC TEST freshmen would probably show kurtosis,
for assessing whether the MEDIANS of having a high peak at age 18. See LEPTO-
multiple samples of ranked data are KURTIC; MESOKURTIC; PLATYKURTIC

191
 Ll
laboratory analogue see ANALOGUE value at a given point in time is related
STUDY. to its value at an earlier time point.
laboratory research scientific study lagged independent variable in
conducted in a laboratory or other such TIME-SERIES ANALYSIS, an explanatory
workplace, where the investigator has variable exhibiting AUTOCORRELATION
some degree of direct control over the in that its value at a given point in time
environment and can manipulate the is related to its value at an earlier time
INDEPENDENT VARIABLES. Although lab- point.
oratory research generally has greater IN-
TERNAL VALIDITY than FIELD RESEARCH, Lagrange m u l t i p l i e r test in REGRES-
it tends to be less generaUzable to the SION ANALYSIS, a procedure used to esti-
real world (i.e., has less EXTERNAL VALID- mate the improvement in the fit of a
ITY). See EXPERIMENTAL RESEARCH. model to observed data that is gained by
the inclusion of additional variables. For
lack o f f i t the degree to which the val- example, a researcher may hypothesize
ues predicted by a modeltypically, that high-school grade point average
one developed in REGRESSION ANALYSIS, and college admission test scores have
STRUCTURAL EQUATION MODELING, Or more influence on grade point average
CONFIRMATORY FACTOR ANALYSISdi- in college than do gender and ethnicity.
verge from the corresponding empiri- Thus, he or she could develop a model
cal values. A large, significant value from that includes only those two items as
a GOODNESS-OF-FiT TEST indicates a predictor variables and then use a
poorly fitting model. Lagrange multiplier test to determine
whether adding gender and ethnicity re-
l a d d e r i n g n. a knowledge elicitation
sults in a significantly better fitting
technique used in interviews to impose a
model. [Joseph Louis Lagrange (1736-
systematic framework upon questioning
1813), Italian-born French mathemati-
so as to reveal complex themes across
cian]
answers. In laddering, a respondent re-
plies to a series of "why?" probes, thus l a m b d a 1. (symbol: X) see EIGENVALUE.
requiring him or her to expose and ex- 2. (symbol: A) see WILKS'S LAMBDA.
plain choices or preferences and justify
behavior in terms of goals, values, and Laplace d i s t r i b u t i o n see DOUBLE EX-
personal constructs. Laddering is con- PONENTIAL DISTRIBUTION. [Pierre Simon
cerned with establishing links between Laplace (1749-1827), French mathema-
concepts elicited from the participant tician]
(e.g., attitudes and beliefs associated
with a particular consumer product) and large-N design a research approach in
provides greater scope for probing sa- which groups comprising numerous in-
lient issues while optimizing the often dividuals are studied. Generally, the
limited time available with respondents. larger the number of observations or
participants in a research study, the
lagged dependent variable in TIME- greater the statistical POWER in hypothe-
SERIES ANALYSIS, an outcome variable sis testing during data analysis. Compare
exhibiting AUTOCORRELATION in that its SMALL-N DESIGN.

192

l a r g e - n u m b e r s l a w see L A W OF L A R G E
NUMBERS.
large-sample method see ASYMP-
TOTIC METHOD.
largest characteristic root see
GREATEST CHARACTERISTIC ROOT.
largest-root test see GREATEST-
CHARACTERISTIC-ROOT TEST.
last observation carried f o r w a r d
(LOCF) a strategy for handling missing
data in L O N G I T U D I N A L DESIGNS:
value immediately prior is substituted
for any missing data points and the
completed data set analyzed as usual. Al-
though this approach preserves the sam-
ple size, it may involve unwarranted
assumptions about the missing data,
which in turn may bias study findings.
latent class analysis (LCA) a method
for finding discrete subgroups of related
cases (latent classes) from multivariate
CATEGORICAL DATA. For example, a
health researcher might use latent class
analysis to determine into which one of
10 diagnostic categories to place each in-
dividual in a group on the basis of the
presence or absence of several distinct
symptoms. The procedure is analogous
to CLUSTER ANALYSIS, in that from a
given sample of cases measured on sev-
eral variables, the researcher attempts
to identify a smaller number of underly-
ing dimensions on which to classify
them. The use of a categorical LATENT
VARIABLE, however, is what distinguishes
latent class analysis from more tradi-
tional clustering approaches such as
FACTOR ANALYSIS, which involve con-
tinuous latent variables. Also caUed
latent class modeling.
l a t e n t c o n s t r u c t see L A T E N T V A R I A B L E .
l a t e n t c u r v e m o d e l i n g see LATENT
G R O W T H C U R V E ANALYSIS.
l a t e n t f a c t o r see L A T E N T VARIABLE.
l a t e n t g r o w t h c u r v e a n a l y s i s a type
of G R O W T H C U R V E ANALYSIS in which a
researcher attempts to estimate a PA-
latent variable
RAMETER (or set of parameters) underly-
ing the observed rate of change in a di-
rectly measured variable. Also called
latent curve modeling; latent
growth curve modeling.
l a t e n t r o o t see EIGENVALUE.
l a t e n t s t r u c t u r e a n a l y s i s any of vari-
ous statistical procedures that attempt
to explain relationships among ob-
served variables in terms of underlying
LATENT VARIABLES, whether continuous
the or categorical. Examples include FACTOR
ANALYSIS, L A T E N T CLASS ANALYSIS, and
STRUCTURAL EQUATION MODELING. Also
called latent structure modeling.
l a t e n t t r a i t t h e o r y a general psy-
chometric theory contending that ob-
served traits, such as intelligence, are
reflections of more basic unobservable
traits (i.e., latent traits). Several quantita-
tive models (e.g., I T E M RESPONSE T H E -
ORY, FACTOR ANALYSIS) have been
developed to allow for the identification
and estimation of these latent traits
from manifest observations,
l a t e n t t r a n s i t i o n a n a l y s i s (LTA) an
extension of L A T E N T CLASS ANALYSIS
that is used to study changes over time
in the unobservable categorical dimen-
sions that underlie measured variables.
For example, an educational researcher
might use LTA to model how the inter-
ests (majors) of a class of college students
change as they progress from their fresh-
man to senior years using data about the
classes each individual has taken. Such
an analysis would yield estimates of the
probability of pursuing a particular
major at a future time given a person's
chosen major at a prior time.
l a t e n t v a r i a b l e a theoretical entity or
construct that is used to explain one
or more MANIFEST VARIABLES. Latent
variables cannot be directly observed or
measured but rather are approximated
through various measures presumed
to assess part of the given construct.
For example, suppose a researcher is in-
terested in student conscientiousness.
193
latent vector

Since conscientiousness is a concept EFFECTS they do not control for CARRY-

that cannot be directly evaluated, he or OVER EFFECTS or PRACTICE EFFECTS. See
she might develop a survey containing also BALANCED LATIN SQUARE; GRAECO-
items pertaining to behavior indicative LATIN SQUARE.
of conscientiousness, such as consis-
law n. 1. a formal statement describing a
tently attending classes, turning in as-
regularity (e.g., of nature) to which no
signments on time, engaging in school-
exceptions are known or anticipated. 2.
sponsored fundraising activities, and so
in science, mathematics, philosophy,
forth. Participants' responses could then
and the social sciences, a theory that is
be analyzed to identify patterns of inter-
widely accepted as correct and that has
relationships from which the values of
no significant rivals in accounting for
the latent variable of conscientiousness
the relationships of interest or the facts
are inferred. Also called latent con-
within its domain.
struct; latent factor; synthetic
variable. See also FACTOR ANALYSIS; law of large numbers a mathemati-
STRUCTURAL EQUATION MODELING. cal principle indicating that as the sam-
latent vector see EIGENVECTOR. ple size increases, the theoretical
expectations of its statistical properties
L a t i n square a type of WITHIN-SUBJECTS will be more and more closely realized.
DESIGN in which treatments, denoted by For example, as the number of replica-
Latin letters, are administered in se- tions of an experiment increases, the av-
quences that are systematically varied erage (MEAN) of the observed results will
such that each treatment occurs equally approach the true average (theoretical
often in each position of the sequence probability in the population) with in-
(first, second, third, etc.). The number of creasing accuracy. Or, as the number of
treatments administered must be the trials of a random process increases, the
same as the number of groups or indi- difference between the expected and ac-
vidual participants receiving them. For tual frequencies of a particular outcome
example, consider the following possi- will decrease to zero. According to the
ble Latin square for an experimental de- strong law of large numbers, the aver-
sign involving four treatments (A, B, C, age of the results obtained from a large
and D) and four people: number of trials will converge upon the
theoretically expected value, becoming
Treatment Order closer as more trials are performed. In
1 2 3 4 distinction, the weak law of large num-
bers states that the average is likely to
Participant 1 A B, C D converge upon the expected value as
Participant 2 B C D A more trials are conducted. Thus, the es-
sential difference between the two laws
Participants C D A B is that the former insists on convergence
Participant 4 D A B C whereas the latter expects convergence.
Also called averages law; Bernoulli's
theorem; large-numbers law; law
As shown, each treatment appears ex- of averages.
actly once in each order position: One
person receives treatment A, then B, l a w o f parsimony the principle that
then C, and then D; a second person re- the simplest explanation of an event or
ceives them in sequence B, C, D, and A; a observation is the preferred explanation.
third person in sequence C, D, A, and B; Simplicity is understood in various ways,
and a fourth person in sequence D, A, B, including the requirement that an ex-
and C. It is important to note that al- planation should (a) make the small-
though Latin squares control for ORDER est number of unsupported assump-

194
 Levene test for equality of variance

tions, (b) postulate the existence of the l e f t censoring the situation in which a
fewest entities, and (c) invoke the fewest researcher cannot determine the precise
unobservable constructs. Also called time at which a target event occurred for
principle of parsimony. See OC- some individuals within a sample be-
CAM'S RAZOR. cause those participants experienced the
event prior to the observation period.
l a w of t o t a l p r o b a b i l i t y see TOTAL
For example, left censoring might arise
PROBABILITY LAW.
in a study of alcohol initiation among
LCA abbreviation for LATENT CLASS high school students because a subset of
ANALYSIS. the sample is Ukely to have initiated use
prior to high school entry. See also
leading diagonal see MAIN DIAGONAL. RIGHT CENSORING.
least significant difference (LSD) l e p t o k u r t i c adj. describing a frequency
see FISHER LEAST SIGNIFICANT DIFFER- distribution that is more peaked than the
ENCE TEST. NORMAL DISTRIBUTION.
least squares c r i t e r i o n in REGRES-
SION ANALYSIS and similar statistical
procedures, the principle that one
should estimate the values of PARAME-
TERS in a way that wUl minimize the
squared error of predictions from the
model. That is, one should strive to
build models that minimize the squared
differences between actual scores (ob-
served data) and expected scores (those
predicted by the model). Also called
least squares principle. See LEAST
SQUARES REGRESSION.

least squares estimator in LEAST It has more scores in the center and
SQUARES REGRESSION, the model or pro- fewer at the two exfremes. See also MESO-
cedure that yields the smallest differ- KURTIC; PLATYKURTIC.
ences between observed data points and level n. in an experimental design, the
their values as predicted by the model. quantity, magnitude, or category of the
See also LEAST SQUARES CRITERION. INDEPENDENT VARIABLE (or Variables)
least squares regression a type of RE- being studied. For example, if a re-
GRESSION ANALYSIS in which the re- searcher is assessing the effect of alcohol
searcher strives to develop a least- on cognition, each specific amount of
squared-errors solution or least squares alcohol included in the study is a level
solution to describe the relationship be- (e.g., 0.0 oz, 0.5 oz, 1.0 oz, 1.5 oz).
tween an outcome or DEPENDENT VARI- level of analysis see UNIT OF ANALYSIS.
ABLE and one or more predictors or
INDEPENDENT VARIABLES. That is, one level o f measurement see MEASURE-
develops a model that is the best fit for MENT LEVEL.
the data because it yields the smallest
squared differences between the actual level of significance see SIGNIFI-
observations and their values as pre- CANCE LEVEL.
dicted by the model. Also called least Levene test f o r equality o f vari-
squares estimation; ordinary least ance a method for evaluating whether
squares (OLS) regression. See LEAST the VARIANCE in a set of scores is equiva-
SQUARES CRITERION. lent across two or more groups being

195
leverage point

studied. Equality of variance is required l i k e l i h o o d f u n c t i o n a formula that

in ANALYSIS OF VARIANCE and other sta- yields the probability of obtaining a par-
tistical techniques in order to obtain ticular distribution of values in a sample
valid results. See HOMOGENEITY OF VARI- for each known value of an associated
ANCE. [Howard Levene (1914-2003), population PARAMETER. In other words,
U.S. statistician and geneticist] it indicates how likely a particular popu-
lation is to produce the observed sample
leverage p o i n t in REGRESSION ANALY- data under certain conditions.
SIS, an observation, value, or score that
is extremely different from the average likelihood principle a foundational
for a given predictor or INDEPENDENT tenet of statistical inference stating that
VARIABLE. Leverage points in a data set where there is an unknown population
may distort regression results. PARAMETER, 9, and an observed sample
distribution, x, all relevant information
lie scale a set of items within a psycho- about the population distribution is
logical instrument (particularly a per- contained in the LIKELIHOOD FUNCTION
sonality assessment) used to indicate for X.
whether a respondent has been truthful
in answering. For example, an honest l i k e l i h o o d ratio (LR) the ratio of two
participant would respond similarly to probabilities, a/b, where a is the proba-
the items "I never regret the life deci- bility of obtaining the data observed if a
sions I have made" and "I've never done particular research hypothesis (A; the
anything I later wished I could take back," NULL HYPOTHESIS) Is true and b is the
which are different ways of presenting probability of obtaining the data ob-
the same concept. Conversely, a respon- served when a different hypothesis (B;
dent trying to present himself or herself t h e A L T E R N A T I V E H Y P O T H E S I S ) IS t r u e .
as positively as possible might answer
such related questions inconsistently. likelihood-ratio chi-square (sym-
bol: G^) a test statistic calculated from
life table a table of the expected or ac- LIKELIHOOD RATIOS and uscd for deter-
tual death rates in a specific area or for mining GOODNESS OF FIT in LOGISTIC
those with specific conditions. For ex- REGRESSION and other multivariate pro-
ample, a life table could be constructed cedures and for assessing INDEPENDENCE
to show the life expectancies for individ- in CONTINGENCY TABLES.
uals who are of a certain age, who have a
particular illness, who smoke, and so likelihood-ratio test a procedure in
forth. Life tables typically yield such in- which LIKELIHOOD RATIOS are used to
formation as the number or proportion evaluate the GOODNESS OF FIT of dif-
of people aUve at given points and their ferent statistical models representing re-
probability of continued survival, as well lationships among variables or to deter-
as the number or proportion who died at mine INDEPENDENCE. If the statistic
those same points. They may be ex- obtainedthe LIKELIHOOD-RATIO CHI-
panded to include additional informa- SQUARE or is large and has a small p
tion and often are used in SURVIVAL VALUE (e.g., p < .05), there is evidence
ANALYSIS for any time-to-event data. that the observed data are significantly
Also called mortality table. different from the data predicted by the
model (i.e., the model is a poor fit). Simi-
l i k e l i h o o d n. in statistics, the probabil- larly, when testing two CATEGORICAL
ity of obtaining a particular set of results VARIABLES for independence a large G^
given a set of assumptions about the DIS- indicates the variables are related (i.e.,
TRIBUTION of the phenomena in the are not independent). A Ukelihood-ratio
population and the PARAMETERS of that test is considered more appropriate than
distribution. a CHI-SQUARE TEST when the sample size

196
 linear model

is small. Also called G test; G test. See the requirement that the total of the
also LOG-LIKELIHOOD RATIO. weights used in any comparison equals
zero (e.g., weights of +1 and -1).
l i k e l i h o o d residual see DELETION RE-
SIDUAL. l i n e a r correlation a measure of the
degree of association between two vari-
L i k e r t scale a type of direct attitude
ables that are assumed to have a LINEAR
measure that consists of statements re-
RELATIONSHIP, that is, to be related in
flecting strong positive or negative eval-
such a manner that their values form a
uations of an object. Five-point scales
straight Une when plotted on a graph. It
are common and a neutral middle point
provides an index of the degree of con-
may or may not be included. For exam-
stant change in the value of one variable
ple, an assessment item using a Likert
(y) for each unit change in the value of
scale response format (i.e., a Likert-type
another variable (;*:).
question) might include the following
statement choices: strongly disagree, dis- linear equation an equation that re-
agree, neither disagree nor agree, agree, and lates elements in a summative fashion
strongly agree. The respondent chooses and involves no exponential powers
the option most representative of his or greater than 1. For example, y = a + bx is
her view (e.g., on whether same-sex mar- a linear equation and will form a straight
riages should be permitted) and these line when depicted graphically. Com-
ratings are summed to provide a total at- pare NONLINEAR EQUATION.
titude score for a topic of interest. Also
called Likert summated rating pro- linear function a mathematical proce-
cedure. [Rensis Likert (1903-1981), U.S. dure that transforms one number, quan-
psychologist] tity, or entity to another through
summation, without involving any ex-
linear adj. describing any relationship ponential powers greater than 1. For ex-
between two variables (x and y) that can ample, y = 2>: + 4A: is a linear function. In
be expressed in the form y = a + bx, contrast, a nonlinear function involves
where a and b are numerical constants. interactions (exponents greater than 1)
No COEFFICIENT Can be raised to a power among the quantities, as in y = 2)^ + 4x^.
greater than 1 or be the denominator of
linear i n t e r p o l a t i o n a method of IN-
a fraction. When depicted graphically,
TERPOLATION (i.e., constmcting new
the relationship is a straight line. Com-
data points between two existing data
pare NONLINEAR.
points) that uses a summative formula
linear combination a pattern in without any exponential powers greater
which the value of one variable is de- than 1. Graphically, linear interpolation
rived by multiplying the values of sev- involves connecting two adjacent known
eral other variables in a set by a constant values with a straight line. Compare
quantity and then adding the results. NONLINEAR INTERPOLATION.
For example, the REGRESSION EQUATION
l i n e a r i t y n. a relationship in which one
y = a + b^Xi + biXx + b^x-^ represents a lin-
variable is associated with another via a
ear combination: y is obtained by adding
the constant (a) and the values of the summative formula in which all COEFFI-
CIENTS are to thefirstpower. See LINEAR
three x variables, each of which is first
COMBINATION; LINEAR CORRELATION.
multiplied by a unique weight (b).
l i n e a r i z i n g n. see LINEAR TRANSFORMA-
linear contrast a method of compar-
TION.
ing the mean values on a variable for
two or more groups using CONTRAST linear model any model for empirical
CODING. The mean value for each group data that attempts to relate the values of
is multiplied by a chosen weight with an outcome or dependent variable to the

197
linear regression

explanatory or independent variables FORMATION of a set of raw data using an

through a LINEAR FUNCTION, that is, one equation that involves addition, sub-
that simply sums terms and includes no traction, multiplication, or division
exponents greater than 1. Most com- with a constant. An example is the trans-
monly used statistical techniques (ANAL- formation of X to y by means of the equa-
YSIS OF VARIANCE, REGRESSION ANALYSIS, tion y = a + bx, where a and b are
etc.) can be represented as linear models. numerical constants. A plot of such
Compare NONLINEAR MODEL. transformed data would form a straight
line. Data are often subjected to li-
linear regression a REGRESSION ANAL- nearizing to determine whether a linear
YSIS in which the predictor or independ- model provides a better fitting or more
ent variables (xs) are assumed to be parsimonious explanation of the vari-
related to the criterion or dependent ables. Compare NONLINEAR TRANSFOR-
variable (y) in such a manner that in- MATION.
creases in an x variable result in consis-
tent increases in the y variable. In other line g r a p h a graph in which data points
words, the direction and rate of change representing a series of individual mea-
of one variable is constant with respect surements are shown connected by
to changes in the other variable. The straight line segments. Consider the ex-
analysis yields a REGRESSION EQUATION ample below using data from the U.S.
that includes an INTERCEPT and REGRES- census.
SION COEFFICIENTS for the predictors. Line graphs often are used to show
Compare NONLINEAR REGRESSION. trends over time, such as population
growth. Also called line chart.
linear relationship an association be-
tween two variables that when subjected line o f best f i t see BEST FIT.
to REGRESSION ANALYSIS and plotted on
a graph forms a straight line. In linear re- l i n k f u n c t i o n in the GENERALIZED LIN-
EAR MODEL, a formula used to obtain ex-
lationships the direction and rate of
change in one variable are constant with pected values for an outcome or
respect to changes in the other variable. dependent variable that is nonlinearly
Compare NONLINEAR RELATIONSHIP. related to a set of predictor or independ-
ent variables. Various link functions are
linear scale a scale of measurement that used, each appropriate for different types
presents actual quantities of the charac- of data and hypothesized outcome DIS-
teristic of interest in EQUAL INTERVALS. TRIBUTIONS (e.g., a POISSON DISTRIBU-
For example, a ruler represents a linear TION).
scale in inches. Compare LOGARITHMIC
SCALE. listwise deletion a strategy for dealing
with the problem of missing data in
linear t r a n s f o r m a t i o n the TRANS- which an entire case record is excluded

1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000
Year
line graph
198
 logarithm

from statistical analysis if values are fluence on one another and that any re-
found to be missing for any variable of lationship between them is to be
interest. For example, consider a re- explained by the latent variable underly-
searcher examining the relationships ing them. In ITEM RESPONSE THEORY
among grade point averages in high local independence is the assumption
school, scores on college admissions that an examinee's responses to the vari-
tests, and grade point averages in col- ous items on a test have no influence
lege. If, for whatever reason, a value has upon one another, such that an answer
not been recorded for one of those vari- to one item will not affect the answer to
ables for certain participants, all of the another. It is assumed that the only fac-
information for those individuals is re- tor influencing the respondent's an-
moved from the data set and analyses swers is the latent variable that the test is
are performed only on the remain- designed to measure.
ing (complete) records. This approach,
which assumes data are MISSING C O M - local m a x i m u m the largest value of
PLETELY AT RANDOM, reduces the effec-
a FUNCTION within a given range of
tive sample size and thus is likely to values. When the function is plotted,
overestimate STANDARD ERRORS and the local maximum is the highest point
reduce POWER. Also called casewise on the graph. Compare LOCAL MINI-
deletion; complete-case analysis. M U M .
Compare PAIRWISE DELETION. local m i n i m u m the smallest value of a
l i t e r a l r e p l i c a t i o n see REPLICATION. FUNCTION within a given range of val-
ues. When the function is plotted, the
L j u n g - B o x test a method used in TIME- local minimum is the lowest point on
SERIES ANALYSIS to evaluate the NULL the graph. Compare LOCAL M A X I M U M .
HYPOTHESIS that any AUTOCOR-
RELATIONS among data values observed local regression (loess; lowess) a
across multiple occasions are random. It form of REGRESSION ANALYSIS In which
is a more sophisticated version of the a model of the relationship between out-
BOX-PIERCE TEST. Also called Box- comes and predictors is obtained by fit-
Ljung test; Ljung-Box Q test. [Greta ting different linear or quadratic
M. Ljung, U.S. mathematician; George E. functions to different segments or inter-
P. Box (1919- ), British statistician] vals ofdata. Variables are assigned differ-
ent WEIGHTS in the REGRESSION
L L R abbreviation for LOG-LIKELIHOOD
EQUATIONS to reflect their relative im-
RATIO. portance at each point and no assump-
In abbreviation for natural LOGARITHM. tions are made about the associations
among the variables. The overall curve
l o a d i n g n. see FACTOR LOADING. obtained by combining the individual
l o a d i n g m a t r i x see FACTOR STRUC- fitted curves for the different data seg-
TURE MATRIX. ments shows the general shape of the re-
local effect any outcome from some lationship between the variables. Also
manipulation in a system or experiment called locally weighted regression.
that has a direct but limited influence. LOCF abbreviation for LAST OBSERVA-
For example, a local effect of an analge- TION CARRIED FORWARD.
sic dmg is numbing around the injec-
tion site. Compare REMOTE EFFECT. loess n. see LOCAL REGRESSION.
local independence in FACTOR ANAL- l o g a r i t h m (log) n. the number of times
YSIS and other models involving LATENT that a given value (the base) has to be
VARIABLES, the basic assumption that in- multiplied by itself to produce a specific
dividual observations have no direct in- quantity. For example, the logarithm for

199
logarithmic distribution

the quantity 81 and the base 3 is the more closely matches the assumptions
number4 ( 3 x 3 x 3 x 3 = 81). Logarithms required in particular statistical analy-
are given by the generic formula y = ses, particularly that of NORMALITY.
log^j;, thus 4 = log381. When the base
logical inference see INFERENCE.
is the particular value known as the
Eulerian number (e = 2.71828...), the logical positivism a philosophical
function is a natural logarithm (or perspective that is committed to the
hyperbolic logarithm) and denoted principle of verification, which holds
y = logeX. that the meaning and truth of all non-
tautological statements is dependent on
logarithmic distribution a PROBA- empirical observation. The positivist view
BILITY DISTRIBUTION derived from a log- of science is evident in psychology's com-
arithmic function. It is used to describe mitment to empirical scienfific methods
the probability that a random variable and focus on testing hypotheses de-
(e.g., the number of words understood duced from theory. Logical positivism
by a young child) equates to a given had waned by the middle of the 20th
value. Logarithmic distributions are century. See POSITIVISM. See also POST-
skewed, with long right tails, and usually POSITIVISM; REDUCTIONISM.
have small MEANS and large VARIANCES:
Values are discrete, peak at x = 1, and logistic d i s t r i b u t i o n a THEORETICAL
cannot be negative. Also called log DISTRIBUTION of contlnuous values that
series distribution. describes a set of relationships between
variables as defined by a LOGISTIC FUNC-
logarithmic function the function TION. Frequently used in studying
y = loggX, which gives the relation be- growth rates, it is similar to the NORMAL
tween a number (x) and its LOGARITHM DISTRIBUTION in Shape but has greater
(y) to the base A. It is equivalent tox = a^. KURTOSIS.
l o g a r i t h m i c scale a measurement that logistic f u n c t i o n a basic function of
uses the LOGARITHMS of the values ob- the form y = c/(l + a exp[-bx]), where y
tained rather than the scale values them- and X are variables, a, b, and c are con-
selves. Logarithmic scales are useful stants, and exp is the EXPONENTIAL
when the obtained data cover a very FUNCTION. When graphed, values de-
large range of values and are more easily rived from a logistic function form an S
described or interpreted when reduced shape called the logistic curve. The logis-
to a smaller, more restricted set of num- tic function is used to describe LONGITU-
bers. For example, loudness is assessed DINAL DATA and growth rates in
on a logarithmic scale, as there is a con- particular, in which an initial stage of
siderable difference between the loudest growth is approximately exponential
and softest sounds that may be detected and then, as saturation begins, slows
by the human ear. Compare LINEAR and ultimately ceases altogether.
SCALE.
logistic regression (LR) a form of RE-
l o g a r i t h m i c t r a n s f o r m a t i o n the GRESSION ANALYSIS used when the out-
conversion of raw data values into an- come or DEPENDENT VARIABLE may
other form via the use of LOGARITHMS. assume only one of two categorical val-
In turning multiplicative relationships ues (e.g., pass or fail) and the predictors
into additive ones and eliminating ex- or INDEPENDENT VARIABLES are either
ponential trends, logarithmic transfor- categorical or CONTINUOUS. For exam-
mations allow researchers to analyze ple, a researcher could use logistic regres-
data using LINEAR MODELS. Addi- sion to determine the likelihood of
tionally, such transformations may be graduating from college (yes or no)
used to convert raw data into a form that given such student information as high-

200
 log-rank test

school grade point average, college ad- rence of an event. Logits are the inverse
missions test score, number of advanced of LOGISTIC TRANSFORMATIONS.
placement courses taken in high school, logit analysis a statistical procedure
socioeconomic status, and gender. The similar to LOGISTIC REGRESSION that
analysis yields a LOGISTIC REGRESSION presents outcome probabilities in terms
EQUATION with an INTERCEPT and RE- of LOGITS.
GRESSION COEFFICIENTS that indicate
the probability of a specific outcome oc- logit model a mathematical model pos-
curring according to the values of the tulating some relation between the
predictors. Also called logistic model- LOGIT of observed probabilities in a sam-
ing. ple and certain unknown but estimated
population PARAMETERS. A logit model
logistic regression equation the for- often fits the data in cases where a linear
mula produced in a LOGISTIC REGRES- model using probabilities does not.
SION analysis. Given in the form y = e"/(l
log l i k e l i h o o d the LOGARITHM of a
- e"), in which e = the natural LOGA-
probability value or estimate.
RITHM and JT = the probability of an out-
come as determined by a + b^x^ + ^2^2 + log-likelihood r a t i o (LLR) a ratio
... bi,x (where a is the INTERCEPT and b whose numerator and denominator
is the regression weight of a predictor comprise LOG LIKELIHOODS. It is used (in
variable), it is the mathematical expres- the log-likelihood test) to evaluate the
sion of the relationship between a bi- GOODNESS OF FIT Of the NULL HYPOTHE-
nary outcome variable and one or more SIS and the ALTERNATIVE HYPOTHESIS in
predictors. A positive regression coeffi- explaining sample data. The simpler
cient indicates that the predictor in- LIKELIHOOD RATIO indicates how many
creases the probability of the outcome, more times likely the observed values are
whereas a negative regression coefficient to occur under one model than under
indicates that the predictor decreases the other, and the LOGARITHMS of those
the probability of that outcome. A large values yield a log-likelihood ratio.
regression coefficient indicates a predic-
log-linear analysis a method of exam-
tor that strongly influences the probabil-
ining relationships between two or more
ity of the outcome, whereas a near-zero CATEGORICAL VARIABLES that involves
coefficient indicates a predictor that has an analysis of the natural LOGARITHMS
little influence on the probabihty of that of frequency counts within a CONTIN-
outcome. GENCY TABLE. Log-linear analyses do not
distinguish between INDEPENDENT VARI-
logistic t r a n s f o r m a t i o n a TRANS- ABLES and DEPENDENT VARIABLES but
FORMATION in which measurements on
rather attempt to model all significant
a LINEAR SCALE are converted into prob-
associations among all variables, includ-
abilities between 0 and 1. It is given by
ing interactions between any combina-
the formula y = e^/(l + e"), where x is the
tion of the variables, using sets of ODDS
scale value and e is the Eulerian number
and ODDS RATIOS for different category
(see LOGARITHM). The inverse of LOGITS,
outcomes. Also called log-linear mod-
logistic transformations are used in such
eling.
statistical procedures as LOGISTIC RE-
GRESSION. log-normal d i s t r i b u t i o n a THEORET-
ICAL DISTRIBUTION in whlch the LOGA-
log^t n. a nonlinear probability function. RITHMS of values on a variable follow
It is given by the formula logit(p) = log the bell-shaped NORMAL DISTRIBUTION.
(p/[l - p]), where p is the probability of
an event. In other words, a logit is the log odds see LOGIT.
logarithm (log) of the ODDS of the occur- log-rank test a NONPARAMETRIC method

201
log series distribution

to compare the patterns of survival for studying change in knowledge about a

two samples that were randomly as- topic following an intervention and ob-
signed to a TREATMENT condition and a tains results indicating that those who
CONTROL CONDITION. If there is signifi- received the treatment had a better out-
cantly longer survival in the treatment come (i.e., had higher knowledge scores)
group, such that there are fewer inci- than those in the control group. If the
dences of a problem (e.g., disease, researcher decides to adjust for baseline
death), then the calculated value of the differences, however, by obtaining
log-rank test statistic will be larger than knowledge scores both before and after
its corresponding CRITICAL VALUE and the intervention, the pattern of findings
the treatment will be seen as successful. could be reversed: Where participants
Also called Cox-Mantel test; Mantel- receiving the treatment had substan-
Cox test. tially more initial knowledge than those
in the control group, it might appear
log series d i s t r i b u t i o n see LOGARITH- that they did not gain from the interven-
MIC DISTRIBUTION. tion despite their high scores at the end
l o n g i t u d i n a l data information ob- of the research. [Frederick M . Lord
tained through multiple measurements (1912-2000), U.S. psychometrician]
of the same individuals over a period of Lorenz curve a graphical representa-
time. For example, a researcher investi- tion of the inequality between the FRE-
gating use of coping strategies in college QUENCY DISTRIBUTION for a variable and
students may evaluate stress levels at the some comparison distribution. For ex-
beginning, middle, and end of the fall ample, Lorenz curves often are used to U-
and spring semesters. lustrate inequities of wealth in a society,
as in the following hypothetical illustra-
l o n g i t u d i n a l design the study of a
tion.
variable or group of variables in the
same cases or participants over a period
of time, sometimes of several years. An
example of a longitudinal design is a
comparative study of the same children
in an urban and a suburban school over
several years for the purpose of record-
ing their cognitive development in
depth. A longitudinal study that evalu-
ates a group of randomly chosen indi-
viduals is referred to as a panel study,
whereas a longitudinal study that evalu-
ates a group of individuals possessing
some common characteristic (usually
age) is referred to as a cohort study. Also
called longitudinal research; longi-
tudinal study. Compare CROSS-SEC- Cumulative % of Population
TIONAL DESIGN.
The curve shows the degree of disparity
Lord's paradox an effect in which the that exists between an arrangement of
relationship between a continuous out- equal wealth distribution and the actual
come variable and a CATEGORICAL inde- distribution. See GINI COEFFICIENT.
pendent variable (e.g., treatment or [Max O. Lorenz (1876-1959), U.S. econ-
control) is reversed when an additional omist]
COVARIATE is introduced to the analysis.
loss f u n c t i o n a formula that specifies a
For example, suppose that a researcher is

202
 lurking variable

penalty or "cost" associated with obtain- lower real l i m i t see REAL LIMIT.
ing an incorrect value estimate from a lower-tail p r o b a b i l i t y the probabil-
statistical model. It provides a quantifi- ity that a random variable will take a
cation of error and DECISION RULES that value considerably lower than the mean
serve to minimize the risk of drawing in- for that variable (i.e., one appearing at
accurate conclusions, the extreme left on the ;ic-axis in the plot
loss to f o l l o w - u p the situation in o f a P R O B A B I L I T Y D I S T R I B U T I O N ) . L o w e r -
which researchers lose contact with tail probabilities are important in certain
some participants in a study and thus statistical tests of experimental hypothe-
can no longer obtain data for them. Loss ses. Compare UPPER-TAIL PROBABILITY.
to follow-up may occur, for example, in lower w h i s k e r in a BOX-AND-WHISKER
a clinical study of a new dmg to treat PLOT, the line extending from the value
cancer if some patients unexpectedly at the 25th PERCENTILE to the smallest
move or otherwise become unreachable. value within one INTERQUARTILE RANGE
A common problem in LONGITUDINAL of that percentile score. Compare UPPER
DESIGNS, loss to foUow-up results in WHISKER.
missing data that may bias study find-
lowess n. see LOCAL REGRESSION.
ings.
LR 1. abbreviation for LIKELIHOOD RATIO.
lot acceptance s a m p l i n g see ACCEP- 2. abbreviafion for LOGISTIC REGRESSION.
TANCE SAMPLING.
LSD abbreviation for least significant dif-
lower hinge the point in a distribution ference. See FISHER LEAST SIGNIFICANT
of values below which lie one fourth of DIFFERENCE TEST.
the data. It is equivalent to the first LSD test see FISHER LEAST SIGNIFICANT
QUARTILE and lies midway between the DIFFERENCE TEST.
MEDIAN and the minimum point of the LTA abbreviation for LATENT TRANSITION
distribution. Compare UPPER HINGE.
ANALYSIS.
lower quartile see QUARTILE. l u r k i n g variable see HIDDEN VARIABLE.

203
 Mm
M symbol for SAMPLE MEAN. scured by, an INTERACTION EFFECT
between variables.
M A abbreviation for MOVING AVERAGE.
M A D 1. abbreviation for MEAN ABSOLUTE M A model abbreviation for MOVING-
AVERAGE MODEL.
DEVIATION. 2. abbreviation for MEDIAN
ABSOLUTE DEVIATION. M A N C O V A acronym for MULTIVARIATE
ANALYSIS OF COVARIANCE.
m a g n i t u d e o f effect see EFFECT SIZE.
Mahalanobis distance (symbol: D^) M a n h a t t a n distance see CITY-BLOCK
a MULTIVARIATE measure of the degree DISTANCE.
to which two sets of values differ from manifest variable a variable whose
one another: the mean distance between values can be directly observed or mea-
two sets of DEVIATION SCORES, relative sured, as opposed to one whose values
to the variances and covariances among must be inferred. In STRUCTURAL EQUA-
the respective variables. Larger values TION MODELING and FACTOR ANALYSIS
indicate greater distance, and hence less manifest variables are used to study LA-
similarity between sets. The Mahalanobis TENT VARIABLES. Also Called indicator
distance is useful for identifying OUTLI- variable.
ERS in a data set and is often applied in
m a n i p u l a t e d variable see EXPERI-
CLUSTER ANALYSIS, DISCRIMINANT ANAL-
MENTAL VARIABLE.
YSIS, and similar techniques. Also called
generalized distance. [Prasanta C. m a n i p u l a t i o n n. in an EXPERIMENTAL
Mahalanobis (1893-1972), Indian statis- DESIGN, the researcher's adjustment of
tician] an independent variable such that one
or more groups of participants are ex-
m a i n diagonal the line of values run-
posed to specific treatments while one
ning from the upper left to the lower
or more other groups experience a CON-
right of a SQUARE MATRIX.
TROL CONDITION. For example, a health
researcher could introduce a manipula-
tion such that a portion of the partici-
pants in a study randomly receive a new
drug, whereas the remaining partici-
pants receive only a PLACEBO. See EX-
As shown, the first value in the main di- PERIMENTAL VARIABLE.
agonal is the number in the first row and m a n i p u l a t i o n check any means by
thefirstcolumn of the matrix, the sec- which an experimenter evaluates the ef-
ond value is the number in the second ficacy of an EXPERIMENTAL VARIABLE,
row and second column, and so on. Also that is, verifies that a manipulation af-
called leading diagonal. fected the participants as intended.
m a i n effect the consistent total effect M a n n - W h i t n e y U test a NON-
of a single INDEPENDENT VARIABLE on a PARAMETRIC TEST of centrality for ORDI-
DEPENDENT VARIABLE over all Other in- NAL DATA that contrasts scores from two
dependent variables in an experimental INDEPENDENT SAMPLES to assess whether
design. It is distinct from, but may be ob- there are significant differences between

204
 marginal homogeneity

the two sets of rankings. The statistic ob- m a r g i n a l d i s t r i b u t i o n a distribution

tained from this test, U, is calculated by of scores that shows the effect of one
summing the number of ranks in one variable of interest after the influence of
group that are smaller than each of the other variables has been removed from
ranks in the other group. A Mann-Whit- the data. It is obtained by averaging out
ney U test is analogous to a ONE-WAY the different scores resulting from differ-
ANALYSIS OF VARIANCE, except that the ent conditions of the variables that are
former is conducted with ranked data to be excluded from the analysis. For ex-
and the latter is conducted with CON- ample, imagine a researcher is conduct-
TINUOUS data. See also W I L C O X O N - ing a study with two independent
MANN-WHiTNEY TEST. [Henry Berthold variables of treatment (treatment or con-
Mann (1905-2000), Austrian-born U.S. trol) and gender (male or female). In
mathematician; Donald Ransom Whit- order to understand the MAIN EFFECT of
ney (1915-2001), U.S. statistician] treatment, he or she could examine the
marginal distribution of means for the
M A N O V A acronym for MULTIVARIATE treatment and control groups, respec-
ANALYSIS OF VARIANCE.
tively, by averaging across the means for
Mantel-Cox test see LOG-RANK TEST. men and women. Similarly, to under-
[Nathan Mantel (1919-2002), U.S. bio- stand the main effect of gender the re-
stafistician; David Cox (1924- ), British searcher could examine the marginal
statistician] distribution of means for the two gender
groups by averaging across means for
Mantel-Haenszel test a statistical the treatment and control groups, re-
procedure used to determine whether spectively. See also MARGINAL MEAN.
two DICHOTOMOUS VARIABLES are inde-
pendent of each other across groups that m a r g i n a l effect a measure of how
differ on a third variable. For example, much an outcome or DEPENDENT VARI-
a health researcher may want to assess ABLE changes when an explanatory or
whether smoking status (smoker vs. INDEPENDENT VARIABLE Is modified by
nonsmoker) and cancer diagnosis (can- one unit. For example, if a researcher is
cer vs. no cancer) are independent in a examining grade point average (GPA)
subsample of men and also in a and the number of hours studied per
subsample of women. If smoking and week, the marginal effect would show
cancer were related in either or both how much GPA would change if study
of the subsamples, the Mantel-Haenszel time was increased by one hour.
statistic obtained from this procedure m a r g i n a l frequency the sum of any
would be larger than a CRITICAL VALUE. one of the rows or columns in a data ma-
Also called Cochran-Mantel-Haens- trix. For example, in a table of students
zel test. [William G. Cochran (1909- classified by sex and area of study, the
1980), British-born U.S. statistician; Na- number of female students, regardless of
than Mantel; WilUam M . Haenszel area of study, would be one marginal fre-
(1910-1998), U.S. biostatistician] quency, and the number of students en-
M A P E abbreviation for MEAN ABSOLUTE rolled in a specific area of study,
PERCENTAGE ERROR.
regardless of sex, would be another. See
also MARGINAL DISTRIBUTION.
M A R abbreviation for MISSING AT RAN-
DOM.
m a r g i n a l homogeneity in a two-way
CONTINGENCY TABLE, the Situation in
m a r g i n a l n. the sums or averages of any which the sum of each row is the same as
of the rows or columns of data in a ma- the sum of the corresponding column. A
trix. See C O L U M N MARGINAL; ROW MAR- test of marginal homogeneity often is
GINAL. conducted to assess agreement between

205
marginal mean

two sets of data, such as performance any CMV, the researcher introduces a
ratings given by two independent fig- marker variable into the study: This is a
ure-skating judges. new variable that is theoretically unre-
lated to at least one of the main variables
m a r g i n a l mean the average score
of interest. The correlation between the
across two or more groups on one vari-
marker variable and the unrelated vari-
able, particularly as calculated in an
able of interest is taken as an estimate of
ANALYSIS OF VARIANCE to determine the
the common method variance in the
existence of a MAIN EFFECT. See also
study.
MARGINAL DISTRIBUTION.
M a r k o v c h a i n a sequence of steps or
m a r g i n a l p r o b a b i l i t y the chance of events in which the probability of each
occurrence of one variable across each transition depends only on the immedi-
level of another variable. For example, a ately preceding step, and not on any ear-
researcher interested in cancer diagnosis lier step. For example, when a person
and smoking status might list diagnosis considers a behavior change such as los-
(cancer vs. no cancer) in two rows and ing weight he or she experiences certain
smoking status (smoker vs. nonsmoker) stages: The individual could move from
in two columns. The marginal probabili- not thinking about losing weight
ties for cancer diagnosis would be ob- (precontemplation) to considering a
tained by summing the individual weight loss plan (contemplation),
probabihties across both levels of smok- through to maintaining a reasonable
ing status, whereas for smoking status weight (maintenance) and conceivably
the marginal probabilities would be ob- back to not wanting to think about
tained by summing the individual prob- weight loss. This process could be de-
abilities across both levels of cancer scribed as a Markov chain. Also called
diagnosis. In other words, the marginal fUscrete-time Markov chain; Mar-
probabilities are the total probability kov process. Compare RANDOM WALK.
values given at the ends of the rows and [Andrei Markov (1856-1922), Russian
the bottoms of the columns. See also mathematician]
MARGINAL DISTRIBUTION.
Markov chain Monte Carlo
m a r g i n a l sum o f squares see TYPE III
method (MCMC method) a statistical
SUM OF SQUARES.
simulation procedure that tries to find a
m a r g i n of error (MOE) a statistic ex- reasonable pattern of probabilities for
pressing the CONFIDENCE INTERVAL as- moving from one designated state to an-
sociated with a given measurement; it is other in a MARKOV CHAIN. It involves
an allowance for a slight miscalculation sampling from various THEORETICAL
or an acceptable deviation. The larger DISTRIBUTIONS to see which is the best
the margin of error for the sample data, fit for observed data. For example, a
the less confidence one has that the re- Markov chain Monte Carlo method
sults obtained are accurate for the entire could be used to find the plausible pat-
population of interest. tern of probabilities of moving from sub-
stance dependence to recovery. See also
mark-and-recapture s a m p l i n g see MONTE CARLO RESEARCH. [Andrei
CAPTURE-TAG-RECAPTURE SAMPLING. Markov]
m a r k e r variable technique in FAC- M a r k o v m a t r i x see TRANSITION MA-
TOR ANALYSIS, a technique used to con- TRIX.
trol for common method variance
(CMV), that is, the systematic error that M a r k o v r a n d o m f i e l d a graphic rep-
can arise from using a single method of resentation of two or more discrete
measurement. In order to account for states or stages in a MARKOV CHAIN with

206
 matching

directional lines between them to indi- matched-pairs design a study involv-

cate the possible patterns of moving ing two groups of participants in which
from one state to the next or back to ear- each member of one group is paired with
lier states. For example, consider a be- a similar person in the other group, that
havior change model in the context of is, someone who matches them on one
quitting smoking, which could be repre- or more variables that are not the main
sented in a Markov random field. A focus of the study but nonetheless could
square may be used to depict one stage, influence its outcome. For example, a re-
smoking regularly, with a line drawn to searcher evaluating the effectiveness of a
a square representing the next stage, new drag in treating Alzheimer's disease
thinking about quitting, and so on might identify pairs of individuals of the
through extended quitting. Lines also same age and intelligence and then ran-
could be drawn from each of those domly assign one person from each pair
stages back to earlier stages to represent to the treatment condition that will re-
the possibility of lapsing during the pro- ceive the drag and the other to the con-
cess. [Andrei Markov] trol condition that will not. A matched-
groups design (matched-samples de-
M A R S acronym for MULTIVARIATE ADAP- sign; matched-subjects design) is a simi-
TIVE REGRESSION SPLINE. lar approach but broader in that it allows
martingale n. a STOCHASTIC process in for the inclusion of more than two
which the probability of any future groups of participants.
event is the same as that of the current matched-pairs signed-ranks test
event, such that knowledge of previous see WILCOXON MATCHED-PAIRS SIGNED-
events does not help one to predict fu- RANKS TEST.
ture events. For example, a game based
on a series of coin tosses would be a mar- matched-pairs t test a statistical pro-
tingale, as the probability of heads or cedure used to test for significant differ-
tails remains the same (50%) at each ences between the two sets of data
toss. In contrast, a game involving the obtained from a MATCHED-PAIRS DE-
drawing of cards from a standard deck SIGN. For example, a researcher could
without replacement would not be a use a matched-pairs t test to assess
martingale, as the probabihty of draw- whether relationship satisfaction signifi-
ing, say, a spade would change depend- cantly differs between two sets of indi-
ing on the suits of the cards drawn viduals who have been matched for age
previously. and gender where one set has been mar-
ried for 5 years and the other for 10
masking n. the obscuring of the effect of years. Also called paired-samples t
one variable by the effect of another test. See T TEST.
variable. For example, a researcher inter-
ested in whether risky sexual behavior is m a t c h i n g n. a procedure for ensuring
related to alcohol use might find that an that participants in different smdy con-
experience of sexual abuse exerts a ditions are comparable at the beginning
stronger influence on sexual behavior, of the research on one or more key vari-
thus masking the effect of alcohol use. ables that have the potential to influ-
mask vb. ence results. After multiple sets of
matched individuals are created, one
matched case-control study a CASE- member of each set is assigned at ran-
CONTROL STUDY in which MATCHING IS dom to the EXPERIMENTAL GROUP and
used to control for CONFOUNDS (e.g., the other to the CONTROL GROUP. For
age, sex). This approach is considered example, a researcher could create two
particularly appropriate where sample groups whose members are of the same
size is limited. sex and have the same family history of

207
matching coefficient

a disease; one group would be given a is broken down into its simpler compo-
treatment while the other would not. nent elements. See also CHOLESKY
Such an approach would enable the re- FACTORIZATION; SINGULAR VALUE DE-
searcher to rule out sex and family his- COMPOSITION.
tory as potential explanations of the
m a t u r a t i o n n. naturally occurring
study outcome, thereby allowing greater
time-related. changes in a participant
validity to attributing any changes be-
(e.g., growth, aging, fatigue, boredom,
tween the groups to the treatment.
attention shifts) that pose a threat to the
When individuals are paired according
INTERNAL VALIDITY of a smdy, particu-
to their PROPENSITY SCORES, the process
larly a longitudinal one. These pro-
is referred to as caliper matching. See
cessesas opposed to the specific
also MATCHED-PAIRS DESIGN. treatment or interventionmay explain
m a t c h i n g coefficient an index of the any changes in participants during the
correspondence between two sets of experiment. For example, a researcher
items, such as scores or other sample may study substance use in a set of indi-
data. For example, a matching coeffi- viduals from young adolescence to late
cient could be used during an investiga- adulthood. In the study, substance use
tion of possible plagiarism to assess how may naturally decline as a function of
much overlap there is between an essay the development of the participants
submitted by a student and an essay rather than because of the influence of
posted online. an experimental intervention. Thus, the
investigator would want to assess and
mathematical statistics see THEO- possibly control for this maturation
RETICAL STATISTICS.
effect in order to maintain the internal
mathematico-deductive method validity of the study.
see HYPOTHETICO-DEDUCTIVE METHOD.
Mauchly's sphericity test in
m a t r i x n. a rectangular ordered arrange- WITHIN-SUBJECTS ANALYSES OF VARI-
ment (ARRAY) of numbers in rows and ANCE, a technique to determine whether
columns. The following is a simple ex- there is HOMOGENEITY OF VARIANCE
ample: across all possible levels of the inde-
1 6 pendent variable (a condition known as
SPHERICITY). Sphericity is a basic as-
4 1
sumption of such analyses and lack of it
L3 7
can distort the calculations of variance.
Individual items in a matrix are caUed The Mauchly test is based on the LIKELI-
ELEMENTS or entries. Many different HOOD RATIO criterion and involves a
types of matrices are used in statistics, scaled comparison between the DETER-
such as the CORRELATION MATRIX and MINANT and the TRACE of the sample
the COVARIANCE MATRIX. covariance matrix. When the signifi-
cance level of Mauchly's test is < .05
m a t r i x algebra a set of mathematical then equality of variance cannot be as-
rales for analyzing large numbers of sumed. Dohn W. Mauchly (1907-1980),
variables arranged in matrices. Matrix U.S. computer scientist]
algebra procedures may involve basic
calculations, such as adding and sub- m a x i m i n strategy in GAME THEORY or
tracting matrices, and also more com- decision making, a tactic in which an in-
plex processes, such as multiplying and dividual chooses the best of a set of
dividing matrices. worst possible outcomes or payoffs. For
example, participants in a study may
m a t r i x decomposition in linear alge- need to choose whichever outcome
bra, a process in which a complex matrix would best maximize a minimum ad-

208
 mean absolute percentage error

vantage: withdrawing from a scenario in MONTE CARLO M A X I M U M LIKELIHOOD

which they are currently performing ESTIMATION.
poorly, or continuing and possibly
M C P abbreviation for MINIMUM CONVEX
doing even worse. Compare MINIMAX
POLYGON.
STRATEGY.
M D A abbreviation for MULTIPLE DIS-
m a x i m u m l i k e l i h o o d a statistical CRIMINANT ANALYSIS.
technique in which the set of possible
values for the PARAMETERS of a distribu- M D L abbreviation for MINIMUM DE-
tion is estimated based on the most SCRIPTION LENGTH.
probable sample of observations that MDS abbreviation for MULTIDIMEN-
one might have obtained from that pop- SIONAL SCALING.
ulation. The values derived from this
procedure are referred to as maximum mean n. the numerical average of a set of
likelihood parameter estimates or maxi- scores, computed as the sum of all scores
mum likelihood estimates and can be divided by the number of scores. For ex-
used to. obtain an overall estimate of ample, suppose a health researcher sam-
how well a given model fits the data. pled five individuals and found their
Maximum likelihood techniques are numbers of hours of exercise per week to
used in several statistical methods, such be 3, 1, 5, 4, and 7, respectively. The
as LOGISTIC REGRESSION, HIERARCHICAL mean number of exercise hours per week
LINEAR MODELING, a n d STRUCTURAL
thus would be (3 + 1 + 5 + 4 + 7)/5 = 20/5
EQUATION MODELING. For example, = 4. The mean is the most widely used
when evaluating a pair of regression statistic for describing CENTRAL TEN-
DENCY. Also called arithmetic mean;
models, a maximum likelihood ratio of
the overall estimates can be used to as- arithmetic average. See also GEO-
sess whether there is a significant differ- METRIC MEAN; HARMONIC MEAN.
ence between the two, which in turn can mean absolute deviation (MAD) a
be used to identify the preferred model. measure of DISPERSION, given as the typ-
Also called maximum likelihood es- ical distance of each score in a set from
timation (MLE). the mean of the set, irrespective of the
positive or negative direction of the dif-
m a x i m u m likelihood-expectation
ference. For example, consider a set of
m a x i m i z a t i o n estimation (ML-EM
estimation) a statistical method for cal- five scores: 7, 1, 10, 8, and 4. To obtain
culating MAXIMUM LIKELIHOOD esti- the mean absolute deviation for the set,
mates for the parameters of a model that one first calculates the mean by sum-
involves LATENT VARIABLES. The process ming all scores and dividing by the total
uses the EM ALGORITHM to obtain the number of scores: 30/5 = 6. One then
closest match between the sample data calculates the difference of each individ-
and the proposed model. ual score from this average regardless of
whether it is positive or negative, yield-
M C A abbreviation for multiple classifica- ing the following: 1, 5, 4, 2, and 2.
tion analysis. See MULTIPLE DIS- Summing these new values and dividing
CRIMINANT ANALYSIS. them by the total number of values pro-
vides the mean absolute deviation: 14/5
M C A R acronym for MISSING C O M - = 2.8. Also called average absolute
PLETELY AT RANDOM. deviation; average deviation (AD);
mean deviation. See also ABSOLUTE
M C M C method abbreviation for DEVIATION.
MARKOV CHAIN MONTE CARLO METHOD.
mean absolute percentage error
M C M L estimation abbreviation for (MAPE) a measure of variance in TIME-

209
mean deviation

SERIES values, given as the average dif- SQUARES divided by its DEGREES OF FREE-
ference of a set of such values from their D O M . It is used primarily in the ANALYSIS
expected values, regardless of the direc- OF VARIANCE, in which an F RATIO is ob-
tion of that difference (i.e., whether it is tained by dividing the mean square be-
positive or negative). For example, con- tween groups by the mean square within
sider the following time series of the groups (see BETWEEN-GROUPS MEAN
number of exercise hours per week for SQUARE; WITHIN-GROUPS MEAN SQUARE).
an individual over five weeks: 3, 1, 5, 4, The mean square also is used to deter-
and 7. If a researcher predicts the weekly mine the accuracy of REGRESSION ANAL-
exercise hours to be 2,2, 5,5, and 6, then YSIS models, indicating the amount of
he or she might wish to determine the variance explained by a model (MS^egres-
general degree of inaccuracy of his or her sion) compared to the amount of error or
predictions. Thus, the mean absolute unexplained variance (MSresiduai)-
percentage error would be calculated as
follows: (13 - 21/3 + 11 - 21/1 + 15 - 51/5 +mean
4 square between see BETWEEN-
GROUPS MEAN SQUARE.
- 51/4 + 7 - 6l/7)/5 = (.333 + 1 + 0 + .25 +
.143)/5 = 1.726/5 = .35 x 100 = 35%. mean squared error (symbol: MSE)
the average amount of ERROR VARIANCE
mean deviation see MEAN ABSOLUTE within a data set, given as the typical
DEVIATION. squared distance of a scorefromthe mean
mean difference a measure of variabil- score for the set. Mean squared error
ity in a data set calculated as the average may be calculated in both ANALYSIS OF
of the distances between each score and VARIANCE and REGRESSION ANALYSIS. In
each of the other scores, disregarding the former it is referred to more specifically
whether the deviation is positive or neg- as the WITHIN-GROUPS MEAN SQUARE and
ative. For example, consider the follow- used as the denominator when calculat-
ing three scores: 1, 3, and 9. The mean ing an F RATIO; in the latter it is known
difference would be calculated as: (11-31 as a residual mean square (or mean-
+ 11-91 + 13-11 + 13-91 + 19-11 + 19- square residual) and gives the mean dif-
3l)/6 = (2 + 8 + 2 + 6 + 8 + 6)/6 = 32/6 = ference between actual scores and those
5.33. predicted by a regression model. A large
mean squared error indicates that scores
mean effect size in a META-ANALYSIS, a are not honiogeneous within groups or
measure of the average EFFECT SIZE are not consistent with prediction, such
across multiple studies. For example, an that there is more "noise" than "signal."
investigator analyzing several studies as- For example, a large mean squared error
sessing a new treatment may determine in gender research would show no sig-
a mean effect size by calculating the av- nificant differences between groups of
erage standardized difference between males and groups of females. Also called
treatment and control groups over all of error mean square.
the studies. Calculations of mean effect
size often take variance and RELIABILITY mean-square deviation the typical
into account by assigning different difference between a set of scores and
WEIGHTS to the values derived from dif- the MEAN of those scores, raised to the
ferent studies. A study with a large sam- second power. It is a measure of variabil-
ple, for instance, or one that uses more ity in a data set generally equivalent to
precise measurement techniques may be the VARIANCE.
weighted to have a greater impact in de- mean-square residual (symbol: MSR)
termining the mean effect size. see MEAN S Q U A R E D ERROR.

mean square (symbol: MS) an estima- mean-square successive d i f f e r -

tor of variance calculated as a SUM OF ence (symbol: MSSD) the average varia-
 measurement scale

tion between each score and the pre- TEST THEORY, any difference between an
vious score in a L O N G I T U D I N A L DESIGN. observed score and the T R U E SCORE.
Used to obtain a more accurate estimate Measurement error may arise from flaws
of a trend over time in simations when in the assessment instrument, mistakes
the mean shifts slightly and could cause in using the instrument, or random or
bias, it is calculated in the same manner chance factors. For example, an investi-
as VARIANCE except that scores at succes- gator may obtain biased results from a
sive time points are used i n place of the survey because of problems with ques-
MEAN. tion wording or response options, ques-
tion order, variability in administration,
m e a n square w i t h i n see WITHIN-
and so forth.
GROUPS MEAN SQUARE.
m e a s u r e m e n t i n v a r i a n c e the situa-
m e a n s u b s t i t u t i o n a method of I M P U -
tion in which a scale or constract pro-
T A T I O N i n which the average value for a
vides the same results across several
set of scores is inserted for each missing
different samples or populations. For ex-
score in the data set. Although this prac-
ample, an intelligence test could be said
tice retains the original sample size, it
to have measurement invariance if it
can artificially reduce the variation
yields similar results for individuals of
among the scores. Mean substitution
varying gender, ethnicity, or age. Mea-
may involve inserting either the G R A N D
surement invariance may apply to a sin-
MEAN or the mean for a specific group;
gle relevant characteristic (e.g., gender
the latter approach typically yields a
only), several characteristics (e.g., gen-
more accurate estimate of any given in-
der and ethnicity), or all possible charac-
dividual's score.
teristics and situations. See also
m e a s u r e n. an item or set of items that MEASUREMENT EQUIVALENCE. Compare
provides an indication of the quantity or SELECTION INVARIANCE.
nature of the phenomenon under smdy.
m e a s u r e m e n t l e v e l the degree of
It is sometimes necessary in research to
specificity, accuracy, and precision in a
have more than one measure for each of
particular set of observations or scores,
the main variables of interest.
as reflected in the M E A S U R E M E N T S C A L E
m e a s u r e m e n t n. the act of appraising used.
the extent of some amount, dimension,
m e a s u r e m e n t m o d e l in STRUCTURAL
or criterionor the resultant descriptive
EQUATION MODELING, a model that
or quantified appraisal itself. A measure-
quantifies the association between ob-
ment is often, but not always, expressed
servations obtained during research (in-
as a numerical value.
dicators) and theoretical underlying
m e a s u r e m e n t e q u i v a l e n c e the situ- constracts or factors. When carrying out
ation in which two instruments yield a CONFIRMATORY FACTOR ANALYSIS, for
virraally the same results for the same example, one assesses a hypothesized
group of individuals, such that either measurement model that specifies the
one could replace the other. For exam- relationships between observed indica-
ple, a researcher constructing ALTER- tors and the LATENT VARIABLES that sup-
NATE FORMS of the same test would want port or affect them.
to ensure measurement equivalence for
the FACTOR LOADINGS of each item m e a s u r e m e n t scale any of four com-
within a subscale, as well as for each test mon methods for quantifying attributes
as a whole. See also M E A S U R E M E N T of variables during the course of re-
search, listed in order of increasing
INVARIANCE.
power and complexity: N O M I N A L SCALE,
measurement error in CLASSICAL ORDINAL SCALE, INTERVAL SCALE, and

211
measurement theory

RATIO SCALE. More specifically, nominal m e a s u r e o f v a r i a b i l i t y see DISPER-

scales consist of named categories with SION MEASURE.
no numerical meaning (e.g., gender,
birthplace), ordinal scales comprise m e d i a n n. the midpoint i n a distribu-
rankings from highest to lowest or vice tion, that is, the score or value that di-
versa (e.g., birth order, contest winners), vides it into two equal-sized halves. The
interval scales provide equal distance be- median is a M E A S U R E O F C E N T R A L T E N -
tween numerical values but have an DENCY that is particularly useful when
arbitrary zero point (e.g., degrees Fahr- analyzing data that have SKEWNESS (i.e.,
enheit, checkbook balance), and ratio lopsidedness), as it is more resistant to
scales provide equal distance between the influence of extreme values.
numerical values with an exact zero
point (e.g., height, weight). median absolute deviation (MAD)
an index of the variance i n a set of
m e a s u r e m e n t t h e o r y a field of study scores, given as the midpoint of the dif-
that examines the attribution of values ference between each score in a set and
to traits, characteristics, or constructs. the median of the set, irrespective of the
Measurement theory focuses on assess- positive or negative direction of the in-
ing the TRUE SCORE of an attribute, such dividual differences. For example, as-
that an obtained value has a close corre- sume that five individuals spent the
spondence with the acraal quantity, following number of hours per week
with high RELIABILITY and littie M E A - using a computer: 6, 10, 15, 21, and 35.
S U R E M E N T ERROR. See also CLASSICAL To obtain the median absolute devia-
TEST THEORY. tion, one first identifies the median for
the set as a wholein this case, 15. Next,
m e a s u r e o f a g r e e m e n t see COEFFI- one calculates the difference of each in-
CIENT OF AGREEMENT. dividual score from this overall median
irrespective of its directionin this case,
m e a s u r e o f a s s o c i a t i o n any of vari-
9, 5, 0, 6, and 20. One then orders
ous indices of the degree to which two or
more variables are related. One of the the differences0, 5, 6, 9, and 20and
most common measures of association is finds the midpoint of this new array,
yielding a median absolute deviation
the CORRELATION COEFFICIENT.
of 6.
m e a s u r e o f c e n t r a l t e n d e n c y any of
a class of descriptive statistics that reflect m e d i a n test a NONPARAMETRIC
the middle or central point of a set of method that assesses the equality of the
scores. The rnost common measures of midpoints (MEDIANS) in two or more
central tendency are the MEAN, MEDIAN, samples of data to determine whether
and MODE, which are the average, mid- they come from the same population.
point, and most frequently occurring Although previously used to indicate
scores in a distribution, respectively. any of several alternative procedures,
Also called measure of location. nowadays the term generally is used to
refer to the W I L C O X O N - M A N N - W H I T N E Y
m e a s u r e o f c o r r e l a t i o n see CORRELA- TEST.
TION COEFFICIENT.
m e d i a t i o n a l p r o c e s s a set of relation-
m e a s u r e o f d i s p e r s i o n see DISPER- ships in which an I N D E P E N D E N T V A R I -
SION MEASURE. ABLE is linked to a M E D I A T O R that in
turn is linked with and influences a DE-
m e a s u r e o f f i t see FIT I N D E X .
P E N D E N T V A R I A B L E . For example, the
m e a s u r e o f l o c a t i o n see MEASURE OF progression from childhood sexual
CENTRAL TENDENCY. abuse to adult risky behavior may be a
 metatheory

mediational process that involves pow- M-estimator n. a type of ESTIMATOR

erlessness. used to calculate the minimum differ-
mediator n. in statistical analyses, an in- ence between an acraal value obtained
termediaty or intervening variable that from a sample and the hypothesized
value in the larger population or the esti-
accounts for an observed relation be-
mated value obtained from a model. The
tween two other variables. For example,
estimators used in LEAST SQUARES RE-
a researcher may posit a model involv-
GRESSION and MAXIMUM LIKELIHOOD
ing an INDEPENDENT VARIABLE of abil- estimation are examples.
ity, a mediator of self-efficacy, and a
DEPENDENT VARIABLE of achievement. meta-analysis n. a quantitative tech-
Thus, ability is hypothesized to influence nique for synthesizing the results of
self-efficacy, which in turn is thought to multiple studies of a phenomenon into
influence achievement. a single result by combining the EFFECT
member check a procedure used in SIZE estimates from each study into a
QUALITATIVE RESEARCH whereby a re- single estimate of the combined effect
searcher returns to a participant and re- size or into a DISTRIBUTION of effect
quests feedback on his or her original sizes. For example, a researcher could
responses as recorded and sometimes on conduct a meta-analysis of several smd-
their interpretation by others. Con- ies on the association between self-
ducting a member check is intended to efficacy and achievement, integrating
ensure the accuracy and INTERNAL VA- the findings into an overall correlation.
LIDITY of a qualitative study. However,
Although meta-analysis is ideally suited
some researchers question the merit of for summarizing a body of literature in
the process, viewing it as inconsistent terms of its impact, limitations, and fu-
with qualitative research in that it as- rare implications, there are conditions
sumes participants' experiences are ob- that limit its applicability. For example,
jective and finite. Also called member there is no minimum number of sradies
validation; respondent validation. nor participants required, and informa-
tion of potential interest may be missing
mesokurtic adj. describing a FRE- from the original research reports upon
QUENCY DISTRIBUTION that is neither which the procedure must rely.
flatter nor more peaked than the NOR-
MAL DISTRIBUTION. metaregression analysis a procedure
similar to a META-ANALYSIS but con-
ducted with two or more REGRESSION
ANALYSES rather than experimental stud-
ies. For example, a researcher could con-
duct a metaregression analysis of relevant
studies predicting college achievement
from high school grades, integrating the
findings into a single overall REGRESSION
EQUATION. Thus, a metaregression analy-
sis investigates the causes of heterogene-
ity (differences) across studies in order to
determine when, where, and for whom a
particular outcome will occur.
That is, a mesokurtic arrangement of
values follows a bell-shaped curve, with metatheory n. a higher order theory
the majority of scores clustered around a about theories, allowing one to analyze,
value at the midpoint and a few extreme compare, and evaluate competing bod-
scores tapering off on either side. See ies of ideas. The concept of a metatheoty
also PLATYKURTIC; LEPTOKURTIC. suggests that theories derive from other

213
method of least squares

theories such that there are always prior methodological t r i a n g u l a t i o n the

theoretical assumptions and commit- use of multiple quantitative and qualita-
ments behind any theoretical formula- tive procedures to collect data so as to
tion. It follows that these prior generate converging evidence on the
assumptions and commitments are wor- topic of study. For example, a researcher
thy of srady in their own right, and an sradying alcohol consumption might
understanding of them is essential to a employ methodological triangulation
full understanding of derivative theo- by measuring participants' blood alco-
ries. metatheoretical adj. hol levels, collecting self-reports on
quantity of alcohol consumed, and ob-
method o f least squares see LEAST taining input from peers on quantity of
SQUARES REGRESSION. alcohol consumed.
method of maximum likelihood methodology n. 1. the science of
see MAXIMUM LIKELIHOOD. method or orderly arrangement; specifi-
method of moments an alternative to cally, the branch of logic concerned
MAXIMUM LIKELIHOOD estimation in with the application of the principles of
which values describing basic character- reasoning to scientific and philosophi-
istics of a sample (known as MOMENTS) cal inquiry. 2. the system of methods,
are used to approximate the correspond- principles, and rules of procedure used
ing values for the larger population. For within a particular discipline. For exam-
example, a researcher could use the ple, in research and experimental design
method of moments to estimate the the term refers to the techniques used to
mean or variance for a population by in- collect information, and in statistics it
serting the appropriate sample value refers to the procedures used to analyze
into an equation that calculates one such data.
from the other. Although simpler than metric n. a scale or system used to ex-
maximum likelihood, the method of press amount or quantity. For example,
moments is less accurate. the Fahrenheit scale is a metric for as-
methodological i n d i v i d u a l i s m a sessing temperature and the system of
theoty that emphasizes the importance IQ points is a metric for assessing intelli-
of each person in determining his or her gence.
own circumstances and minimizes any Metropolis-Hastings a l g o r i t h m a
possible influence from larger societal SIMULATION procedure for generating
groups or structures. For example, a re- samples from a known PROBABILITY DIS-
searcher who endorsed methodological TRIBUTION. It is among the most com-
individualism would be less apt to con- monly used MARKOV CHAIN MONTE
sider how socioeconomic staras might CARLO METHODS. [Nlcholas C. Metropo-
contribute to the level of achievement lis (1915-1999), U.S. physicist; W. Keith
for a sample of individuals, preferring Hastings (1930- ), Canadian statisti-
instead to investigate variables believed
cian]
to be under each person's control, such
as motivation, determination, or ability. M l abbreviation for MODIFICATION
INDEX.
methodological p l u r a l i s m the belief
that various approaches to conducting m i c r o d a t a pl. n. scores or other infor-
research, qualitative and quantitative, mation collected on individual partici-
each have their respective strengths and pants or units. In many research studies,
weaknesses such that no one method is microdata are consolidated into total
inherently superior to any other and no scores, scale scores, or summaty statistics
single method is best overall. known as AGGREGATE DATA.
 minimum volume ellipsoid

microgenetic method a research ap- most an opponent will gain. For exam-
proach that examines developmental ple, a health researcher may propose an
change within a single set of individuals intervention that would be the least
over relatively brief periods of time, usu- aversive treatment for a serious disease,
ally days or weeks, in order to determine thereby minimizing the adverse effects
how it occurs. For example, health re- patients may expect to experience as a
searchers could use a microgenetic result of the disease. Compare MAXIMIN
method to study the weekly improve- STRATEGY.
ment of patients with traumatic brain
injury during a one-year period. m i n i m u m chi-square a method of
identifying acceptable PARAMETER esti-
middle-range theory an approach to mates that involves ttying to find the
the construction of THEORY that aims to smallest possible CHI-SQUARE value
combine existing formulations with ex- when assessing the differences between
perimental findings and in the process observed and expected data points. Min-
to generate new hypotheses that are imum chi-square is an alternative to
open to empirical testing. Representing MAXIMUM LIKELIHOOD for finding pop-
a level of investigation below that of ulation values that are close to observed
a METATHEORY, middle-range theories sample values.
often are used in sociology and psychol-
ogy to identity the constracts and mech- m i n i m u m convex polygon (MCP) a
anisms that best explain relevant data. procedure for estimating the boundaries
For example, a researcher could identity of an area in which a designated POPU-
the main concepts related to organiza- LATION may be found, based on sets of
tional change by investigating previous location points collected over a period of
formulations and findings in this area. time for different individuals. For exam-
ple, a health researcher may draw a min-
m i d p o i n t n. see MEDIAN. imum convex polygon around an area
believed to hold the set of individuals
midrange . the average of the lowest who may have been exposed to a serious
and highest scores in a set of data. The illness, based on reports about individu-
midrange is a MEASURE OF CENTRAL TEN- als with the illness.
DENCY more prone to bias than the
MEAN, MEDIAN, and MODE since it relies minimum description length
solely upon the two most extreme (MDL) in STRUCTURAL EQUATION MOD-
scores, which potentially are OUTLIERS. ELING, the principle that the best model
For example, consider the following for a given set of data is the one that pro-
hours per week spent using a computer vides the most compact description of
for five individuals: 6,10,15, 21, and 35. this data. See also OCCAM'S RAZOR.
The midrange is 20.5, the mean is 17.4,
m i n i m u m variance b o u n d see
and the median is 15. The midrange
CRAMER-RAO LOWER BOUND.
score thus is larger than the mean or me-
dian due to the influence of the outHer m i n i m u m variance unbiased esti-
of 35 hours. mator (MVUE) see EFFICIENT ESTIMA-
TOR.
midspread rz. see INTERQUARTILE RANGE.
MIMIC model acronym for MULTIPLE
m i n i m u m volume ellipsoid a statis-
INDICATORS-MULTIPLE CAUSES MODEL.
tical procedure for finding the smallest
space that would encompass most of the
m i n i m a x strategy in GAME THEORY or points in a MULTIVARIATE data set, as
decision making, a tactic in which indi- in the generic depiction overleaf.
viduals attempt either to minimize their Identitying the minimum volume el-
own maximum losses or to reduce the lipsoid, for example, would help a re-

215
misclassification cost

item than are the women). If so, the val-

ues would be missing at random. When
data are missing at random but not MISS-
ING COMPLETELY AT RANDOM there is a
danger of biasing srady results.
.a
ra
missing completely at random
(MCAR) the situation in which the ab-
sence of certain points from a data set is
unrelated to the nature of the variable
being measured or to any other variables
1 1 1 1 r in the research. For example, assume a
researcher identifies 100 items to ask
Variable x
young children in order to measure 10
searcher to observe the general location concepraal variables. Given the still-
and DISPERSION of the main portion of developing cognitive skills of the partici-
the data and also ascertain any OUTLIER pants, however, no more than 40 ques-
points outside the structure that may be tions may be used with each child in
considered for deletion due to their ex- order to obtain a sufficient response
tremity. quality. Thus, if the researcher presents a
computer-chosen set of 40 of the origi-
misclassification cost an estimate of nal items to each individual, the 60
the costs that could arise from assigning dropped items would be missing com-
an incorrect category or status to a situa- pletely at randomthey have no sys-
tion or condition. For example, it would tematic pattern and do not bias the
be worthwhile to estimate the mis- srady results.
classification cost of prematurely telling
the public that previously contaminated missing values procedure any strat-
town water is now safe to drink, versus egy for addressing the problem of empty
waiting longer than needed to let people items in a data set. A variety of alterna-
know they can drink the water. In this tives exist, each having different merits
case, the misclassification cost of an and drawbacks. Traditional missing val-
early announcement would be much ues procedures include LISTWISE DELE-
more seriouspossibly involving severe TION, in which all of the data for a
illness or even loss of lifethan that of a participant or case are deleted if even
later announcement. one item is left blank; PAIRWISE DELE-
TION, in which certain data are retained
missing at r a n d o m (MAR) the situa-
for a participant even where some other
tion in which the absence of certain
items are missing; and MEAN SUBSTITU-
points from a data set is unrelated to the
TION, in which each missing value is re-
nature of the particular items but may be
placed by the average value of all points
explained by other variables. For exam-
in the data set. More recent techniques
ple, assume a researcher surveyed 100
include IMPUTATION, in which a value is
students on a college campus about their
inserted into the data to substitute for a
weekly amount of alcohol, drug, and
missing one, and its many variations.
cigarette use but 12 individuals failed to
Notable among them is MULTIPLE IMPU-
note their use for an item. If all 12 peo-
TATION, whereby several possible values
ple are male, and if there is no consis-
are inserted into the data to replace
tency as to which item was left blank by
omitted ones and then the average and
which person, then the pattern of re-
standard deviation of those replacement
sponses on the survey could be related to
values are obtained.
sex (perhaps the men on campus are less
likely to answer any type of evaluative misspecification n. the situation in

216
 mixed-effects model

which the number of variables, factors, m i x e d d i s t r i b u t i o n a DISTRIBUTION

PARAMETERS, or some combination of in which the variables may take on both
these was not correctly specified in a sta- a DISCRETE and a CONTINUOUS range of
tistical model, with the result that the values. This may occur when a continu-
model does not offer a reasonable repre- ously measured variable is abruptly
sentation of obtained data. Mis- topped off or bottomed out at an arbi-
specification is indicated by such things trary value for a number of participants
as large RESIDUALS between the data and in the sample. For example, consider a
that predicted by the model, large STAN- health researcher studying the effects of
DARD ERRORS for the parameter esti- a smoking cessation intervention who
mates, values for the parameter estimates records the number of days since a par-
outside of the normal range (e.g., HEY- ticipant quit smoking. For those partici-
WOOD CASES), or a nonsignificant result pants who quit, the number of days will
from a GOODNESS-OF-FIT TEST. reflect an infinite series of potential val-
ues, whereas for those who continue to
m i x e d design a srady that combines smoke the number of days will be the
features of both a BETWEEN-SUBJECTS single finite value of zero.
DESIGN and a WITHIN-SUBJECTS DESIGN.
Thus, a researcher examines not only mixed-effects design a vague term
the potential differences between two or used to denote an approach to either
more separate groups of participants but research or analysisthe MIXED DESIGN
also assesses change in the individual or the MIXED-EFFECTS MODEL, respec-
members of each group over time. For tively.
example, a researcher might use a mixed
design to study the influence of different mixed-effects logistic regression a
types of music on relaxation. He or she statistical procedure for analyzing a bi-
could divide participants into a CON- nary outcome in which at least one of
TROL GROUP (listening to no music) and the predictors is a FIXED EFFECT (i.e., spe-
two E.XPERIMENTAL GROUPS (one listen- cifically selected by the researcher for
ing to classical music and one listening srady) and one or more other predictors
to rock music). The researcher could ad- is a RANDOM EFFECT (i.e., chosen by
minister a PRETEST to participants in all chance from a range of possible choices).
groups in order to determine the base- For example, an educational researcher
line level of physiological arousal prior might use mixed-effects logistic regres-
to hearing any music and then introduce sion to examine the odds of a student se-
the music and test participants while lecting a quantitative major in college
they listen. After stopping the music, he given his or her gender (a fixed predic-
or she could administer another test (a tor) and whether he or she has taken one
POSTTEST) to determine what specific re- or more specific quantitative classes in
duction in arousal may have occurred high school (random predictors), such
throughout the listening period. In this as quantitative psychology or research
situation, music type is a between- methodology.
subjects factor (each participant hears
only a single genre of music) and physio- mixed-effects model any statistical
logical arousal is a within-subjects factor procedure or experimental design that
(each participant is evaluated on this uses one or more independent variables
variable on multiple occasions and the whose levels are specifically selected by
different assessments compared). the researcher (FIXED EFFECTS; e.g., gen-
der) and one or more additional inde-
mixed-design analysis of variance pendent variables whose levels are
see MIXED-MODEL ANALYSIS OF VARI- chosen randomly from a wide range of
ANCE. possible values (RANDOM EFFECTS; e.g..

217
mixed-methods research

age). Also called m i x e d m o d e l . Com- m i x t u r e d i s t r i b u t i o n see FINITE M I X -

pare FIXED-EFFECTS MODEL; RANDOM- TURE DISTRIBUTION.
EFFECTS MODEL.
M L E abbreviation for M A X I M U M LIKELI-
H O O D estimation.
m i x e d - m e t h o d s r e s e a r c h a srady
that combines aspects of both QUALITA- M L - E M e s t i m a t i o n abbreviation for
TIVE RESEARCH and Q U A N T I T A T I V E R E - MAXIMUM LIKELIHOOD-EXPECTATION
SEARCH so as to more fully understand MAXIMIZATION ESTIMATION.
the phenomenon of interest. For exam-
m o d a l class for grouped data, the inter-
ple, a researcher studying a disease could
val or categoty containing the M O D E
conduct a focus group with a set of indi-
(most frequently occurring score).
viduals who would share their experi-
ences in dealing with the disease, and m o d a l f r e q u e n c y the number of re-
then supplement those qualitative find- spondents who have the score that oc-
ings by surveying a different set of indi- curs most often in a set of data (i.e., the
viduals to obtain quantitative knowl- M O D E ) . For example, in a small sample
edge of risk factors for the disease. of college freshmen whose ages are 18,
18, 18, 19, 19, 20, 20, 21, 23, and 25, the
mixed model see MIXED-EFFECTS modal frequency for age is 3: The most
MODEL. commonly occurring score is 18, and
there are three people with such a score.
mixed-model analysis of variance
a TWO-WAY ANALYSIS OF VARIANCE in
mode n. the most frequently occurring
which one independent variable has score in a set of data, which is sometimes
fixed levels specifically chosen for inves- used as a measure of CENTRAL TEN-
tigation (e.g., gender) and the other has DENCY. Also called modal value.
levels randomly selected from among m o d e l n. a graphic, theoretical, or other
many possible conditions (e.g., doses type of representation of a concept or
of a drug). It is one of many types process that can be used for various in-
of MIXED-EFFECTS MODELS. Also called vestigative and demonstrative purposes,
mixed-design analysis of variance. such as enhancing understanding of the
concept, proposing hypotheses, show-
m i x e d s a m p l i n g a method of collect- ing relationships, or identitying pat-
ing data that incorporates more than terns.
one sampling strategy. For example, a re-
searcher initially may choose partici- m o d e l I a n a l y s i s o f v a r i a n c e a rare
synonym of FIXED-EFFECTS A N A L Y S I S O F
pants by RANDOM S A M P L I N G , in which
each person in an identified population VARIANCE.
(e.g., cancer patients within a particular m o d e l II a n a l y s i s o f v a r i a n c e a rare
region) has an equal probability of being synonym of R A N D O M - E F F E C T S A N A L Y S I S
selected. However, the researcher may OF VARIANCE.
then need to switch to a STRATIFIED
SAMPLING or CLUSTER SAMPLING ap- m o d e l b u i l d i n g the process of positing
proach if he or she discovered that cer- several alternative statistical models of a
tain subgroups (e.g., women with lung phenomenon by adding various P A R A M -
cancer, men with colon cancer) were not ETERS to or deleting various parameters
adequately represented in the sample from the model originally proposed and
and thus needed to be sampled more then comparing the different versions to
specifically. The two approaches to- identity the simplest way to accurately
gether (e.g., simple random sampling represent the data.
and stratified random sampling) consti- m o d e l c a l i b r a t i o n the process of im-
tute mixed sampling. proving the accuracy of a statistical

218
 modification index

model by adjusting the number or na- tors have a significant interaction be-
ture of its PARAMETERS until the output tween them, and this interaction
from the model is seen to match an ob- changes the narare of the relationships
served sample of data. (e.g., the REGRESSION COEFFICIENTS)
among the other predictors when it is
model f i t assessment the process of
included in the analysis, then moder-
determining how close a proposed
ated multiple regression would be more
model of a relationship between vari-
appropriate.
ables is to the actual sample data ob-
tained. Model fit assessment often moderating effect the effect that oc-
involves a GOODNESS-OF-FIT TEST, an ex- curs when a third variable changes the
amination of the RESIDUALS between nature of the relationship between a pre-
the proposed representation and the ac- dictor and an outcome, particularly in
tual data, and an evaluation of whether analyses such as MULTIPLE REGRESSION.
the proposed representation is consis- For example, STRUCTURAL EQUATION
tent with relevant theory. Statistical pro- MODELING can be used to assess whether
cedures, such as STRUCTURAL EQUATION a predicted association between quanti-
MODELING and HIERARCHICAL LINEAR tative skill and performancefitsequaUy
MODELING, use some or all of these vari- well across different teaching style
ous methods to gauge model fit. groups (e.g., lecture based vs. hands-on
learning). If the prediction is different
model-free test a statistical procedure across the two groups, then teaching
for identifying the underlying structure style is said to have produced a moderat-
or pattern of associations for a phenom- ing effect. Also called moderator ef-
enon that does not require any initial as- fect. See also MODERATED MULTIPLE
sumptions about the nature of this REGRESSION.
pattern. A NONPARAMETRIC TEST is an
example of a model-free test. m o d e r a t o r n. an INDEPENDENT VARI-
that changes the narare of the rela-
ABLE
model i d e n t i f i c a t i o n see IDENTIFICA-
tionship between other variables. For
TION.
example, if a researcher examined the
m o d e l i n g effect a type of EXPERI- relationship between gender and math
MENTER EFFECT in which a participant is performance, a significant difference
unwittingly influenced to give responses might emerge. However, if teaching
similar to the responses the experi- style were taken into account, such that
menter would give if the experimenter those who learned math by applied,
were a participant. hands-on methods performed better
model misspecification see MIS- than those who learned with traditional
lecrare styles, regardless of gender, one
SPECIFICATION.
could say that teaching style was a mod-
model-to-data f i t see FIT. erator of the relationship between gen-
der and math performance. Also called
moderated m u l t i p l e regression a
moderating variable. See also MOD-
statistical procedure that is appropriate
ERATING EFFECT.
when there is a single CONTINUOUS out-
come, two or more PREDICTOR VARI- m o d i f i c a t i o n index (MI) a measure
ABLES, and one or more interactions that indicates the extent to which a
between the predictor variables. For ex- model could be improved if a specific
ample, a researcher might use tradi- PARAMETER were added, or alternatively
tional MULTIPLE REGRESSION to predict whether a parameter could be reason-
mathematics performance from the vari- ably deleted without significantly alter-
ables of quantitative skill, gender, and ing model fit. Although procedures such
teaching style. If the latter two predic- as STRUCTURAL EQUATION MODELING

219
modified replication

often provide a set of modification in- able to draw accurate conclusions. For
dexes as part of the output, researchers example, a researcher could conduct
should only consider amending a model momentaty time sampling to assess the
when using INDEPENDENT SAMPLES and occurrence of smoking at 100 specific
when the modification is consistent time points during a month-long pe-
with relevant theory. riod, using a small portable device that
m o d i f i e d r e p l i c a t i o n see REPLICA-
prompts participants to record their
smoking behavior at each point.
TION.
modulus n. see ABSOLUTE VALUE. moment generating f u n c t i o n a for-
mula for calculating the various MO-
MOE abbreviation for MARGIN OF ERROR. MENTS of a distribution for a random
m o l a r approach any theory or variable. The values obtained from a mo-
method that stresses comprehensive ment generating function describe the
concepts or overall frameworks or struc- range and shape of the set of possi-
tures. For example, METATHEORY uses a ble scores on that variable. See also
molar approach to understanding and CUMULANT G E N E R A T I N G FUNCTION;
developing ways of thinking about PROBABILITY GENERATING FUNCTION.
knowledge and research.
monomethod bias the lack of RELI-
molecular approach any theory or ABILITY or VALIDITY that may occur
method that stresses the components of when measuring a phenomenon with
a phenomenon, process, or system, mak- a single item, scale, or observation. For
ing use of elemental units in its analysis. example, monomethod bias could be
In a molecular approach the specific de- present if a researcher used only one
tails are more important than the overall questionnaire to measure a characteris-
perspective. tic or trait, or if a teacher used scores on
moment n. the power to which the EX- only one exam to evaluate students' per-
PECTED VALUE of a RANDOM VARIABLE is
formance for the entire academic year.
raised. Thus, E(x'') is the *c"^ moment of Also called mono-operation bias.
X. The first moment is usually the MEAN monotonic adj. denoting a variable that
of a variable, the second moment refers either increases or decreases as a second
to VARIANCE, the third moment relates variable either increases or decreases, re-
to SKEWNESS, and the fourth moment spectively: The relationship is not neces-
concerns KURTOSIS. Knowing each of sarily LINEAR but there are no changes
these moments provides a complete pic- in direction. A monotonically increasing
ture of the DISTRIBUTION for a set of variable is one that rises consistently as a
scores: A researcher knows the center second variable increases, for example,
point of the data, how spread out the level of performance in relation to
values are, whether they are lopsided, amount of practice if this were observed
and whether they are peaked (LEPTO- to be the case. In contrast, depression
KURTIC) or flat (PLATYKURTIC). would be a monotonically decreasing
moment about the mean see CEN- variable if its severity were found to fall
TRAL MOMENT.
consistently as a person's level of
perseveration declined.
momentary t i m e s a m p l i n g a proce-
dure in which the researcher indicates monotonic regression a NONPARA-
whether a particular behavior occurred METRIC method used when an outcome
during a designated interval of observa- variable is expected to systematically in-
tion. It is important to collect momen- crease or decrease as a FUNCTION of one
tary time sampling data for a large or more predictor variables. For exam-
number of time periods in order to be ple, monotonic regression could be used
 moving average

to assess the relationship between hours shape, and structure are important not
of exercise and body mass index, only in biology but also in neuro-
whereby the latter consistently decreases psychology and medical psychology.
as the former consistently increases. For example, researchers interested in
morphometries could study the form
monotonic relationship any associa- and stracture of the brain in individuals
tion between two variables in which in- who have various physical or psycholog-
crease or decrease in one produces a ical disorders. By contrast, PSYCHO-
corresponding increase or decrease in METRICS specificaUy focuses on the
the other. measurement of psychological phenom-
m o n o t r a i t - m u l t i m e t h o d model see ena.
MULTITRAIT-MULTIMETHOD MODEL.
m o r t a l i t y effect the degree to which
Monte Carlo m a x i m u m l i k e l i h o o d circumstances or behavior increase or
estimation (MCML estimation) in decrease the incidence of death. For ex-
MONTE CARLO RESEARCH, a Simulation ample, a behavioral health researcher
procedure for obtaining a good approxi- could srady the mortality effects of a
mation of a PARAMETER value (i.e., one fatty diet and lack of exercise, which
that closely matches the sample values could lead to early death from heart dis-
generated in the simulation). MCML es- ease.
timation often is used in BAYESIAN IN- m o r t a l i t y odds ratio a measure of the
FERENCE. chance of dying from a specific illness as
Monte Carlo research a SIMULATION the result of a specific behavior or cir-
technique in which a large number of cumstance. For example, a medical psy-
samples with specific selected properties chology researcher could examine the
(e.g., NORMALITY, slze, modeltype)are ODDS RATIO for dying from cancer for
generated by computer in order to assess individuals who smoke a pack or more
the behavior of a statistical procedure or of cigarettes per day. The mortality odds
PARAMETER Under vatying conditions. ratio wiU be greater than 1.0 (as in this
For example, an investigator might con- example) when the chance of the out-
duct Monte Carlo research with a large come (i.e., death) occurring is high and
number of normally distributed samples less than 1.0 when the characteristic
of various sizes (e.g., N = 50, 100, 200, would lead to a smaller chance of death
400, 800) in which a structural model is (e.g., as for the characteristic of a healthy
applied to characterize the data. Results diet).
would help the researcher determine the m o r t a l i t y rate a measure of how often
conditions under which the model be- death occurs, usually with respect to a
haves correctly (i.e.,fitsthe data) as well specific illness, characteristic, behavior,
as shows its limits (e.g., not fitting well or population. For example, a researcher
with sample sizes less than 200). Also could estimate the mortality rate for in-
called Monte Carlo method. dividuals who have been diagnosed with
m o r b i d i t y rate the incidence of dis- cancer.
ease, expressed as a ratio denoting the m o r t a l i t y table see LIFE TABLE.
number of people in a population who
are iU or have a specific disease com- m o v i n g average (MA) a form of aver-
pared with the number who are well. age comprising the means of successive
subsets of data within a longer set of ob-
morphometries n. the theory and servations. For example, a three-term
technique associated with the physical moving average of the sequence 1, 3, 2,
measurement of living organisms and 4, 3, and 5 would be (1 + 3 + 2)/3 = 2, (3 +
their component parts. Studies of size. 2 + 4)/3 = 3, (2 + 4 + 3)/3 = 3, and (4 + 3 +

221
moving-average model

5)/3 = 4. Often plotted graphically, a MSR symbol for MEAN-SQUARE RESID-

moving average generally is used to de- UAL.
scribe or discern the pattern in TIME-
MSSD symbol for MEAN-SQUARE SUCCES-
SERIES data and may be adjusted (or
SIVE DIFFERENCE.
smoothed) to allow for seasonal or cycli-
cal trends. For example, a behavioral M T M M 1. abbreviation for MULTITRAIT-
health researcher could plot moving av- MULTIMETHOD MATRIX. 2. abbreviation
erage values for the occurrence of a spe- for MULTITRAIT-MULTIMETHOD MODEL.
cific behavior in several individuals with
m u (symbol: ji) n. see POPULATION MEAN.
autism over a one-year period in order to
identify a pattern that is less erratic thanm u l t i c o l l i n e a r i t y n. in MULTIPLE RE-
if individual values were plotted without GRESSION, the state that occurs when
grouping and averaging. Also called several INDEPENDENT VARIABLES are ex-
rolling average; running average. tremely highly interrelated, making it
difficult to determine separate effects
moving-average model (MA model) on the DEPENDENT VARIABLE and thus
a statistical depiction of the pattern of yielding unstable REGRESSION ESTI-
change over time for a series of mean MATES with large STANDARD ERRORS. For
(average) values on a variable, as col- example, if a researcher includes predic-
lected in a LONGITUDINAL DESIGN. tors of self-esteem, self-efficacy, and self-
concept in an analysis with a dependent
moving medians a method of describ-
variable of achievement, multicol-
ing the pattern in a TIME SERIES in which
linearity most likely will be present.
different windows of n values are identi-
fied and the middle value from each is m u l t i d i m e n s i o n a l i t y n. 1. the qual-
used to represent CENTRAL TENDENCY, ity of a CONSTRUCT that cannot be ade-
with the process continuing until a quately described by measuring a single
SMOOTHED CURVE can be drawn to show trait or attribute. 2. the quality ofa scale,
the general trend in the data over time. test, or so forth that is capable of mea-
Moving medians are less susceptible to suring more than one DIMENSION of
the influence of OUTLIERS or shocks that a construct. For example, a psycho-
exist in the series than are MOVING AV- metrician may be interested in investi-
ERAGES. Also called running medians. gating the multidimensionality of a new
scale to measure cognitive functioning.
M o z a r t effect a temporaty increase in Compare UNIDIMENSIONALITY.
the affect or performance of research
participants on tasks involving spatial- m u l t i d i m e n s i o n a l scaling (MDS) a
temporal reasoning after listening to the SCALING method that represents per-
music of Austrian composer Wolfgang ceived similarities among stimuli by ar-
Amadeus Mozart (1756-1791). More ranging similar stimuli in spatial
generally, the term refers to the possibil- proximity to one another, while dispa-
ity that listening to certain types of rate stimuli are represented far apart
music enhances inherent cognitive from one another. For example, a re-
functioning. The notion of the Mozart searcher may use multidimensional scal-
effect has entered into popular culture to ing to assess the dimensions underlying
carty the as-yet-unsupported suggestion attributions for maternal behavior, as in
that early childhood exposure to classi- the illustration opposite.
cal music benefits mental development Multidimensional scaling is an alterna-
or inteUigence. tive to FACTOR ANALYSIS for dealing with
large matrices of data or stimuli.
MS symbol for MEAN SQUARE.
multidimensional unfolfling a
MSE symbol for MEAN SQUARED ERROR. MULTIDIMENSIONAL SCALING procedure

222
 multimodal distribution

for mapping, usually along two axes, amount of caffeine intake (e.g., 0 mg, 50
how a set of individuals rate their prefer- mg, 150 mg) affect rats' performance on
ences for or similarity to various charac- a particular maze task. When only one
teristics, siraations, or entities. For independent variable with multiple con-
example, a researcher could use multidi- ditions is involved, it is termed a SINGLE-
mensional unfolding to graph the pref- FACTOR MULTILEVEL DESIGN.
erences of a set of journal editors for
m u l t i l e v e l model see HIERARCHICAL
research article characteristics such as
LINEAR MODEL.
experimental srady, theoretical review,
qualitative inquity, and methodological m u l t i m e t h o d approach a design
application. Editors closer to one axis that uses more than one procedure for
might prefer theoretical or qualitative measuring the main characteristic or
articles, whereas those along the other construct of interest. For example, a re-
axis would favor publishing articles that searcher could use a multimethod ap-
are experimental or methodological. proach to understanding relationship
m u l t i d i m e n s i o n a l variable a con- satisfaction by simultaneously collect-
struct that cannot be fully described or ing data from a survey, asking one or
measured using a single underlying FAC- both of the partners to give their own
TOR. For example, intelligence could be self-report, and systematically observing
described as a multidimensional variable the degree of relatidnship satisfaction.
that involves verbal, quantitative, and See also MULTITRAIT-MULTIMETHOD
other aspects. MODEL.
m u l t i f a c t o r design see FACTORIAL DE- multimethod-multitrait model
SIGN. see MULTITRAIT-MULTIMETHOD MODEL.
m u l t i l e v e l design any research study m u l t i m o d a l d i s t r i b u t i o n a set of
involving two or more INDEPENDENT data in which there is more than one
VARIABLES with multiple conditions MODE or score that occurs most fre-
(levels) of investigation. An example quently, as shown in the generic graphic
is a study examining how time of day overleaf.
(e.g., morning, afternoon, evening) and For example, the ages of a sample of

Stable

Instinct
Knowledge Family upbringing
Financial
Necessity Age of mother
R o l e of ="PP'*
Personality partner

Internal External
Breastfeeding
Emotional support
Complications
Amount of time
spent with child
Desire to Amount of
have child time spent at
hospital

Temporary
multidimensional scaling

223
multinomial

(i.e., C ); three different ways of getting

an A in two classes and a B in the other
(e.g., AAB, ABA, BAA); and so on
through the six different ways to get an
A in one class, a B in another class, and a
C in the other class (e.g., ABC, ACB,
BAC, BCA, CAB, CBA). A similar process
for only two events is called a BINOMIAL
EXPANSION.

m u l t i n o m i a l logistic regression a
statistical procedure to characterize the
relationship between a set of predictors
college students would form a multi- and a single outcome with several cate-
modal distiibution since the largest num- gories, usually expressed as the odds of
ber of people are either 18, 19, or 20 years falling into one of the outcome catego-
old, with the remaining individuals aged ries (usually the most extreme) given
17 or 21 through 70. See also BIMODAL each individual predictor. For example,
DISTRIBUTION; UNIMODAL DISTRIBUTION. a multinomial logistic regression could
be conducted to assess the likelihood of
m u l t i n o m i a l adj. describing a measure- being included in a high-risk category
ment that can have more than two cate- on a four-category risk outcome (where
gories or outcomes. For example, a 1 = low and 4 = high) depending on pre-
professor assigning grades of A, B, C, D, dictors of smoking history, cholesterol
or F to students in his or her course is level, and weight.
making a multinomial decision, whereas
assigning grades of pass or fail would be m u l t i n o r m a l d i s t r i b u t i o n see MUL-
a binomial decision. TIVARIATE NORMAL DISTRIBUTION.

m u l t i n o m i a l d i s t r i b u t i o n a proba- m u l t i p l e baseline design an experi-

bility distribution that describes the mental approach in which two or more
theoretical distribution of n objects sam- behaviors are assessed to determine their
pled at random from a population of k initial, stable expression (i.e., baseline)
kinds of things with regard to the num- and then an intervention or manipula-
ber of each of the kinds that appears in tion is applied to one of the behaviors
the sample. By contrast, a BINOMIAL DIS- while the others are unaffected. After a
TRIBUTION involves just two variables, period, the manipulation is then applied
categories, or objects. to the next behavior while the remain-
ing behaviors are unaltered, and so forth
m u l t i n o m i a l expansion a mathe- until the experimental manipulation
matical procedure for finding the num- has been applied in sequential fashion
ber of different ways in which a set of to all of the behaviors in the design. In
more than two events can occur when successively administering a manipula-
raised to a certain power. For example, a tion to different behaviors after initial
multinomial expansion could be used to behaviors have been recorded, a multi-
determine the number of different ways ple baseline design allows for inferences
a student could achieve grades of A, B, or about the effect of the intervention.
C in three different courses. The out-
multiple-choice question see FIXED-
come would be equal to (A + B + C)"* = A''
A L T E R N A T I V E QUESTION.
+ B^ + C^ + 3A^B + 3A^C + 3B^A + 3B^C +
3C^A + 3C^B + 6ABC. That is, the out- multiple-choice test any test in which,
comes would correspond to a sequence for each item, the examinee chooses one
of all As (i.e., A^); all Bs (i.e., B^); aU Cs of several given alternatives as being cor-

224
 multiple indicators-multiple causes model

rect. A multiple-choice test contiasts with no training and an outcome variable of

an ESSAY TEST, in which respondents are reading achievement might include
free to state answers in their own words. multiple covariates of initial reading
See FIXED-ALTERNATIVE QUESTION. level, parents' education, and socioeco-
nomic staras. In this way, the researcher
multiple classification analysis could determine whether a phoneme-
(MCA) see MULTIPLE DISCRIMINANT
trained group obtains significantly
ANALYSIS. higher reading achievement than a
multiple comparisons a set of com- nontrained control group, after ruling
parisons made between samples to iden- out the effects of the multiple covariates
tify significant differences among their that could also be related to reading
mean values. Multiple comparisons are level.
generally done in a post hoc manner multiple discriminant analysis
(i.e., are unplanned) in order to keep the (MDA) a MULTIVARIATE Statistical pro-
TYPE I ERROR rate controlled at a pre- cedure for examining the relationship
specified level. Also called multiple between a set of more than two catego-
contrasts. ries representing an outcome variable
m u l t i p l e comparison test any of and a set of predictor variables. It often is
various statistical procedures used to fol- used to verify that the predictors (e.g.,
low up on a significant result from family histoty, self-esteem, number of
an ANALYSIS OF VARIANCE by determin- sleeping hours) are accurately classifying
ing which groups in particular differ individuals into the appropriate catego-
in their mean values. Examples include ries of the outcome variable (diagnosis
DUNCAN'S MULTIPLE RANGE TEST, the of depression). See DISCRIMINANT ANAL-
FISHER LEAST SIGNIFICANT DIFFERENCE YSIS. Also called multiple classifica-
TEST, the REGWQTEST, the SCHEFFETEST, tion analysis (MCA).
and TUKEY'S HONESTLY SIGNIFICANT DIF- m u l t i p l e i m p u t a t i o n a method for
FERENCE TEST. See also POST HOC C O M - addressing missing data in which several
PARISON. possible simulated values are inserted
multiple correlation coefficient into a data set to replace omitted values,
(symbol: R) a numerical index of the de- and then the mean and STANDARD DEVI-
gree of relationship between a particular ATION of the set are calculated to arrive
variable and two or more other vari- at an estimate to substitute for the miss-
ables. Its value ranges from -1 to +1, ing value. Multiple imputation is con-
with the former indicating a strong neg- sidered less biased than other missing
ative relationship and the latter a strong values procedures, such as LISTWISE DE-
positive relationship. Also called multi- LETION, PAIRWISE DELETION, and Single
ple R. IMPUTATION.

m u l t i p l e correlation coefficient multiple inflicators-multiple causes

squared see COEFFICIENT OF MULTIPLE model (MIMIC model) a MULTI-
VARIATE statistical procedure in which
DETERMINATION.
several INDEPENDENT VARIABLES are hy-
m u l t i p l e covariates two or more cor- pothesized to be predictors of one or
related variables that are included in an more LATENT VARIABLES (factors). For
ANALYSIS OF COVARIANCE tO help rulC example, a health researcher may hy-
out potential CONFOUNDS when exam- pothesize that measurements on several
ining group differences. For example, an background variablessuch as age, fam-
educational researcher analyzing the ily histoty of illness, socioeconomic sta-
covariance between an INDEPENDENT tus, and amount of exercise per week
VARIABLE of phoneme training versus are linked to a latent factor of heart

225
multiple linear regression

health, which in rarn has several indica- sessment Test (SAT) reading score, and
tors, such as systolic and diastolic blood SAT mathematics score. The multiple re-
pressure readings, pulse, and cholesterol gression equation would be
level. This multiple indicators-multiple
College GPA = intercept +
causes model would be depicted as four
B,(High School GPA) +
(predictor) measures having arrows
B2(SAT Reading) +
pointing toward the factor, which in
B3(SAT Mathematics) +
turn would have four lines emanating
prediction error
outward to represent its four indicators.
Also called multiple regression
m u l t i p l e linear regression see MUL-
model.
TIPLE REGRESSION.
multiple time series a set of measures
m u l t i p l e R see MULTIPLE CORRELATION
on two or more variables or individuals
COEFFICIENT.
taken over numerous occasions. For ex-
m u l t i p l e range test see DUNCAN'S ample, a researcher could examine the
MULTIPLE RANGE TEST. daily recordings of speech and interac-
tion for children with autism across a
multiple regression a statistical tech-
one-year period. The use of several paral-
nique for examining the linear relation-
lel-running TIME SERIES to gather data in
ship between a continuous DEPENDENT
a LONGITUDINAL DESIGN Is Called a mul-
VARIABLE and a set of two or more INDE-
tiple time-series design, whereas the
PENDENT VARIABLES. It Is Often used to
MULTIVARIATE ANALYSIS used to identity
predict a single outcome variableftoma
relationships among the variables in
set of predictor variables. For example,
such data is called multiple time-series
an educational psychology researcher
analysis. Alternatively, the researcher
could use multiple regression to predict
could use an INTERRUPTED TIME-SERIES
college achievement (e.g., grade point
DESIGN to see if the pattern of data for
average) from the variables of high
the two variables changed noticeably
school grade point average. Scholastic
after the introduction of an intervention
Assessment Test (SAT) reading score, SAT
during the recording period. Also called
mathematics score, and SAT writing
multivariate time series.
score. When a single predictor and a sin-
gle outcome are involved the process is m u l t i p l e t test a T TEST used to make
known as LINEAR REGRESSION. Also more than one comparison between
called multiple linear regression. pairs of group MEANS. The increased
number of calculations results in an in-
m u l t i p l e regression equation a RE-
creased chance of making a TYPE I ERROR
GRESSION EQUATION generated from the
(i.e., of concluding there is a significant
process of MULTIPLE REGRESSION, which
difference between groups when there is
represents the relationship between a
not).
single outcome variable and the best
LINEAR COMBINATION of a Set of predic- multiplication rule a rule stating that
tors. In the equation an outcome score the JOINT PROBABILITY of two indepen-
(y) is presented as a function of a con- dent events occurring together or in suc-
stant called an INTERCEPT, plus the cession is equal to the probability of the
products of a REGRESSION COEFFICIENT first event times the probability of the
(B) times the values of various predictor second event. For example, the multipli-
variables (x), plus some prediction error. cation rule would indicate that the prob-
For example, assume a researcher is ability of drawing a heart followed by
studying the relationship of college the probability of drawing a spade is
grade point average (GPA) to the predic- equal to 13/52 x 13/52 = .25 x .25 =
tors of high school GPA, Scholastic As- .0625. Also called and rule; multipli-
 multitrait-multimethod matrix

cation law; multiplicative law. sure a sample of the subunits in any cho-
Compare ADDITION RULE. sen unit to avoid uneconomically mea-
suring all of them. When two grouping
m u l t i p l i c a t i v e model a description sets are involved, the process is also
of the effect of two or more predictor known as two-stage sampling, when
variables on an outcome variable that al- three sets are involved it is also called
lows for INTERACTION EFFECTS among three-stage sampling, and so on.
the predictors. This is in contrast to an
ADDITIVE MODEL, which sums the indi- multistate model a statistical repre-
vidual effects of several predictors on sentation of several possible stages for an
an outcome. For example, a health re- event, either over the course of time or
searcher could use a multiplicative across a set of individuals; such models
model to examine the interaction effect are often used to understand disease pro-
of number of cigarettes smoked per day gression or health promotion. For exam-
and length of smoking habit on the ple, a researcher may hypothesize a
onset of cancer; the results might be multistate model of behavior change,
compared to those obtained from an ad- involving various stages from pre-
ditive model that examines the separate contemplation, when an individual is
effects of amount of cigarettes and not thinking of making a change,
length of a smoking habit on cancer through to maintenance, when an indi-
onset. vidual has successfully made a behavior
multisite study research and data col- change (e.g., quitting smoking) for more
lection conducted over several locations than six months. A MARKOV C H A I N is
or geographical areas; this allows greater similar to a multistate model but in-
generalization offindingsthan research volves stages with some degree of de-
conducted in a single place. Also called pendence among them.
cooperative study.
multitrait-multimethod matrix
multistage s a m p l i n g a technique in (MTMM) a matrix showing correlations
which samples are drawn first from among two or more measurement tech-
higher order groupings (e.g., states) and niques used to assess two or more con-
then from successively lower level structs or traits, as obtained from a
groupings (e.g., counties within states, MULTITRAIT-MULTIMETHOD MODEL. It
towns within counties) in order to avoid includes correlations among the same
the necessity of having a SAMPLING traits with different methods (i.e.,
FRAME for the entire population. That is, monotrait-heteromethod) and among
if subunits within a selected unit give different traits with the same method
similar results, one can select and mea- (i.e., heterotrait-monomethod). The for-

Peer ratings Association test

A, Bl Cl Dl AI Bl C, Dl
Peer ratings
Courtesy Al (.82)
Honesty B, .74 (.80)
Poise Cl .63 .65 (.74)
Scliool drive D. .76 .78 .65 (.89)

Association test
Courtesty Al .13 .14 .10 .14 (.28)
Honesty B, .06 .12 .16 .08 .27 (.38)
Poise Cl .01 .08 -10 .02 .19 .37 (.42)
School drive D, .12 .15 .14 .16 .27 .32 .18 (.36)

multitrait-multimethod matrix

227
multitrait-multimethod model

mer are expected to be the largest, thus ing assumptions about their underlying
demonstrating CONVERGENT VALIDITY, functional relationships. For example, a
whereas the latter are expected to be researcher studying the relationship be-
smallest, demonstrating DISCRIMINANT tween achievement motivation and aca-
VALIDITY. Consider the example below, demic performance could use MARS to
in which VALIDITY DIAGONALS are more accurately model the inverted U-
shown in italics and RELIABILITY DIAGO- shaped pattern expected to emerge. That
NALS in parentheses. is, individuals who have very low
HETEROTRAIT-MONOMETHOD COEFFI- achievement motivation would be ex-
CIENTS as large or larger than monotrait- pected to have relatively low perfor-
heteromethod coefficients indicate that mance scores, which would increase as
some method variance is present, sug- achievement motivation increases. After
gesting that participants are responding a certain point, however, performance
similarly across different traits simply scores could be expected to decline for
because they are being assessed with the individuals who have too much achieve-
same method. ment motivation, thus forming a non-
linear pattern. See SPLINE FUNCTION.
multitrait-multimethod model
(MTMM) a procedure for examining m u l t i v a r i a t e analysis 1. a set of sta-
CONSTRUCT VALIDITY that assesses the tistical procedures for studying the rela-
correlations among two or more charac- tionships between one or more
teristics where these are each measured predictors and several outcome or DE-
in two or more ways. For example, a re- PENDENT VARIABLES. Examples include
searcher studying self-concept and CANONICAL CORRELATION ANALYSIS for
achievement as measured by both self- assessing the relationships among two
reports and teacher .evaluations could sets of variables; FACTOR ANALYSIS for as-
use a multitrait-multimethod model to sessing the relationships among a large
evaluate the associations between the set of measures and a small set of under-
self-reports and teacher evaluations on lying factors; MULTIVARIATE ANALYSIS
each one of the characteristics. The set of OF VARIANCE and MULTIVARIATE ANALY-
correlations resulting from such an anal- SIS OF COVARIANCE for assessing poten-
ysis is displayed in a MULTITRAIT- tial group differences on several
MULTIMETHOD MATRIX. In the related dependent variables; and STRUCTURAL
monotrait-multimethod model, a re- EQUATION MODELING, which examines
searcher examines a single characteristic a theoretically based pattern of relation-
using several different methods. Also ships among multiple independent, de-
called multitrait-multimethod an- pendent, and even mediating variables.
alysis. Also called multivariate statistics.
Compare UNIVARIATE ANALYSIS. 2. more
m u l t i v a r i a t e adj. consisting of or oth-
erwise involving a number of distinct generally, any procedure for under-
variables. For example, a multivariate standing any large set of variables,
srady of ability could involve multiple whether dependent or independent.
measures of intelligence and achieve-
m u l t i v a r i a t e analysis of co-
ment. Compare UNIVARIATE. See also
variance (MANCOVA) a statistical
BIVARIATE. procedure for assessing possible group
m u l t i v a r i a t e adaptive regression differences on a set of outcome or DE-
spline (MARS) a type of nonpara- PENDENT VARIABLES, after taking into ac-
metric REGRESSION ANALYSIS that exam- count the scores on one or more
ines NONLINEAR aS Well aS LINEAR COVARIATES. For example, a researcher
relationships among dependent and in- could conduct a multivariate analysis of
dependent variables but without mak- covariance to assess whether two groups
 multivariate normal distribution

of participants in a teaching style study together. For example, a mathematical

differ significantiy on a set of achievement health researcher could examine the
variablessuch as quiz scores, homework multivariate distribution function for
scores, exam scores, and project scores the JOINT PROBABILITY of taking little
after taking into account the initial grade exercise, having high cholesterol, and
point average of each of the participants. having heart trouble in a sample of mid-
It is an extension of the univariate ANAL- dle-aged individuals. A bivariate distri-
YSIS OF COVARIANCE, which examines a bution function is a form of multivariate
single dependent variable. distribution function specific to two
multivariate analysis of variance variables.
(MANOVA) a statistical procedure for m u l t i v a r i a t e generalizability the-
assessing possible group differences on a ory a body of theory concerning the
set of outcome or DEPENDENT VARI- identification and control of one or
ABLES. For example, a researcher could more sources of error over several vari-
conduct a multivariate analysis of vari- ables or measures. For example, a re-
ance to assess whether a group of partici- searcher may wish to assess the
pants who receive a new educational RELIABILITY of Several measures of quan-
method differ significantly from an- titative achievement, such as number of
other group of participants who are courses taken, mathematics score on the
taught with a traditional method on a Scholastic Assessment Test or Graduate
set of achievement variables, such as Record Examination, and score on a
quiz scores, homework scores, exam quantitative skills test. He or she could
scores, and project scores. It is an exten- do so by identifying several possible
sion of the univariate ANALYSIS OF VARI- sources of error, known as FACETS, and
ANCE, which examines a single then examining the different VARIANCE
dependent variable. and COVARIANCE components for the
various facets. See GENERALIZABILITY
m u l t i v a r i a t e d i s t r i b u t i o n the DIS-
TRIBUTION of scores obtained on several
THEORY.
variables, as in the following example. m u l t i v a r i a t e kurtosis measure an
index of the degree of KURTOSIS in a dis-
Exercise Cholesterol Blood pressure
tribution of scores on several variables. It
Participant (hrs/day) (mg/dL) {mm/Hg) indicates the extent to which such a dis-
A 1 233 160/100
tribution differs from NORMALITY with
respect to the presence of high peaks in
B 5 175 100/75
the data (points at which many people
C 3 192 105/80 or items obtain particular scores). An ex-
D 1 216 122/85 ample is Mardia's multivariate kurtosis
measure: A value of zero indicates nor-
E 1 221 135/86
mality, and increasingly larger values in-
dicate increasingly greater peakedness or
A health researcher may want to exam- kurtosis.
ine these scores on level of exercise, cho-
lesterol level, and blood pressure in a multivariate normal distribution
sample of individuals at risk for heart a pattern of values on several variables in
disease. Compare UNIVARIATE DISTRIBU- which a graph of the data forms a bell-
TION. shaped NORMAL DISTRIBUTION. Evi-
dence for a multivariate normal distribu-
m u l t i v a r i a t e d i s t r i b u t i o n f u n c - tion may be obtained using a QUANTILE-
t i o n a mathematical function giving QUANTILE PLOT, in whlch MAHALANOBIS
the probability that certain values on DISTANCE scores are plotted against a set
several RANDOM VARIABLES will occur of points from a normal distribution. If

229
multivariate normality

the plot shows a series of points along a when comparing the mean values be-
diagonal line, indicating a lack of any tween two groups on two or more DE-
discrepant scores (i.e., any MULTI- PENDENT VARIABLES. It Is an extension of
VARIATE OUTLIERS), then the data have a the T DISTRIBUTION used when testing
multivariate normal distribution. Also for potential differences in situations
called multinormal distribution. that involve more than one outcome.
See also MULTIVARIATE NORMALITY. For example, a researcher might consult
a multivariate t distribution in order to
m u l t i v a r i a t e n o r m a l i t y the situa-find the CRITICAL VALUE that would in-
tion in which the values for a set of vari- dicate whether treatment and control
ables have an even distribution, with groups differ in their physical health
most scores falling in the middle of the and psychological well-being outcomes.
range and a smaller number of high and
low scores. In other words, a set of scores m u l t i v a r i a t e test any of various statis-
demonstrating multivariate NORMALITY tical procedures involving two or more
follows a MULTIVARIATE NORMAL DIS- outcome or DEPENDENT VARIABLES. An
TRIBUTION and lacks OUTLIERS or ex- example is the MULTIVARIATE ANALYSIS
treme scores. OF VARIANCE, whlch extends the ANALY-
SIS OF VARIANCE examining one de-
m u l t i v a r i a t e outlier a data point pendent variable to include additional
whose values on several variables are dependent variables, FACTOR ANALYSIS
very different from the others in a set, and MULTIPLE REGRESSION sometlmes
such that it may BIAS the results of a sta- are considered multivariate tests as well,
tistical analysis, particularly one based even though the former involves a sin-
on an ASSUMPTION of NORMALITY. For gle set of variables that are not necessar-
example, an individual participating in ily independent or dependent and the
a study who obtained an IQ score of 150 latter allows for multiple INDEPENDENT
(when the mean IQ is 100 and the stan- VARIABLES but only one outcome.
dard deviation is 15), a performance
m u l t i v a r i a t e time series see MULTI-
score of 100 (when the mean is 50 and
PLE TIME SERIES.
the standard deviation is 20), and a so-
cial skills score of 1 (when the mean is 10 m u l t i v a r i a t e t test see HOTELLING'S T^
and the standard deviation is 5) would TEST.
be a multivariate outlier, MAHALANOBIS m u n d a n e realism the extent to which
DISTANCE, COOK'S DISTANCE, and Other an experimental siraation resembles a
statistical procedures may be used to de- real-life siraation or event. See also EX-
termine whether multivariate outliers PERIMENTAL REALISM.
are present in a data set. Compare
UNIVARIATE OUTLIER. m u t u a l l y exclusive events 1. two or
more events that have no common ele-
m u l t i v a r i a t e research a study con- ments, that is, they are disjoint (see DIS-
ducted to simultaneously assess the rela- JOINT SETS). 2. in probability theory, two
tionships among multiple DEPENDENT or more events that cannot co-occur:
VARIABLES and INDEPENDENT VARI- The occurrence of one precludes the si-
ABLES. Compare UNIVARIATE RESEARCH. multaneous or subsequent occurrence of
the other(s). For example, the alterna-
m u l t i v a r i a t e statistics see MULTI- tives "heads" and "tails" in a single toss
VARIATE ANALYSIS. of a coin are mutually exclusive events.
m u l t i v a r i a t e t d i s t r i b u t i o n the dis- M V U E abbreviation for minimum vari-
tribution of possible values for the statis- ance unbiased estimator. See EFFICIENT
tic obtained from HOTELLING'S T^ TEST ESTIMATOR.
 Nn
n symbol for the number of scores or ob- experiments. Also called naturalistic
servations obtained from a particular ex- design; naturalistic research.
perimental condition or subgroup.
naturalistic observation data collec-
N symbol for the total number of cases tion in a field setting, without labora-
(participants) in an experiment or study. toty controls or manipulation of
variables. These procedures are usually
naive p a r t i c i p a n t a participant who carried out by a trained observer, who
has not previously participated in a par- watches and records the everyday be-
ticular research study and has not been havior of participants in their natural
made aware of the experimenter's hy- environments. Examples of naturalistic
pothesis. observation include an ethologist's
study of the behavior of chimpanzees
narrative analysis atypeof QUALITA- and a developmental psychologist's ob-
TIVE ANALYSIS in which a researcher col- servation of playing children. See also
lects and examines stories from SYSTEMATIC NATURALISTIC OBSERVA-
individuals about a variety of concrete TION. Compare ANALOGUE OBSERVA-
life siraationsranging from first ro- TION; SELF-MONITORING OBSERVATION;
mantic involvements to larger issues, STRUCTURED OBSERVATION.
such as divorce, aging, and life satisfac-
tion. The goal is to understand how indi- n a t u r a l l o g a r i t h m (In) see LOGA-
viduals experience certain events, RITHM.
structure them into coherent sequences, nay-saying n. answering questions nega-
and give them subjective meaning. Also tively regardless of their content, which
called narrative inquiry; narrative can distort the results of surveys, ques-
research. tionnaires, and similar instmments. Also
called response deviation. Compare
n a t u r a l experiment the study of a Y E A - S A Y I N G .
naturally occurring siraation as it un-
folds in the real world. The researcher NCE abbreviation for NORMAL CURVE
does not exert any influence over the sit- E Q U I V A L E N T .
uation but rather simply observes indi-
viduals and circumstances, comparing N=l design see SINGLE-CASE DESIGN.
the current condition to some other nearest neighbor see SINGLE-LINKAGE
condition. For example, an investigator C L U S T E R I N G .
might evaluate the influence of a new
community policing program by observ- negative b i n o m i a l d i s t r i b u t i o n a
ing neighborhood activities after it has T H E O R E T I C A L D I S T R I B U T I O N Of d i s c r e t e
been implemented and comparing the values that describes the number of tri-
outcome to that for neighborhoods in als that will occur before a success in a
which the policy has not yet been imple- sequence of n BERNOULLI TRIALS, with a
mented. Since such real-life events can- given likeliness of failure and success
not be manipulated or prearranged, across the sequence. Also called Pascal
natural experiments are QUASI-EX- distribution. Compare POSITIVE BINO-
PERIMENTAL DESIGNS rather than true MIAL DISTRIBUTION.

231
negative case analysis
n e g a t i v e case a n a l y s i s see D E V I A N T
CASE ANALYSIS.
n e g a t i v e c o r r e l a t i o n a relationship
between two variables i n which the
value of one variable increases as the
value of the other decreases. For exam-
ple, in a study about babies crying and
being held, the discovery that those who
are held more tend to cty less is a nega-
tive correlation. Also called i n d i r e c t
c o r r e l a t i o n ; inverse c o r r e l a t i o n .
Compare POSITIVE C O R R E L A T I O N .
n e g a t i v e p r e d i c t i v e p o w e r the pro-
portion of cases correctly identified by a
test as not having a particular condition,
indicating the probability that someone
predicted not to have a condition by this n e s t e d c a s e - c o n t r o l s t u d y a research
test in fact does not have that condition.
For example, the negative predictive
power of a diagnostic test for depression
would be determined by dividing the
percentage of people whose results cor-
rectly show they do not have the disor-
der (i.e., valid negatives) by the
percentage of all people whose results
are negative (i.e., both valid negatives
and false negatives). Compare POSITIVE
PREDICTIVE POWER.
n e g a t i v e r e l a t i o n s h i p an association
in which one variable decreases as the (controls) into another. Although an ef-
other variable increases, or vice versa.
Also called inverse r e l a t i o n s h i p . See
also INDIRECT RELATIONSHIP. Compare
DIRECT RELATIONSHIP.
n e g a t i v e s k e w see SKEWNESS.
n e g a t i v i s t i c - s u b j e c t r o l e behavior
adopted by a participant in a research n e s t e d d e s i g n see HIERARCHICALLY
study who intentionally tries to be
"bad," responding in a contrary or ran-
dom fashion to obstruct the stuciy. Com- n e s t e d f a c t o r see NESTING.
pare APPREHENSIVE-SUBJECT ROLE; n e s t e d
FAITHFUL-SUBJECT ROLE; GOOD-SUBJECT
ROLE.
n e s t i n g n. i n an experimental design,
n e i g h b o r h o o d c o n t r o l a group of in-
dividuals selected from the same region
or area as a targeted study group to serve
as a comparison group. The assumption
is that such a group will share similar ex-
232
periences, risk exposure, and other rele-
vant characteristics to the study group,
given their geographical proximity.
Used especially in epidemiological or
clinical research, neigfiborhood controls
generaUy are chosen according to some
rule-based procedure and matched to
members of the target group on certain
attributes, such as age or sex. For exam-
ple, a researcher investigating cancer
might go to evety second house on the
same city block as each person in the
study group to identify similar individu-
als who do not have the disease to be-
come part of a neighborhood control.
Also called c o m m u n i t y c o n t r o l .
Compare DISTAL C O N T R O L .
method in which individuals from a de-
fined COHORT are sampled and divided
into groups according to whether or not
they have a disease or other condition.
It is thus a combination of a CASE-
C O N T R O L STUDY and a cohort srady (see
LONGITUDINAL DESIGN). For example, a
researcher investigating a group of Viet-
nam War veterans (the cohort) might
place participants who have attempted
suicide (cases) into one subgroup and
those with little to no suicidal ideation
ficient strategy when the budget is lim-
ited or the condition or outcome of
interest is rare, a nested case-control
study has such disadvantages as greater
risk of confounding (see C O N F O U N D )
and reduced ability to generalize or
make causal inferences.
NESTED DESIGN.
m o d e l see HIERARCHICAL
MODEL.
the appearance of the levels of one factor
(the nested factor) only within a single
level of another factor. For example,
classrooms are nested within a school
because each specific classroom is found
 nominal data

only within a single school; similarly, proximations to the root of a function.

schools are nested within school dis- Also called Newton method. [Isaac
tricts. See HIERARCHICALLY NESTED DE- Newton (1642-1727), British physicist
SIGN. and mathematician; Joseph Raphson
(1648-1715), British mathematician]
n e t w o r k analysis the study of the re-
lations among sampling units (e.g., indi- Neyman-Pearson theory an ap-
viduals) within an interconnected group proach to formulating two competing
of such units (e.g., a friendship network) hypotheses (the NULL HYPOTHESIS and
and the implications of these networks an ALTERNATIVE HYPOTHESIS) and iden-
for the system in which they occur. In tifying appropriate statistical procedures
organizational contexts, for example, for choosing between them. It focuses
network analysis involves identifying upon identifying CRITICAL REGIONS,
patterns of communication, influence, minimizing errors in estimation, and
liking, and other interpersonal behav- obtaining an appropriate balance in the
iors and attitudes among employees and probability of committing TYPE I ERRORS
quantifying them in statistical and and TYPE II ERRORS. IJerzy Neyman
graphical models. Properties of systems (1894-1981), Russian-born U.S. statisti-
are assumed to be emergent, that is, not cian; Egon S. Pearson (1895-1980), Brit-
immediately predictable from a knowl- ish statistician]
edge of networks among individuals.
N F l abbreviation for normed fit index.
n e t w o r k s a m p l i n g a sampling tech- See BENTLER-BONETT INDEX.
nique in which members of a srady
group recruit their peers to participate in N H abbreviation for NULL HYPOTHESIS.
the research. It is simUar in process, ben- NHST abbreviation for null hypothesis
efits, and drawbacks to SNOWBALL SAM- SIGNIFICANCE TESTING.
PLING but distinct in that new
participants are acquired directly by ex- NNT abbreviation for NUMBER NEEDED
isting ones according to certain rules TO TREAT.
(e.g., no more than four recraits per per-
N-of-1 design see SINGLE-CASE DESIGN.
son) and that various incentives are pro-
vided for participation and recruitment. no-intercept model see REGRESSION
THROUGH THE ORIGIN.
Newman-Keuls m u l t i p l e compari-
son test a statistical procedure in n o m i n a l adj. denoting a number that
which sets of means are compared fol- indicates membership in a categoty,
lowing a significant result from an ANAL- such as coding political party affiliations
YSIS OF VARIANCE. The mean values of all with a 0 for Democrat, a 1 for Republi-
experimental groups are arranged in can, and a 2 for Independent. A nominal
order of size, formed into pairs, and the value is no indication of rank order or
differences between members of each magnitude. Compare CARDINAL; ORDI-
pair evaluated against a critical value NAL.
known as the STUDENTIZED RANGE STA-
TISTIC. Also called Newman-Keuls n o m i n a l data numerical values that
multiple range test; Newman- represent membership in specific cate-
Keuls test. [D. Newman, British statisti- gories. For example, the categoty male
cian; M. Keuls, Dutch horticulraralist] could be labeled 0 and the categoty fe-
male labeled 1, and each person within
Newton-Raphson method a tech- the population of interest (e.g., a partic-
nique for solving equations that is based ular town) assigned the number corre-
on a process of linear approximation, sponding to their sex. Nominal data are
that is, finding progressively closer ap- similar to CATEGORICAL DATA, and the

233
nominal scale

two terms are often used interchange- that characterize the average person or
ably. case. Compare IDIOGRAPHIC
n o m i n a l scale a sequence of numbers nonadditive adj. describing values or
that do not indicate order, magnitude, measurements that cannot be meaning-
or a true zero point but rather identity fully summarized through addition be-
items as belonging to mutually exclusive cause the resulting total does not
categories. For example, a nominal scale correctly reflect the underlying proper-
for the performance of a specific group ties of and associations between the
of people on a particular test might arbi- component values. For example, if two
trarily use the number 1 to denote pass variables a and b interact to influence
and the number 2 to denote fail. Since another variable y, the addition of the
the numbers represent categoty labels, separate effects of a and b will not equal
they cannot be manipulated mathemat- the total effect since the contribution of
ically or otherwise quantitatively com- the interaction needs to be included.
pared. A nominal scale is one of four Compare ADDITIVE.
types of measurement scale, the others
being an ORDINAL SCALE, an INTERVAL noncentral d i s t r i b u t i o n a DISTRIBU-
SCALE, and a RATIO SCALE. See also CATE- TION in which the NONCENTRALITY PA-
GORICAL SCALE. RAMETER is not equal to zero: In
statistical SIGNIFICANCE TESTING, such a
n o m i n a l variable a variable whose . distribution is obtained when the NULL
possible values are unordered categories HYPOTHESIS under test is false. The
or labels. For example, choice of college noncentral version of a distribution has
major is a nominal variable. a different mean, SKEWNESS, and VARI-
nomological n e t w o r k a conceptual ANCE (among other properties) from its
network: a broadly integrative theoreti- corresponding central distribution, as
cal framework' that identifies the key well as a larger proportion of numbers
constructs associated with a phenome- beyond the CRITICAL VALUES. For exam-
non of interest and the associations ple, the CHI-SQUARE DISTRIBUTION, F
DISTRIBUTION, a n d T D I S T R I B U T I O N a l l
among these constructs. For example,
have corresponding noncentral versions
psychopathy is a complex notion in-
signified by a noncentrality parameter
volving a significant nomological net-
that is not equal to zero.
work of knowledge and speculations
about components, causes, correlates, noncentrality parameter in many
and consequences as well as their inter- PROBABILITY DISTRIBUTIONS USed i n SIG-
relationships and means of measure- NIFICANCE TESTING, a PARAMETER that
ment or evaluation. has a value different from zero when the
nomological v a l i d i t y the degree to NULL HYPOTHESIS under test is false.
which a measure assesses the specific This parameter is important in deter-
construct it is designed to assess, as for- mining the POWER of a statistical proce-
mulated hom the NOMOLOGICAL NET- dure. See NONCENTRAL DISTRIBUTION.
WORK for the construct being measured.
n o n d i r e c t i o n a l hypothesis a hy-
See CONSTRUCT VALIDITY.
pothesis that one experimental group
nomothetic adj. relating to the formu- will differ from another without specifi-
lation of general laws as opposed to the cation of the expected direction of this
study of the individual case. A effect. For example, a researcher might
nomothetic approach involves the hypothesize that college students will
study of groups of people or cases for the .perform differently from elementary
purpose of discovering those general school students on a memory task with-
and universally valid laws or principles out predicting which group of students

234
 nonlinear interpolation

will perform better. Also called equivalent control-group design.

nondirectional alternative hy- Compare EQUIVALENT-GROUPS DESIGN.
pothesis; two-tailed (alternative) nonexperimental adj. of a research
hypothesis. Compare DIRECTIONAL project, lacking manipulation of INDE-
HYPOTHESIS. PENDENT VARIABLES by a researcher or
n o n d i r e c t i o n a l test a statistical test of RANDOM ASSIGNMENT of participants
an experimental hypothesis that does to treatment conditions, as in OBSERVA-
not specify the expected direction of an TIONAL STUDIES and QUASI-EXPERI-
effect or a relationship. Also called MENTAL DESIGNS for example.
nondirectional alternative hy- nonexperimental research see
pothesis test; nondirectional hy- QUASI-EXPERIMENTAL RESEARCH.
pothesis test; two-tailed test. See
NONDIRECTIONAL HYPOTHESIS. Com- nonignorable missing value a piece
pare DIRECTIONAL TEST. of information whose loss from a data
set is not random and consequently
nondirective i n t e r v i e w see U N - needs to be accounted for in parameter
STRUCTURED INTERVIEW. ESTIMATION and other statistical analy-
nonequivalent-groups design a ses. See IMPUTATION. .
QUASI-EXPERIMENTAL DESIGN in which n o n i n f o r m a t i v e p r i o r in BAYESIAN
the responses of a treatment group and a analysis, a theoretical set of feasible val-
comparison group are compared on ues for a parameter or unknown charac-
measures coUected at the beginning and teristic of interest in a population, where
end of the research: In psychology and there is no specific or definite informa-
other social sciences, these designs often tion about the parameter but certain val-
involve self-selection, in which the ues are assumed to be more likely to
members of the treatment group are occur than others. One of two main
those who volunteer or otherwise seek types of PRIOR DISTRIBUTION, it is used
the treatment whereas the comparison to calculate an associated POSTERIOR
group members do not. For example, DISTRIBUTION. Compare INFORMATIVE
a researcher might wish to study the rel- PRIOR.
ative value of medication for patients
participating in psychotherapy for de- nonlinear adj. describing any relation-
pression. He or she might identity two ship between two variables (x and y) that
groups of individualsone currently re- cannot be expressed in the form y = a +
ceiving therapy only and the other re- bx, where a and b are numerical con-
ceiving therapy plus a certain dragand stants. The relationship therefore does
collect PRETEST data on their severity of not appear as a straight line when de-
depression. The investigator would then picted graphically. Compare LINEAR.
collect POSTTEST data on the severity of n o n l i n e a r equation an equation in
depression at some subsequent point which the variables are related in an in-
(e.g., 12 weeks later) and compare the teractive fashion (e.g., x^X2) or are raised
outcomes of the two groups in order to to powers greater than 1. For example, y
estimate the efficacy of the antidepres- = a + b^x^ + ^2*2^ is ^ nonUnear equation
sant medication. Since participants are and will form a curve when depicted
not assigned to conditions at random, graphically. Compare LINEAR EQUA-
the two groups are likely to exhibit TION.
preexisting differences on both mea-
sured and unmeasured factors that must n o n l i n e a r f u n c t i o n see LINEAR FUNC-
be taken into account during statistical TION.
analyses. Also called nonequivalent n o n l i n e a r i n t e r p o l a t i o n a method
comparison-group design; non- of constracting new data points within a

235
nonlinear model

known range of such points that in- n o n m e t r i c adj. describing data that are
volves use of a NONLINEAR EQUATION. N O M I N A L or ORDINAL, as opposed to IN-
Compare LINEAR I N T E R P O L A T I O N . See TERVAL DATA or RATIO DATA. Such data
INTERPOLATION. cannot be precisely quantified. Exam-
ples include yes/no answers or a list
n o n l i n e a r m o d e l any model that at-
ranking individuals on some attribute.
tempts to relate the values of an out-
come or dependent variable to the n o n m e t r i c s c a l i n g an analytic tech-
explanatory or independent variables nique i n which similarities among data
using a NONLINEAR E Q U A T I O N . Compare points are determined based on a rela-
LINEAR MODEL. tive ordering of their indexed difference
values. A variation of standard (metric)
n o n l i n e a r r e g r e s s i o n a procedure MULTIDIMENSIONAL SCALING, which
for analyzing the relationship between uses the actual quantified spatial dis-
an INDEPENDENT VARIABLE (x) and a DE- tances among points (i.e., DISSIMILAR-
PENDENT VARIABLE (y) where the RE- ITY COEFFICIENTS and SIMILARITY
GRESSION EQUATION involves exponen- COEFFICIENTS), nonmetric scaUng exam-
tial forms of X. That is, the changes in y ines only the rankings of those distance
are not consistent for unit changes in values to identify similarities.
the X variable(s) but are a function of the
particular values of x. For example, a n o n n o r m a t i v e adj. not conforming to
nonlinear regression model is given by y or reflecting an established N O R M . For
= a + b^Xi + ^2^1^ + e, where a indicates example, an individual test score well
the place where the line of BEST FIT above or well below the mean for a class-
crosses the y-axis, b^ indicates the num- room of students is a nonnormative
ber of units that y changes when x is value. Compare N O R M A T I V E .
changed by 1 point, and e is error. Also
called c u r v i l i n e a r regression. Com- n o n p a r a m e t r i c adj. describing any an-
pare LINEAR REGRESSION. See REGRES- alytic method that does not involve
SION ANALYSIS. making ASSUMPTIONS about the data of
interest. Compare P A R A M E T R I C
n o n l i n e a r r e l a t i o n s h i p an associa-
n o n p a r a m e t r i c h y p o t h e s i s test see
tion between two variables in which the
N O N P A R A M E T R I C TEST.
direction and rate of change flucraate.
That is, the amount of change i n a D E - n o n p a r a m e t r i c r e g r e s s i o n a form of
PENDENT VARIABLE (y) varies as a func- REGRESSION ANALYSIS in which the rela-
tion of the particular value or level of the tionship between an outcome or D E -
INDEPENDENT VARIABLE (x). Compare PENDENT VARIABLE and one or more
LINEAR RELATIONSHIP. predictors or I N D E P E N D E N T V A R I A B L E S is
analyzed without the assistance of a pre-
n o n l i n e a r t r a n s f o r m a t i o n a TRANS- existing model. KERNEL REGRESSION is
F O R M A T I O N of a data set that uses a func- an example. In contrast to traditional re-
tion to change the linear relationship gression, i n which the structure of the
between variables. For example, a nor- REGRESSION EQUATION is known and
malizing transformation creates a new only the REGRESSION COEFFICIENTS are
set of scores that approximate a N O R M A L estimated from the data set, non-
DISTRIBUTION from the original non- parametric regression requires that both
normal data set, and a L O G A R I T H M I C the equation and coefficients be deter-
TRANSFORMATION may be used to con- mined from the data.
vert raw data into a form that more
closely matches the ASSUMPTIONS re- n o n p a r a m e t r i c s t a t i s t i c s statistical
quired in particular statistical analyses. procedures in which the narare of the
Compare LINEAR T R A N S F O R M A T I O N . data being analyzed is such that certain

236
 norm

common assumptions about the distri- nonrecursive m o d e l a set of relation-

bution of the attribute (or attributes) in ships in which a particular variable is
the population being tested, such as nor- sometimes a cause and sometimes an ef-
mality and homogeneity of variance, are fect, thus suggesting there is a reciprocal
not necessary or applicable. Also called relationship between that variable and
distribution-free statistics. Com- others in the model. For example, high
pare PARAMETRIC STATISTICS. ; job satisfaction may lead to increased
pay and job responsibility, which in
nonparametric test a type of HY- rarn may further enhance job satisfac-
POTHESIS TEST that does not make any tion. Compare RECURSIVE MODEL.
assumptions (e.g., of NORMALITY or HO-
MOGENEITY OF VARIANCE) about' the nonresponse n. a participant's failure
population of interest. Nonparametric to answer one or more survey, question-
tests generally are used in situations in- naire, or test items or to provide a mea-
volving NOMINAL or ORDINAL data. Also surement on some study variable. If data
called distribution-free test; non- values are MISSING COMPLETELY AT RAN-
parametric hypothesis test. Com- DOM, then nonresponse does not distort
pare PARAMETRIC TEST. results. In most research situations,
however, nonresponse is not a purely
n o n p r o b a b i l i t y s a m p l i n g non- random phenomenon; participants who
random selection: any process of choos- do not answer questions differ in some
ing a subset of participants or cases from important, systematic way from those
a larger population in which it is impos- who do answer. The basic method for
sible to precisely determine each unit's compensating for such nonresponse (or
likelihood of being selected. Examples nonresponder) bias involves estimating
include CONVENIENCE SAMPLING and the probability that each sample case
QUOTA SAMPLING. Nonprobability sam- will become a respondent. See also MISS-
pling makes it difficult to determine ING VALUES PROCEDURE.
how well the target population is repre-
sented by the subset, thus limiting the nonresponse rate see REFUSAL RATE.
GENERALIZABILITY offindings.Compare
n o n s i g n i f i c a n t adj. see NOT SIGNIFI-
PROBABILITY SAMPLING. CANT.
n o n r a n d o m i z e d c l i n i c a l t r i a l a nonsingular m a t r i x a SQUARE MA-
type of CLINICAL TRIAL in which the par- TRIX whose DETERMINANT is not equal
ticipants are not assigned by chance to to zero. Where A is a nonsingular ma-
the different treatment groups or inter- trix, A is an INVERSE MATRIX of the ma-
ventions. For example, if participants trix B, B is an inverse of A, and the
choose whether to receive a new treat- product of AB equals the product of BA.
ment or an existing standard of'care Also called invertible matrix. Com-
then the trial would be nonrandomized. pare SINGULAR MATRIX.
Compare RANDOMIZED CLINICAL TRIAL.
nonspecific effect a result or conse-
n o n r a n d o m i z e d design any of a quence whose specific cause or precipi-
large number of research designs in tating factors are unknown. An example
which participants or cases are not as- would be the effect on a patient of the
signed to experimental conditions via a belief that he or she has received medi-
chance process. For example, FIELD EX- cation or some other intervention when
PERIMENTS often are nonrandomized. no trae treatment has been given (see
Compare RANDOMIZED DESIGN. PLACEBO EFFECT).
nonreactive measure see UNOBTRU- n o r m n. 1. a standard or range of values
SIVE MEASURE. that represents the typical performance

237
normal approximation

of a group or of an individual (of a cer- The normal distribution has several

tain age, for example) against which primaty characteristics: It is symmetri-
comparisons can be made. 2. a conver- cal, it has both upper and lower ASYMP-
sion of a raw score into a scaled score TOTES, and its mean, MEDIAN, and MODE
that is more easily interpretable, such as are the same value. Perhaps most impor
percentiles or IQ scores. normative tant, however,fixedproportions of val-
adj. ues fall within defined sections of the
n o r m a l a p p r o x i m a t i o n 1. a proce- distribution. For example, 34.13% of
dure in which part of an observed distri- values fall between the mean and one
bution of values is approximated to part STANDARD DEVIATION above the mean,
of the symmetrical bell-shaped NORMAL and a corresponding 34.13% of values
DISTRIBUTION. Many sample data sets fall between the mean of the distribu-
can be normally approximated, thus sat- tion and one standard deviation below
isfying certain ASSUMPTIONS required the mean. Many statistical models are
for the use of particular analytic tech- based on the assumption that data fol-
niques and often simpUfying the associ- low a normal distribution. For example,
ated calculations. 2. the distribution of it is reasonable to expect that human
values resulting from such a process. height follows a normal distribution
with a mean of 5 feet several inches,
n o r m a l curve equivalent (NCE) a such that very few adults are less than 3
standardized test score based on a NOR- feet or greater than 7 feet tall. Also called
MAL DISTRIBUTION with a mean of 50 Gaussian curve; Gaussian distribu-
and a STANDARD DEVIATION of 21.06. It tion; normal curve.
is a measure of academic development
or performance created by the U.S. De- n o r m a l i t y n. the condition in which a
partment of Education to allow test data set presents a NORMAL DISTRIBU-
scores from different contexts and scales TION of values.
to be compared. NCEs range from 1 to normalize vb. to apply a TRANSFORMA-
99 in value and are an alternative to T TION to data in order to produce a new
SCORES, Z SCORES, and other types of set of scores that approximately follow
STANDARDIZED SCORES. the NORMAL DISTRIBUTION.
n o r m a l d i s t r i b u t i o n a THEORETICAL normal probability plot a type of
DISTRIBUTION in which values pile up in PROBABILITY PLOT in which Observed
the center at the MEAN and fall off into values are plotted against the values tha
tails at either end. When plotted, it gives would be consistent with a NORMAL DIS-
the familiar bell-shaped curve expected TRIBUTION. If the observed values ex-
when variation about the mean value is hibit normality, they will appear on a
random, as shown in the following ge- diagonal line from the lower left to the
neric depiction. upper right corner of the graph. Also
called full-normal plot.
n o r m a l r a n d o m variable any CON-
TINUOUS RANDOM VARIABLE whose val-
0.3-
ues follow a NORMAL DISTRIBUTION that
is centered around the MEAN for the
2< 0.2-
population and that has a STANDARD DE-
VIATION matching that of the popula-
2 0.1
Q.
tion. A STANDARD NORMAL VARIABLE IS
similar but has a mean of 0 and a stan-
dard deviation of 1.
n o r m a l range see REFERENCE INTERVAL.

238
 nuisance parameter

normal score see STANDARDIZED Thus, if a student obtains a score of 70%

SCORE. but the standard test score (norm) for
those of the same age is 90%, then the
n o r m a l variable see NORMAL RANDOM sradent has done relatively poorly. See
VARIABLE. also CRITERION-REFERENCED TEST; DO-
normative adj. relating to a particular MAIN-REFERENCED TEST. .
NORM or norms generally. Thus, a nor- no-treatment c o n t r o l group a CON-
mative score for a test (e.g., of language TROL GROUP who are not exposed to any
skills) reflects the average performance experimental manipulation or interven-
of a specific group (e.g., lO-year-old chU- tion, thus serving as a neutral compari-
dren) against which the score of a given son for other study groups receiving the
individual from that group is evaluated. treatment under investigation. For ex-
Similarly, normative research refers to ample, a researcher investigating the ef-
any empirical investigation designed to fectiveness of a new antidepressant
determine such standards of compari- might divide participants into a treat-
son. Compare NONNORMATIVE. ment group that receives the new medi-
normative scale any evaluative instru- cation and a separate no-treatment
ment on which the respondent provides control group that receives no medica-
ratings for a series of items or chooses tion whatsoever, and then compare the
scores to indicate his or her agreement outcomes of each group to identify any
with a series of statements. Unlike an significant differences. In contrast, an
IPSATIVE SCALE, there is no requirement ordinary control group may be adminis-
for these scores to sum to a particular tered some standard treatment instead
total (e.g., 100%). For example, a super- of the experimental one.
visor using a normative scale to assess an not significant (NS) denoting a result
employee's job performance might be from a statistical hypothesis-testing pro-
asked to choose a number from 1 to 5 to cedure that does not allow the re-
indicate how well the employee per- searcher to conclude that differences in
formed in each of several different areas, the data obtained for different samples
such as communication, timekeeping, are meaningful and legitimate. In other
and quality of work. The scores given in words, a result that is not significant
any area would not be affected by those does not permit the rejection of the
given in any of the others (e.g., the su- NULL HYPOTHESIS; any observed differ-
pervisor could award all 5s or all Is if he ences are considered to be due to chance
or she thought this was merited). or random factors. Also called non-
n o r m e d f i t index (NFl) see BENTLER- significant.
BONETT INDEX. nuisance parameter in statistical hy-
n o r m group see STANDARDIZATION pothesis testing, a population PARAME-
GROUP. TER of secondary interest that must be
accounted for in order to obtain an esti-
norm-referenced test any assessment mated value for a parameter of primaty
in which scores are interpreted by com- interest. For example, assume a re-
parison with a NORM, generally the aver- searcher wishes to determine whether
age score obtained by members of a male and female schoolchildren differ in
specified group. For example, a teacher their mathematics ability. He or she
might administer a norm-referenced might administer a math test to a sample
reading test to the students in his or her of children at a particular school and use
classroom, with each person's score in- those data to estimate male and female
dicating how well that test taker reads means for all children. If, however, the
relative to other examinees of that age. researcher needed to calculate the vari-

239
nuisance variable

ance of the scores before attempting to diction of a significant finding (e.g., a

determine the means, the former would significant difference between sample
be a nuisance parameter. means, a correlation that is significantly
different from zero). Statistical proce-
nuisance variable a type of EXTRANE- dures are applied to research data in an
OUS VARIABLE that does not differ sys- attempt to disprove or reject the NH at a
tematically across levels or conditions of predetermined SIGNIFICANCE LEVEL. See
the independent variable under investi- HYPOTHESIS TESTING.
gation but whose variation nonetheless
may contribute to an increase in experi- n u l l hypothesis significance test-
mental error. Participant characteristics i n g (NHST) see SIGNIFICANCE TESTING.
and environmental conditions often are
nuisance variables. For example, indi- null matrix a SQUARE MATRIX whose
viduals in a learning study who are dis- elements are all zeros. In MATRIX ALGE-
tracted by noise in a nearby room may BRA null matrices serve many of the
not perform as weU as they would other- same functions that a zero does in other
wise. contexts. Also called zero matrix.
n u l l d i s t r i b u t i o n in statistical testing, n u l l result see NULL FINDING.
the PROBABILITY DISTRIBUTION o f valueS
for a particular test statistic that is ob- n u l l set see EMPTY SET.
tained when the NULL HYPOTHESIS is
true. For example, the F RATIO from an number needed to treat (NNT) the
ANALYSIS OF VARIANCE folloWS the F DIS- number of people who must receive a
TRIBUTION if the null hypothesis is cor- particular therapeutic intervention in
rect. Also called n u l l hypothesis order to prevent a single adverse out-
distribution. Compare ALTERNATIVE come. An EFFECT SIZE commonly used in
HYPOTHESIS DISTRIBUTION. clinical research, it is the inverse of the
absolute risk reduction (ARR). For exam-
null finding the siraation in which the ple, if a new therapy for depression was
outcome of a statistical hypothesis- found to reduce the risk of suicide from
testing procedure indicates that there is 50% to 30%, its ARR is .20. Thus, the
no relationship, or no significant rela- NNT would be 1/.20 or 5, meaning that
tionship, between experimental vari- five people need to be treated for evety
ables. Also called null result. one suicide that is avoided. Smaller
NNTs indicate larger treatment effect
n u l l hypothesis (NH; symbol: HQ) a sizes.
statement that a srady will find no
meaningful differences between the n u m e r i c a l variable see QUANTITA-
groups or conditions under investiga- TIVE VARIABLE.
tion, such that there is no relationship
among the variables of interest and that Nuremberg code a set of 10 guidelines
any variation in observed data is the re- for conducting research with human
sult of chance or random processes. For participants that were established in
example, if a researcher is investigating a 1949 following public discovery of the
new technique to improve the skills of atrocities associated with the work of
children who have difficulty reading, Nazi scientists during World War II. The
the null hypothesis would predict no Nuremberg code introduced several im-
difference among the average reading portant principles, including INFORMED
performance of those children who re- CONSENT procedures and the right of in-
ceive the intervention and those who do dividuals to withdraw participation at
not. The NH is contrasted with the AL- any time. See INSTITUTIONAL REVIEW
TERNATIVE HYPOTHESIS, which is a pre- BOARD; RESEARCH ETHICS.

240
 Oo
objective adj. 1. having verifiable exis- ments, techniques of control, and obser-
tence in the external world, indepen- vations are as free from bias as possible.
dently of any opinion or judgment. 2. Compare SUBJECTIVITY.
impartial or uninfluenced by personal
o b l i m i n n. a class of OBLIQUE ROTATION
feelings, interpretations, or prejudices.
methods used in FACTOR ANALYSIS.
Compare SUBJECTIVE.
oblique adj. describing a set of axes that
objective e x a m i n a t i o n see OBJEC-
do not form right angles; in graphical
TIVE TEST. representations of mathematical com-
objective p r i o r see PRIOR DISTRIBU- putations (such as FACTOR ANALYSIS),
TION. this indicates correlated (not independ-
ent) variables. Compare ORTHOGONAL.
objective scoring scoring a test by
means of a key or formula, so that differ- oblique r o t a t i o n a transformational
ent scorers will arrive at the same score system used in FACTOR ANALYSIS when
for the same set of responses. It is con- two or more factors (i.e., LATENT VARI-
trasted with subjective scoring, in which ABLES) are correlated. Oblique rotation
the score depends on the scorer's opin- reorients the factors so that they fall
ion or interpretation of participant re- closer to clusters of vectors representing
sponses to items. MANIFEST VARIABLES, thereby simplify-
ing the mathematical description of the
objective test a type of assessment in-
manifest variables. It is one of two types
stmment consisting of a set of items or
questions that have specific correct an- of FACTOR ROTATION used to identify
swers (e.g., How much is 2 + 27), such a simpler stracrare pattern or solution,
that no interpretation, judgment, or per- the other being ORTHOGONAL ROTATION.
sonal impressions are involved in scor- oblique s o l u t i o n in FACTOR ANALYSIS,
ing. Examples include MULTIPLE-CHOICE any of various FACTOR STRUCTURE pat-
TESTS and TRUE-FALSE TESTS. In contrast, terns identified using OBLIQUE ROTA-
short-answer and essay examinations are TION methods.
SUBJECTIVE TESTS. Also called objective
examination. observation n. 1. the careful, close ex-
amination of an object, process, or other
objectivism n. the position that judg- phenomenon for the purpose of collect-
ments about the external world can be ing data about it or drawing conclusions.
established as true or false independent See CONTROLLED OBSERVATION; NATU-
of personal feelings, beliefs, and experi- RALISTIC OBSERVATION; PARTICIPANT OB-
ences. Compare SUBJECTIVISM. ob- SERVATION. 2. a piece of informafion (see
jectivist n., adj. DATA). observational adj.
objectivity n. 1. the tendency to base observational study research in
judgments and interpretations on exter- which the experimenter passively ob-
nal data rather than on subjective fac- serves the behavior of the participants
tors, such as personal feelings, beliefs, without any attempt at intervention or
and experiences. 2. a quality of a re- manipulation of the behaviors being ob-
search study such that its hypotheses, served. Such studies fypically involve
choices of variables studied, measure- observation of cases under naturalistic

241
observation coding system

conditions rather than the random as- EXPECTED FREQUENCY for the major. In-
signment of cases to experimental condi- deed, in the CHI-SQUARE TEST and other
tions: Specially trained individuals record statistical procedures observed frequen-
activities, events, or processes as precisely cies are compared with expected fre-
and completely as possible without quencies, with large differences suggest-
personal interpretation. Also called ob- ing a poorfitfor models or explanations
servational design; observational proposed to describe the data. Also
method; observational research. See called obtained frequency.
also NATURALISTIC OBSERVATION. observed score a data value recorded
observation coding system a for a variable via actual measurement or
scheme or list of mutually exclusive la- observation, as opposed to an estimated
bels, categories, and so fortheach of or predicted score (or to a score on a LA-
which characterizes a coherent dimen- TENT VARIABLE, which is derived or ap-
sion of interestused for classifying in- proximated).
formation obtained by observing others. observed-score e q u a t i n g see SCORE
It is an essential component of any OB- EQUATING.
SERVATIONAL STUDY as it clarifies what
data should be collected and how, pro- observed variable see MANIFEST VARI-
viding definitions of each code along ABLE.
with examples. For example, a re- observer bias any expectations, beliefs,
searcher investigating infant behavior or personal preferences of a researcher
might develop the following observa- that unintentionally influence his or her
tion coding system: (1) quietly alert, (2) recordings during an OBSERVATIONAL
crying, (3) fussing, and (4) sleeping. STUDY. See EXPERIMENTER EFFECT.
Thus, for every time during a specific ob-
servation period that an observer sees observer d r i f t gradual, systematic
the baby acting alert, he or she would re- changes over a period of time by a partic-
cord a 1 on his or her data form; for ular observer in his or her application of
every time the observer sees the baby criteria for recording or scoring observa-
crying, he or she would record a 2, and tions. See EXPERIMENTER DRIFT.
so forth. o b t a i n e d f r e q u e n c y see OBSERVED
observed d i s t r i b u t i o n the DISTRIBU- FREQUENCY.
TION of values on a variable as actually Occam's razor the maxim that given a
obtained from a SAMPLE, as opposed to a choice between two hypotheses, the one
THEORETICAL DISTRIBUTION based on involving the fewer assumptions should
the laws of probability. For example, one be preferred. In other words, one should
might draw a card from a standard deck apply the LAW OF PARSIMONY and
of playing cards, note the suit, replace choose simpler explanations over more
the card in the deck, and repeat the pro- complicated ones. [WUUam of Occam or
cess 100 times. An observed distribution Ockham (c. 1285-1347), English Francis-
would show how many times a heart, a can monk and Scholastic philosopher]
diamond, a club, and a spade were cho-
sen in the trial of 100 draws. odd-even r e l i a b i l i t y a method of as-
sessing the reliability of a test by corre-
observed frequency the counts of val- lating scores on the odd-numbered
ues or categories on a variable as ob- items with scores on the even-numbered
tained from a sample. For example, the items. The CORRELATION COEFFICIENT
number of people majoring in psychol- between the two halves is adjusted using
ogy at a particular college would be the the SPEARMAN-BROWN PROPHECY FOR-
observed frequency of that major, which MULA, which accounts for the decreased
may or may not match a theoreticaUy size of the odd and even item sets com-
 one-bend transformation

pared to the test as a whole. Qdd-even on the proportion of variance of one

reliability is a special case of SPLIT-HALF measure predictable from variance in
RELIABILITY. other measures. In ANALYSIS OF VARI-
ANCE it indicates how much variation in
odds n. the ratio of the probability of an
a DEPENDENT VARIABLE Can be explained
event occurring to the probability of the
by variation in one or more INDEPEN-
event not occurring, usually expressed
DENT VARIABLES.
as the ratio of two integers (e.g., 3:2).
odds r a t i o (OR) the quotient of two o m i t t e d variable bias the situation
ODDS. For example, in a study on a drug, in which values calculated from a statis-
the odds ratio is calculated as the odds of tical model systematically overestimate
an effect in a treated group divided by or underestimate a degree of relationship
the odds of the same effect in a control or other quantity of interest because an
group. A measure of EFFECT SIZE, it varies important variable has been left out of
from 0 to infinity, with a value of 1 indi- the model. For example, a researcher
cating no effect and a value of less than 1 could hypothesize a Unear REGRESSION
indicating a negative effect. EQUATION in which stressful life events
and lack of social support predict depres-
ogive n. the somewhat flattened S- sion. If coping skiUs also are highly rele-
shaped curve typically obtained by vant to predicting depression, the
graphing a CUMULATIVE FREQUENCY researcher's failure to include that ele-
DISTRIBUTION. Consider the following ment in his or her concepraalization
hypothetical Ulustration of test results would create an omitted variable bias.
from students in a classroom. The exclusion of important variables
from models may constrain the validity
of srady findings. See also BIASED ESTI-
MATOR;

omnibus test 1. any statistical test of

70
Exam Score
Cumulative frequency values are given
along the vertical y-axis and obtained
test scores are given along the horizontal
X-axis. The plot increases slightly at ei-
ther end, including that few students re-
ceived vety low or very high scores, but
rises much more steeply in the center,
indicating that the majority of students
received average scores.
OLS abbreviation for ordinaty least
squares. See LEAST SQUARES REGRESSION.
omega squared (symbol: oo^) a measure
of the STRENGTH OF ASSOCIATION based one-bend
significance in which more than two
conditions are compared simulta-
neously or in which there are two or
more INDEPENDENT VARIABLES. The
ANALYSIS OF VARIANCE is an example.
For instance, assume a researcher col-
lects data from an experimental design
having two independent variables, each
of which has three different levels or
conditions. He or she could conduct an
analysis of variance to concurrently ex-
amine the mean values for all levels of
each of the variables (and their combi-
nations thereof) in order to determine
whether there are any significant differ-
ences among them. 2. a type of exam
that simultaneously measures several
different abilities. For example, a mathe-
matics omnibus test may include items
assessing numeracy, algebraic skills, geo-
metric proficiency, and trigonometric
competency.
transformation a se-

243
one-factor analysis o f variance

quence of P O W E R F U N C T I O N S used to CONFIDENCE INTERVAL in which Only an

convert a M O N O T O N I C relationship be- upper or lower boundaty is specified. For
tween two variables into a linear one, example, if a researcher has a negatively
such that a single bend in the curve de- skewed data set (see SKEWNESS), most
picting their values on a graph is data points will be at the upper end of
straightened out. See also T W O - B E N D the range of values. Thus, he or she likely
TRANSFORMATION. would focus upon that upper area and
specify only a maximum value for the
o n e - f a c t o r a n a l y s i s o f v a r i a n c e see
population PARAMETER (i.e., a value
ONE-WAY ANALYSIS OF VARIANCE.
above which the parameter is not likely
one-group pretest-posttest design to fall).
a variation of the PRETEST-POSTTEST DE-
SIGN in which only a single set of partici- o n e - t a i l e d h y p o t h e s i s see DIREC-
pants is measured on a DEPENDENT T I O N A L HYPOTHESIS.
VARIABLE of interest, exposed to a treat- o n e - t a i l e d t e s t see D I R E C T I O N A L TEST.
ment or intervention, and then mea-
sured again to determine the change or o n e - w a y a n a l y s i s o f c o v a r i a n c e an
difference between the initial (pre-) and ANALYSIS OF C O V A R I A N C E in which there
second (post-) measurement. The lack of is a single I N D E P E N D E N T V A R I A B L E and
a CONTROL GROUP in this type of re- one or more COVARIATES whose poten-
search design makes it difficult to attrib- tial influence needs to be accounted for
ute gains in the posttest score to the statistically. For example, assume a re-
intervention, as other elements (e.g., searcher is evaluating how threetypesof
participant MATURATION) may have instructional methods (three levels of
contributed to any change observed. the independent variable) affect scores
Also called one-group pre-post de- on an academic achievement test (the
sign. D E P E N D E N T V A R I A B L E ) . If the researcher
believes general intelligence, gender, and
one-parameter m o d e l see RASCH other unmeasured factors could also af-
MODEL. fect scores, he or she might use a one-
o n e - s a m p l e r u n s test see SINGLE- way analysis of covariance to control for
SAMPLE RUNS TEST. such covariates when assessing the ex-
perimental data.
o n e - s a m p l e test see SINGLE-SAMPLE
TEST. o n e - w a y a n a l y s i s o f v a r i a n c e an
o n e - s a m p l e test f o r t h e m e d i a n see ANALYSIS OF VARIANCE that evaluates
SINGLE-SAMPLE TEST FOR THE MEDIAN. the influence of different levels or condi-
tions of a single INDEPENDENT VARIABLE
o n e - s a m p l e t test see SINGLE-SAMPLE T upon a DEPENDENT VARIABLE. The mean
TEST. values of two or more samples are exam-
o n e - s a m p l e z test see SINGLE-SAMPLE Z ined in order to determine the probabil-
TEST. ity that they have been drawn from the
same population. Also called one-
o n e - s h o t case s t u d y a research design factor analysis of variance; single-
in which a single group is observed on a factor analysis of variance.
single occasion after experiencing some
event, treatment, or intervention. Since o n e - w a y b l o c k e d a n a l y s i s o f v a r i -
there is no CONTROL GROUP against which a n c e see R A N D O M I Z E D B L O C K O N E - W A Y
to make comparisons, it is a weak design; ANALYSIS OF VARIANCE.
any changes noted are merely presumed
o n e - w a y d e s i g n an experimental de-
to have been caused by the event.
sign in which a single INDEPENDENT
one-sided c o n f i d e n c e i n t e r v a l a VARIABLE is manipulated to observe its
 operational research

influence on a DEPENDENT VARIABLE. For open study a research project in which

example, a researcher could use a one- new participants can be added after the
way design to examine the effect of dif- project has begun.
ferent amounts of daily exercise (e.g., 0 operating characteristic curve for
minutes, 30 minutes, 60 minutes) on
a given SIGNIFICANCE TESTING proce-
mood. Also called single-factor de-
dure, a plot showing the probability of
sign.
the NULL HYPOTHESIS being correct at
open-ended i n t e r v i e w an interview different values of a PARAMETER. Typi-
in which the interviewee is asked ques- cally, the former is given along the verti-
tions that cannot be answered with a cal y-axis and the latter along the hori-
simple yes or no. For example, a human zontal Ai-axis, as in the following generic
resources staff member interviewing a depiction.
candidate for employment might ask,
"What were the major responsibilities of
your most recent job?" Qpen-ended in-
terviews encourage interviewees to talk
freely and extensively, thus providing
information that might not be obtained
otherwise. The general questions and
their order may be planned in advance
(see STRUCTURED INTERVIEW) or a single
initial question can be planned and the
subsequent discussion allowed to pur-
sue various areas of interest as they arise
(see UNSTRUCTURED INTERVIEW).

open i n t e r v a l a set of numbers that

falls between two defined endpoints.
Parameter Value
For example, an open interval (0, 3)
around x indicates that the value of x
will fall between 0 and 3 but not equal Operational definition a description
either. In contrast, the CLOSED INTERVAL of something in terms of the operations
[0,3] (written with square brackets) indi- (procedures, actions, or processes) by
cates that X may be equal to 0 or 3 or as- which it could be observed and mea-
sume any value in between. sured. For example, the operational defi-
nition of anxiety could be in terms of a
open-label study a CLINICAL TRIAL in test score, withdrawalfroma siraation,
which both the researcher and the par- or activation of the sympathetic nervous
ticipant know the treatment the partici- system. The process of creating an opera-
pant is receiving. tional definition is known as opera-
tionalization.
open question 1. in a test or survey, an
item that does not come with any multi- operationalism n. the position that
ple-choice options for the respondent. the meaning of a scientific concept de-
Compare FIXED-ALTERNATIVE QUES- pends upon the procedures used to es-
TION. 2. in an interview, a question that tablish it, so that each concept can be
encourages the respondent to answer defined by a single observable and mea-
freely in his or her own words, providing surable operation. Also called opera-
as much or as little detail as desired. For tionism.
example, a candidate in a job interview
may be asked the open question, "How operational research see OPERA-
would you describe yourself?" TIONS RESEARCH.

245
 operationism

operationism n. see OPERATIONALISM. o r a l history background information

operations research the application about a person provided verbally by that
of advanced analytical methodssuch individual during an interview. Oral his-
as mathematical modeling, OPTIMIZA- tories may describe such things as per-
TION, and computer simulationto the ceptions and thoughts, experiences of
study of complex situations so as to ob- important events, and family relation-
tain a comprehensive understanding ships. They can be collected, tran-
that allows for accurate predictions of scribed, and even analyzed as part of
outcomes and estimates of risk and en- research studies.
ables more rational, effective decision o r a l test any assessment in which the
making. It is used particularly in man- questions are posed and answered ver-
agement science and business contexts. bally.
opinion survey a technique in which a
ordered scale see ORDINAL SCALE.
large number of people are polled to de-
termine their collective views, beliefs, or order effect 1. in WITHIN-SUBJECTS DE-
attitudes about a particular topic of in- SIGNS, the influence of the order in
terest. Information so obtained often is which treatments are administered, for
extrapolated to a broader population example, the effect of being the first ad-
with a given MARGIN OF ERROR. For ex- ministered treatment (rather than the
ample, during an election year one second, third, and so forth). As individu-
might conduct an opinion survey in dif- als participate in first one and then an-
ferent states of preferences for potential other treatment condition, they may
U.S. presidential candidates. Also called experience increased fatigue, boredom,
opinion poll. and familiarity with or practice with re-
o p i n i o n n a i r e n. a type of measure for acting to the independent variable. Any
assessing the attitudes or beliefs of an in- of these conditions could affect the par-
1^ dividual about particular topics. It com- ticipants' responses and C O N F O U N D the
prises a list of various statements that results of the study. Researchers often
the respondent is asked to endorse or re- use COUNTERBALANCING to control for
ject. order effects. See also CARRYOVER EF-
FECT; SEQUENCE EFFECT. 2. the influence
o p p o r t u n i t y s a m p l i n g see CONVE- of the order in which items or state-
NIENCE SAMPLING. ments are listed on surveys and ques-
o p t i m a l design an approach to experi- tionnaires. Three basictypesof question
mental design in which the conditions order effect have been identified: (a) un-
to be studied and the assignment of par- conditional, in which the answer to a
subsequent question is affected by hav-
ticipants to those conditions are deter-
ing responded to the prior question but
mined so as to best fulfill particular
not by the response given on that prior
statistical criteria. See A-OPTIMAL DE-
question; (b) conditional, in which the
SIGN; D-OPTIMAL DESIGN; E-OPTIMAL DE-
answer to a subsequent question de-
SIGN. pends on the response given to the prior
o p t i m a l scaling see DUAL SCALING. question; and (c) associational, in which
the correlation between the prior and
o p t i m i z a t i o n n. a statistical process in subsequent questions changes depend-
which ALGORITHMS are applied to data ing on which is asked first. Order effects
from a SAMPLE so as to obtain the best present serious problems when measur-
estimate of a value for a PARAMETER of ing change over timeunless the ques-
interest in the broader population. tion order is the same for each data
OR abbreviation for ODDS RATIO.
collection it is difficult to know whether

246
 orthogonal

change reflects legitimate respondent vere dementia. It is important to note

differences or question effects. that although the results between
groups vary depending on the levels of
order o f magnitude the approximate the variables under study, they do not
degree or strength of a number or value reverse their direction of influence as
within a range, usually to the nearest they would in a DISORDINAL INTERAC-
power of 10. For example, 2,500 (2.5 x TION.
10^) and 4,300 (4.3 x 10^) are of the
same order of magnitude, but both are o r d i n a l scale a sequence of numbers
one order of magnirade greater than 240 that do not indicate magnirade or a true
(2.4 X 10^). zero point but rather reflect a rank order-
ing on the attribute being measured. For
o r d i n a l adj. pertaining to rank, order, or example, an ordinal scale for the perfor-
position in a series. Compare CARDINAL; mance of a specific group of people on a
NOMINAL. particular test might use the number 1 to
o r d i n a l data numerical values that rep- indicate the person who obtained the
resent rankings along a continuum from highest score, the number 2 to indicate
lowest to highest, as in a judge's assign- the person who obtained the next high-
ment of a 2 to denote that a particular est score, and so on. It is important to
athlete's performance was fair and a 3 to note that an ordinal scale does not pro-
denote that a subsequent athlete's per- vide any information about the degree
formance was better. Ordinal data may of difference between adjacent ranks
be counted (i.e., how many athletes ob- (e.g., it is not clear what the actual point
tained a 2, how many a 3, etc.) and ar- difference is between the rank 1 and 2
ranged in descending or ascending scores). Compare INTERVAL SCALE; NOM-
sequence but may not be manipulated INAL SCALE; RATIO SCALE.
arithmeticallysuch as by adding, sub- o r d i n a l variable a variable whose pos-
tracting, dividing, or multiplying any sible values have a clear rank order. For
rank by any otherbecause the actual example, attitude is an ordinal variable
difference in performance between adja- as it may be denoted with ordered points
cent values is unspecified and may vary. indicating increasing or decreasing val-
In other words, one does not know how ues, such as 1 = strongly disagree, 2 = dis-
much better a rank of 3 is than a 2, and agree, 3 = agree, and 4 = strongly agree.
the difference between a 2 and a 3 may Values on an ordinal variable indicate
not be the same as the difference be- that one data point is higher or lower
tween a 3 and a 4. than another but do not define the ex-
tent of the difference between them.
o r f l i n a l i n t e r a c t i o n in a FACTORIAL
Compare INTERVAL VARIABLE.
ANALYSIS OF VARIANCE, a simation in
which the effect of one INDEPENDENT ordinary least squares regression
VARIABLE is greater for certain condi- see LEAST SQUARES REGRESSION.
tions of another independent variable. It
is indicated by a pattern of converging ordinate n. the vertical coordinate in a
lines when plotting the MAIN EFFECTS graph or data plot; that is, the y-axis. See
from the analysis. For example, an ordi- also ABSCISSA.
nal interaction would exist if a re-
searcher evaluating the effect of two or rule see ADDITION RULE.
dementia treatments (A and B) found orthogonal adj. 1. describing a set of
that with both A and B functioning im- axes at right angles to one another,
proved significantly in individuals with which in graphical representations of
a slight to moderate degree of dementia mathematical computations (such as
but had little benefit for those with se- FACTOR ANALYSIS) and other research in-

247
 orthogonal coding

dicates uncorrelated (unrelated) vari- particular set of payments, one set being
ables. Compare OBLIQUE. 2. denoting a paid to each participant.
research design in which there are an
equal or proportional number of partici- outcome research a systematic inves-
pants across study conditions. See OR- tigation of the effectiveness of a type or
THOGONAL DESIGN. technique of intervention (e.g., a new
form of psychotherapy for treating de-
orthogonal coding see CONTRAST pression) or of the comparative effec-
CODING. tiveness of different intervention types
or techniques (e.g., cognitive behavior
orthogonal contrast in a FACTORIAL therapy vs. drug therapy for depression).
ANALYSIS OF VARIANCE, a set of non- In other words, outcome research fo-
redundant comparisons among the cuses on determining whether partici-
mean values for different samples. In pants benefit from receiving the
other words, the set contains no overlap- intervention. Compare PROCESS RE-
ping information, such that the results SEARCH.
from one comparison provide no infor-
mation about the results of the second outcome variable see DEPENDENT
comparison. Also called orthogonal VARIABLE.
comparison.
outlier n. an extreme observation or
orthogonal design an experimental measurement, that is, a score that signif-
design involving multiple INDEPENDENT icantly differs from all others obtained.
VARIABLES in which each level of one For instance, assume a researcher ad-
variable is combined with each level of ministered an intelligence test to a
every other and all of the resulting con- group of people. If most individuals ob-
ditions contain an equal or proportional tained scores near the average IQ of 100
number of participants or observations. yet one person had an IQof 150, the lat-
ter score would be an outlier. OutUers
JM orthogonal rotation a trans- can have a high degree of influence on
formational system used in FACTOR summaty statistics (e.g., the MEAN and
ANALYSIS in which the different under- STANDARD DEVIATION can be pulled se-
lying or LATENT VARIABLES are required verely toward outliers) and on estimates
to remain separated from or un- of PARAMETER values, and they may dis-
correlated with one another. It is one of tort research findings if they are the re-
t w o types o f F A C T O R R O T A T I O N USCd tO sult of error.
identify a simpler structure pattern or
solution, the other being OBLIQUE ROTA- overdispersion n. in CATEGORICAL
TION. Also called rigid rotation. DATA ANALYSIS, 3 Situation in which ob-
tained observations display more varia-
orthogonal solution in FACTOR tion (DISPERSION) than is predicted by a
ANALYSIS, any of various FACTOR STRUC- model. Overdispersion distorts STAN-
TURE patterns identified using OR- DARD ERROR and CONFIDENCE INTERVAL
THOGONAL ROTATION methods. estimates and often is associated with
OUTLIERS or model MISSPECIFICATION.
outcome n. 1. the result of an experi-
ment, treatment, intervention, or other overfitted model a model of a data set
event. For example, the test scores of stu- in which there are too many predictor
dents in a classroom who have been variables or PARAMETERS relative to the
taught with a new lecture method com- number of sample observations. An
prise an outcome, as does any individual overfitted model is unnecessarily com-
element from a SAMPLE SPACE. 2. in plex for the amount of avaUable data
GAME THEORY, the factor determining a upon which it is based and does not rep-

248
 o-x-o
licate well nor accurately predict re- to the treatment rather than to sex or
sponses. age differences. If, however, the re-
searcher were to pair the groups on such
o v e r i d e n t i f i c a t i o n n. in STRUCTURAL additional factors as area of residence
EQUATION MODELING and Similar statis- and household income, overmatching
tical techniques, the presence of more would be present and likely to mask the
PARAMETERS in a model than are re- true narare of the relationship under in-
quired to correctly specity the relation- vestigation and lead to statistical BIAS,
ships among the variables of interest. In such as by reducing the POWER and EFFI-
other words, the number of known pa- CIENCY of analyses.
rameters exceeds the number of free pa-
rameters, allowing one or more values to oversampling n. a sampling strategy in
be estimated in more than one way. which certain subsets of participants are
Compare UNDERIDENTIFICATION. overrepresented in a study group com-
pared to the larger population from
overidentified model a model of a which they are drawn. Oversampling in-
data set in which the number of PARAM- volves deliberately selecting greater
ETERS estimated exceeds the number of numbers of such participants than
data points or pieces of unique informa- would be obtained via RANDOM SAM-
tion in the data. In other words, an PLING so as to enhance the accuracy of
overidentified model contains elements PARAMETER values estimated through
that are redundant, with more known statistical procedures. For example, a re-
than free parameter values. Compare searcher selecting firefighters from a
JUST-IDENTIFIED MODEL; UNDERIDENTI- candidate pool might include equal
FIED MODEL. numbers of males and females to ensure
a sufficient sample of females for statisti-
overmatching n. unnecessary MATCH- cal analysis, even though males com-
ING: the pairing of research participants prise the majority of firefighters in the
on an excessive number of characteris- pool.
tics or on characteristics having little
or no potential influence upon the out- oversaturated model see SATURATED
come of interest. For example, a re- MODEL.
searcher investigating a new drug O-X-O shorthand for a ONE-GROUP PRE-
treatment for cancer might create two TEST-POSTTEST DESIGN, in which the re-
groups whose members are of the same searcher observes and measures a single
age and sex, administering the drag to set of participants (O), introduces an in-
one group while the other receives a pla- tervention (X), and then measures the
cebo. Such group comparability would participants (Q) again to determine
allow the researcher greater validity in whether the intervention resulted in
attributing any changes between them any change.

249
 Pp
p symbol for PROBABILITY.
paired comparison method a sys-
tematic procedure for scaling and com-
paring a set of stimuli or other items. A
pair of stimuli is presented to the partici-
pant, who is asked to compare them on a
particular dimension: This is often a
physical characteristic, such as size,
loudness, or brightness, but may also in-
clude personal traits, abilities, or perfor-
mance on some task. The process is
continued until evety item in the set has
been compared with evety other item; in
a work setting, for example, the process
would be complete when evety em-
ployee has been evaluated relative to
every other employee supervised by the
rater. The number of pairs is found by
the formula n x ( n ~ 1)12, where n is the
number of objects or people to be rated.
Thus, if there are five objects or people
to be compared, the number of pairs is
10; if there are 20 objects or people to be
rated, the number of pairs is 190. This
latter example illustrates a limitation of
the procedure; when the number of ob-
jects to be rated becomes large, the num-
ber of paired comparisons becomes
excessive and burdensome for the rater.
See PAIRED COMPARISON EXPERIMENT.
paired comparison experiment a
study design in which sets of paired al-
ternatives are offered to respondents,
who have to indicate which of the alter-
natives they prefer or rank more highly
on a specified dimension. The alterna-
tives offered may be real or hypothetical
and judgments can be qualitative or
quantitative. The purpose of the paired
comparison experiment is to assess the
importance of various attributes to the
respondents and to determine the
weights the respondents attach to the
250
different levels of these attributes. Also
called paired comparison design.
See PAIRED COMPARISON METHOD.
paired sample any sample in which
each participant is matched on a partic-
ular variable to a participant in a second
sample. This ensures that any differ-
ences on an outcome variable cannot be
due to differences between participants
on the matching variables. A simple ex-
ample of a paired sample is a PRETEST-
POSTTEST DESIGN in which a variable is
measured before and after an interven-
tion; in this case the participant is
matched to him- or herself. More fre-
quently, paired samples are achieved by
matching individuals on personal char-
acteristics such as age and gender.
Typically, each member of the pair is
randomly assigned to the treatment or
control group. See MATCHED-PAIRS DE-
SIGN.
paired-samples t test see MATCHED-
PAIRS T TEST.
pairwise comparison in a FACTORIAL
DESIGN where the variable being investi-
gated has more than two levels, a proce-
dure in which the data obtained from
each level of the variable are compared
separately to the data from evety other
level. For example, where the research
interest is the differences in student
achievement resulting from three differ-
ent types of content presentation, statis-
tical tests would be used to evaluate the
different outcomes of each pair of con-
tent methods (i.e.. Method 1 and
Method 2, 1 and 3, and 2 and 3).
DUNNETT'S MULTIPLE COMPARISON
TEST, T U K E Y ' S H O N E S T L Y SIGNIFICANT
DIFFERENCE TEST, a n d the N E W M A N -
KEULS MULTIPLE COMPARISON TEST aU
 parameter estimation

involve pairwise comparisons. See also framework within, or a general perspec-

MULTIPLE COMPARISONS. tive of, a discipline.
pairwise deletion a method in which parallel f o r m see ALTERNATE FORM.
data for a variable pertinent to a specific parallel-forms r e l i a b i l i t y see ALTER-
assessment are included, even if values NATE-FORMS RELIABILITY.
for the same individual on other vari-
ables are missing. For example, consider parallel-groups design a research de-
a researcher studying the influence of sign that compares two conditions or
age, education level, and current salaty treatments (e.g., relaxation training vs.
on the socioeconomic staras of a sample no training) such that both groups are
of employees. If assessing specifically sradied simultaneously and each partici-
how education and salary influence so- pant receives one treatment only. A par-
cioeconomic Staras, he or she could in- allel-groups design thus contrasts with a
clude all participants for whom that data CROSSOVER DESIGN, in which each indi-
had been recorded even if they were vidual participates in each condition.
missing information on age, as the latter parallelism n. 1. in general, the quality
variable is not of interest in the current or condition of being parallel, stractur-
analysis. Also called available-case ally similar, or having corresponding
analysis. Compare LISTWISE DELETION. features. 2. in philosophy, the proposi-
panel data observations on multiple tion that, although mind and body con-
phenomena collected over multiple stimte separate realities, they function
time periods for the same group of indi- in parallel such that responses seem ho-
viduals or other units. Repeated observa- listic and the two realms seem to assert
tions permit the researcher to srady the causal control over each other.
dynamics of change using techniques parallel research a srady that exam-
such as TIME-SERIES ANALYSIS. Panel ines an issue by simultaneously using
data have characteristics of both LONGI- two or more methods, such as using
TUDINAL DATA and cross-sectional data both focus groups and individual inter-
(see CROSS-SECTIONAL ANALYSIS). For views.
example, the values of the gross annual
income for each of 500 randomly cho- p a r a l l e l threshold method in CLUS-
sen households in New York City col- TER ANALYSIS, a Strategy in which several
lected for each of 10 years would be cluster centers are determined, then ob-
panel data. jects that are within a predetermined
threshold of these centers are grouped
panel study see LONGITUDINAL DESIGN; together. Compare SEQUENTIAL THRESH-
PANEL DATA. OLD METHOD.
paper-and-pencil test a test in which parameter n. 1. a characteristic of a
the questions or problems are written, POPULATION, such as the mean or STAN-
printed, or drawn and the answers are DARD DEVIATION, that is described or es-
written down. timated by a STATISTIC obtained from
sample data. For example, the mean
p a r a d i g m n. 1. a model, pattern, or rep-
score on a national exam for a sample of
resentative example, as of the functions
colleges provides an estimate of this pa-
and interrelationships of a process, a be-
rameter in the population of colleges. 2.
havior under study, or the like. 2. an ex-
any of the variables in a statistical model
perimental design or plan of the various
that is sradied or used to explain an out-
steps of an experiment. 3. a set of as-
come or relationship.
sumptions, attirades, concepts, values,
procedures, and techniques that consti- parameter estimation see ESTIMA-
tutes a generally accepted theoretical TION.

251
parameter space

parameter space the set of experimen- ables. This can help a researcher to get
tal variables that are used as input into a a clearer understanding of the relation-
model and serve as the basis for estima- ship between x and y. Also called semi-
tion of PARAMETERS in a population. partial correlation. Compare PARTIAL
CORRELATION; ZERO-ORDER CORRELA-
parametric adj. describing any analytic
TION.
method that makes ASSUMPTIONS about
the data of interest. Compare NON- p a r t i a l autocorrelation in TIME-
PARAMETRIC SERIES DATA, the relation between a
parametric hypothesis test see measure at one time point and the same
PARAMETRIC TEST. measure at a subsequent time point,
controlling for study variables that are
parametric statistics statistical pro- thought to affect the magnirade of the
cedures that are based on assumptions over-time relation. For example, mood
about the distribution of the attribute ratings from one day to the next may be
(or attributes) in the population being affected by whether or not a positive so-
tested, for example, that there is a NOR- cial interaction occurred. See AUTO-
MAL DISTRIBUTION of values. Compare CORRELATION; PARTIAL CORRELATION.
NONPARAMETRIC STATISTICS.
parametric test a HYPOTHESIS TEST p a r t i a l correlation the association
that involves one or more assumptions between two variables, x and y, with the
about the underlying distribution of the influence of one or more other variables
population (typicaUy assumed to be a (zj, Z2) statistically removed, controlled,
NORMAL DISTRIBUTION) from which the or held constant; the effect of the z vari-
sample is drawn. Common parametric able is removed from both x and y. For
hypothesis tests include ANALYSIS OF example, partial correlation between sal-
VARIANCE, REGRESSION ANALYSIS, CHI- ary and education level can be examined
SQUARE TESTS, TTESTS, and Z TESTS. Also
after the effects of age on each are re-
called distribution-dependent test. moved. It is often of interest to learn
Compare NONPARAMETRIC TEST. whether a correlation is significantly re-
duced in magnitude once a third vari-
Pareto d i s t r i b u t i o n a distribution able is removed. Also called higher
whose values follow a POWER FUNCTION, order correlation; higher order
such that there is a narrow peak at the partial correlation; partial associ-
extreme left of the x-axis and a long tail ation; partial relationship. Com-
to the right. Pareto distributions occur pare PART CORRELATION; ZERO-ORDER
in a diverse range of phenomena, such CORRELATION.
as the distribution of city populations or
income, that are not adequately de- p a r t i a l eta squared (symbol: y\p) in
scribed by a NORMAL DISTRIBUTION. The ANALYSIS OF VARIANCE, an EFFECT SIZE
marked SKEWNESS of the Pareto distribu- measure given as the ratio of the BE-
tion is often summarized in the so-called TWEEN-GROUPS S U M OF SQUARES (SSB)
80:20 rule, which states that, for exam- relative to the sum of the SSg and the
ple, 20% of cities will have 80% of the ERROR SUM OF SQUARES. It expresses the
total population, 20% of people will proportion of variance in the dependent
earn 80% of total income, and so on. or outcome variable that is accounted
for by the factors or independent vari-
p a r s i m o n y n. see LAW OF PARSIMONY. ables in the model. A higher value
part correlation the association be- means that a higher proportion of the
tween two variables, x and y, with the in- variance was accounted for by the fac-
fluence of a third variable, z, removed tors included in the model. See also ETA
from one (but only one) of the two vari- SQUARED; GENERALIZED ETA SQUARED.
 partial replication

p a r t i a l i n t e r a c t i o n a type of POST pendent and independent variables).

HOC TEST in which different aspects of a Partialing helps clarify a specific rela-
complex INTERACTION EFFECT are exam- tionship by excluding the effects of
ined one at a time. The procedure makes other variables that may also be associ-
use of CONTRAST CODING to determine ated. In a MULTIPLE REGRESSION model,
whether the effect of one variable is the for example, a researcher may want
same across all levels or conditions of to remove the effects of a set of
another variable. For example, a re- COVARIATES before examining the final
searcher whose study data show that REGRESSION COEFFICIENTS of Interest.
diet, gender, and exercise jointly influ- This process can be accomplished using
ence weight loss may use partial interac- a hierarchical method, entering the
tions to identity and characterize the covariates in thefirststep and the vari-
specific contributions of each. ables of interest in subsequent steps.
Also called partialing out.
partial-interval recording a me-
thod for measuring or coding instances p a r t i a l l y sequential sum of
of behavior within particular time peri- squares see TYPE II SUM OF SQUARES.
ods of observation. For frequently occur-
p a r t i a l regression coefficient a
ring behaviors or behaviors with a long
measure of the relationship between an
duration, a researcher can identify time
outcome variable and a particular inde-
intervals in advance of the coding and
pendent variable with the effects of the
have judges indicate whether a particu-
other predictor variables in the equation
lar behavior occurs (or not) within each
controlled for. It gives the amount by
period. For example, in videos of mock
which the outcome variable increases
job interviews the method could be used
when one independent variable.is in-
to record whether or not the candidate
creased by one unit and all the other in-
had eye contact with the interviewer
dependent variables are held constant.
within different periods of the inter-
For example, a researcher may want to
view. See also WHOLE-INTERVAL RE-
understand how several independent
CORDING.
variables together predict salaty and also
p a r t i a l least squares regression a how each variable explains salaty while
statistical procedure that generalizes and controlling for the influence of the other
combines fearares from PRINCIPAL C O M - variables in the equation. The specific
PONENTS ANALYSIS and MULTIPLE RE- contribution of each independent vari-
GRESSION. It is used when a researcher able to the REGRESSION EQUATION is as-
wants to estimate the outcomes on a set sessed by the partial regression coeffi-
of dependent variables from a (very) cient associated with each variable. This
large set of predictors, especially when coefficient corresponds to the increment
these predictors have a high degree of in EXPLAINED VARIANCE obtained by
MULTICOLLINEARITY or exceed the num- adding this variable to the regression
ber of obtained observations. With this equation after all the other independent
approach, a researcher can use varying variables already have been included.
measurement levels for both the inde- Also called partial regression weight;
pendent and the dependent variables partial slope coefficient. See SIMUL-
and can also use small samples. It is, TANEOUS REGRESSION.
however, particularly important to vali-
date findings in independent data sets. p a r t i a l regression plot see ADDED-
VARIABLE PLOT.
p a r t i a l i n g n. an approach to the STA-
p a r t i a l relationship see PARTIAL COR-
TISTICAL CONTROL of the influence of
RELATION.
a variable or set of variables on other
variables of interest (typically, the de- p a r t i a l r e p l i c a t i o n a REPLICATION of

253
partial residual plot

an empirical study in which only a sub- researcher can become enmeshed with
set of the srady's design and methodol- the group to the extent that he or she is
ogy are repeated. Often, a researcher will no longer able to document it in an un-
choose to conduct a partial replication biased way. Also called participative
to show that the general findings of research; participatory research.
a study remain the same, despite the
methodological changes. p a r t i c i p a n t report a SELF-REPORT
provided by an individual who takes
partial residual plot a graphical rep- part in a study. These data are useful be-
resentation of the relationship between cause an individual has a unique access
an INDEPENDENT VARIABLE (x^) and a to his or her own thoughts, emotions,
DEPENDENT VARIABLE (y) with the other and behaviors. On the other hand, an
independent variables in the model (X2, individual may show biases due to mem-
x^, ... x^) statistically removed or con- oty recall and the desire to be seen in a
trolled for. It allows the researcher to see particular light.
the unique effects of an independent
variable on the outcome. These plots are participants' rights in a study ap-
less informative when the independent proved by an INSTITUTIONAL REVIEW
variables are highly correlated. Also BOARD, a set of conditions relating to
called added-variable plot; compo- participants in the srady and their role
nent-plus-residual plot. in the research. Participants normally
should be informed about the purpose
p a r t i c i p a n t n. a person who takes part of the study (but see DECEPTION RE-
in an investigation, study, or experi- SEARCH), its procedures (i.e., what spe-
ment, such as by performing tasks set by cifically is expected to occur) and the
the experimenter or by answering ques- associated costs and benefits; that their
tions set by a researcher. In an experi- data from the study will be kept confi-
mental design, the person may be dential; whom they can contact if they
further identified as an experimental have any concerns; and that they can
participant (see EXPERIMENTAL GROUP) leave the study at any time without pen-
or a control participant (see CONTROL alty. Also called subjects' rights. See
GROUP). Participants may also be re- also INFORMED CONSENT; RESEARCH
ferred to as SUBJECTS, although the for- ETHICS.
mer term is now often preferred when
referring to humans rather than ani- p a r t i c i p a n t variable see SUBJECT
mals. See also RESPONDENT. VARIABLE.
participative research see PARTICI-
p a r t i c i p a n t m a t c h i n g see MATCH- PANT OBSERVATION.
ING.
participatory a c t i o n research a
p a r t i c i p a n t observation a QUASI- form of ACTION RESEARCH that empha-
EXPERIMENTAL RESEARCH method in
sizes collaboration between researchers
which a trained investigator studies a and members of the disadvantaged com-
preexisting group by entering it as a
munity of interest.
member, while avoiding a conspicuous
role that would alter the group processes p a r t i c i p a t o r y evaluation a type of
and bias the data. The researcher's role PROGRAM EVALUATION in which the in-
may be known or unknown to the other dividuals who provide a service (e.g.,
members of the group. Cultural anthro- professional staff) take a direct role in
pologists become participant observers evaluating the service or program; cli-
when they enter the life of a given cul- ents of the service provided may also be
rare to study its stracture and processes. involved. The official evaluator facili-
A downside to the approach is that the tates proceedings rather than actually

254
 path diagram

conducting the appraisal. Also called called deception by omission. Com-

collaborative evaluation; empow- pare ACTIVE DECEPTION.
erment evaluation.
p a t h analysis a type of STRUCTURAL
p a r t i c i p a t o r y research see PARTICI- EQUATION MODELING used to examine a
PANT OBSERVATION. set of simultaneous LINEAR RELATION-
p a r t i t i o n 1. vb. in statistics and mea- SHIPS between variables. Results are dis-
surement, to divide VARIANCE observed played using a figure in which boxes
in the DEPENDENT VARIABLE into com- denote measured variables, bidirectional
ponent elements on the basis of its na- arrows show correlations, directional ar-
ture and origin. 2. n. a specific rows show causal or predictive relations,
component of the observed variance and arrows coming from nowhere show
(e.g., variance due to a treatment effect error associated with the prediction
and variance due to RANDOM ERROR). model. Each variable is expressed as a
LINEAR FUNCTION of the preceding vari-
Parzen w i n d o w estimator see KER- able as weU as by a unique LATENT VARI-
NEL DENSITY ESTIMATOR. [Emanuel ABLE; the weights and STANDARD
Parzen (1929- ), U.S. stafisfician] ERRORS for each coefficient provide in-
Pascal d i s t r i b u t i o n see NEGATIVE BI- formation about the size of the PARAME-
NOMIAL DISTRIBUTION. TER. GOODNESS OF FIT Is tested by
comparing models with more pathways
Pascal's triangle a set of numbers ar- to models with fewer pathways. Also
rayed as a pyramid: Apart from the ones, called path modeling. See PATH DIA-
each entty is the sum of the adjacent en- GRAM.
tries above.
p a t h coefficient in PATH ANALYSIS,
1 any of a set of weights (PARTIAL REGRES-
1 1 SION COEFFICIENTS) that reflect the
strength of the different facets of the re-
1 2 1 lationships among MANIFEST VARIABLES
1 3 3 1 in a system. For example, if the model
1 4 6 4 1 indicates that A ^ B ^ C , there would be a
1 5 10 10 5 1 path coefficient for the relationship be-
tween A and B and another for the rela-
It is used to calculate probabUities and tionship between B and C; these would
binomial coefficients, which indicate be direct effects, whereas the relation-
the number of possible COMBINATIONS ship between A and C would be an indi-
for a set of objects, [named after but not rect effect, working through the impact
originated by Blaise Pascal (1623-1662), of B. The size of the weight can be tested
French mathematician] for STATISTICAL SIGNIFICANCE. If the
weight is statistically significant, it is
passive deception the withholding of generally drawn as part of the PATH DIA-
certain information from research par- GRAM, whereas if it is not, it is not in-
ticipants, such as not informing them of cluded.
the true focus of the study. As long as
certain conditions are met (e.g., DE- p a t h d i a g r a m a figure describing the
BRIEFING), this approach is generally hypothesized relations tested in an anal-
considered acceptable, because research- ysis of simultaneous LINEAR RELATION-
ers may not want to reveal their hypoth- SHIPS among measured or MANIFEST
eses to study participants in case this VARIABLES; the object is to identify the
leads themconsciously or uncon- causal relationships or logical ordering
sciouslyto adjust their behavior. Also among variables. A diagram generaUy
uses curved lines to represent un-

255
path modeling

analyzed associations. Each measured vidual elements i n a FACTOR P A T T E R N

variable is designated by a box and la- MATRIX. The direction and magnitude
tent variables are represented by ovals. of the weight help researchers to evalu-
COVARIANCES 01 Correlations between ate the factors.
EXOGENOUS VARIABLES and between er-
rors are represented by curved lines with p a t t e r n m a t r i x see F A C T O R PATTERN
arrowheads at both ends. Paths are rep- MATRIX.
resented by straight lines with an arrow- p a t t e r n - m i x t u r e m o d e l an approach
head pointing from the INDEPENDENT to analyzing L O N G I T U D I N A L D A T A or
VARIABLE toward the D E P E N D E N T V A R I -
nested data (see NESTING) in which dif-
ABLE. Consider the generic example ferent strata of cases are identified and
below. Notice that associated with each separate PARAMETER estimates obtained
path is an asterisk or a number. The as- for each. Researchers also use this ap-
terisks indicate free PARAMETERS, whose proach when accounting for missing
values wUl be estimated from the data, data, including those that are not MISS-
while the numbers indicate fixed param- ING AT RANDOM.
eters, whose values do not change as a
function of the data. Also called c a u s a l p a t t e r n v a r i a b l e a variable consisting
d i a g r a m ; p a t h m o d e l . See also PATH of two or more N O M I N A L or C A T E G O R I -
ANALYSIS. CAL items converted into a single item
p a t h m o d e l i n g see P A T H ANALYSIS.
that includes all combinations of the rel-
evant information. For two items, A and
p a t t e r n a n a l y s i s a class of methods B, that can score either correct or incor-
(e.g., CLUSTER ANALYSIS, FACTOR ANALY- rect, a single item can be created consist-
SIS, DISCRIMINANT ANALYSIS) that are ing of (a) zero items correct; (b) Item B
used by researchers to recognize and correct, Item A incorrect; (c) Item A cor-
find systematic regularity within a much rect. Item B incorrect; and (d) both items
larger field of data. Often these methods correct.
use computer modeling and simulation
approaches and involve D A T A M I N I N G , P C A abbreviation for P R I N C I P A L C O M P O -
image processing, and the study of net- NENTS ANALYSIS.
works.
P D F abbreviation for PROBABILITY D E N -
p a t t e r n c o e f f i c i e n t in EXPLORATORY SITY FUNCTION.
FACTOR ANALYSIS, One of a set of variable
WEIGHTS expressing the significance of P e a r s o n c h i - s q u a r e test see C H I -
the underlying factors: one of the indi- SQUARE TEST FOR I N D E P E N D E N C E . [Kari

path diagram
 period effect

Pearson (1857-1936), British mathema- percent error see RELATIVE ERROR.

tician]
percentile n. the location of a score in a
Pearson product-moment correla- distribution expressed as the percentage
t i o n coefficient see PRODUCT-MO- of cases in the data set with scores equal
MENT CORRELATION COEFFICIENT. [Karl to or below the score in question. Thus,
Pearson] if a score is said to be in the 90th percen-
tile this means that 90% of the scores in
Pearson's r see PRODUCT-MOMENT COR- the distribution are equal to or lower
RELATION COEFFICIENT. [Karl Pearson]
than that score. Also called percentile
penalized m a x i m u m l i k e l i h o o d rank.
estimation (PMLE) a M A X I M U M LIKE- percentile n o r m see NORM.
LIHOOD approach to the estimation of
population PARAMETERS that is particu- percent point f u n c t i o n see INVERSE
larly useful in cases where the model DISTRIBUTION FUNCTION.
may be overfitted (i.e., have too many per-comparison error rate see
parameters). For example, the accuracy TESTWISE E R R O R R A T E .
of prediction models for drug benefits
generally turns out to be lower than esti- perfect correlation a relationship be-
mated when these drugs are given to tween two variables, x and y, in which
new samples of patients. This may be the change in value of one variable is ex-
because of differences between the new actly proportional to the change
samples and those from which the pre- in value of the other. That is, knowing
dictions were generated, or because the the value of one variable exactly predicts
developed model was overfitted, or the value of the other variable (i.e., r^y =
both. The PMLE method makes an ad- 1.0). When plotted graphically, a perfect
justment for overfitting that is directly correlation forms a perfectly straight
built into the model development, in- line. If the variables change in the same
stead of relying on strategies such as direction (i.e., they both increase or
SHRINKAGE. both decrease), the correlation is perfect
positive, whereas if the variables change
penalty f u n c t i o n a WEIGHT applied to in opposite directions (i.e., one increases
a statistic or index that lowers its value as the other decreases or vice versa),
because of some undesired characteris- the correlation is perfect negative. It is
tic. In STRUCTURAL EQUATION MODEL- uncommon to observe this type of rela-
ING, for example, several indices used to tionship in acraal data. See also CORRE-
assess GOODNESS OF FIT include such a LATION COEFFICIENT.
mechanism to downweight the index
where a researcher's model exhibits a performance test any test of ability re-
high level of complexity. See also quiring primarily motor, rather than
SHRINKAGE ESTIMATOR. verbal, responses, such as a test requiring
manipulation of a variety of different
percentage n. a fraction or ratio with objects or completion of a task that in-
100 as the assumed denominator. For volves physical movement.
example, a national study of computer
use might show that a high percentage period effect any outcome associated
of adolescents (89 for every 100 sam- with living during a particular time pe-
pled) no longer use e-mail as their pri- riod or era, regardless of how old one
mary form of communication. was at the time. Certain historical
events, such as war, the introduction of
percentage of variance accounted social media, or the events occurring on
f o r see PROPORTION OF VARIANCE AC- September 11, 2001, may affect re-
COUNTED FOR. sponses of participants in research stud-

257
periodicity

ies. Period effects may be difficult to can provide basic information about the
distinguish from AGE EFFECTS and CO- underlying process that may otherwise
HORT EFFECTS in research. be obscured.
periodicity n. 1. the state of recurring period prevalence see PREVALENCE.
more or less regularly, that is, at inter-
p e r m u t a t i o n n. an ordered arrange-
vals. Phenomena that exhibit periodic-
ment of elements from a set. A permuta-
ity include the cycle of the seasons,
tion is similar to a COMBINATION but
human circadian rhythms, and con-
distinguished by its emphasis on order.
sumer spending. 2. in SYSTEMATIC SAM-
For example, if there are three colored
PLING, a problem that can arise when
objectsred (R), white (W), and blue
the process of selection (e.g., choosing
(B)there are six possible permutations
every fifth unit from a list) interacts with
of these objects: RWB, RBW, WBR, WRB,
a hidden periodic trait (e.g., gender)
BRW, and BWR.
within the population. If the sampling
technique coincides with the periodicity permutation test see RANDOMIZA-
of the trait (e.g., evety fifth person on TION TEST.
the list is a male), the sampling tech-
nique will no longer be random and the perpendicular n. a line that is at a right
representativeness of the sample is com- angle (90) to another line or plane.
promised. Typically, in plotting a graph, the y-axis
is perpendicular to the x-axis.
periodic survey an assessment, often a
personality p r o f i l e 1. a presentation
questionnaire, that is administered with
of results from psychological testing in
some regularity but not necessarily at
graphic form so as to provide a summary
equal time intervals and that generally
of a person's traits or other unique at-
includes new respondents each time. tributes and tendencies. Because various
The national census is an example of a scores appear in one display, a researcher
periodic survey with equal time inter- can see the pattern of high and low scale
vals. scores for a given person. For example, a
periodogram n. in SPECTRAL ANALYSIS profile of a client's test results from the
or TIME-SERIES ANALYSIS, a plot that Minnesota Multiphasic Personality In-
breaks down the complex sources of ventory would display his or her scores
variance and noise in the function into on measures of depression, hysteria,
component elements. A periodogram masculinity/femininity, paranoia, social

8 50 -
(A

1 1 1 1 1~
Depression Hysteria Masculinity/ Paranoia Introversion
femininity

Attribute

personality profile
 Phillips-Perron test

introversion, and other attributes (as phenomenology n. 1. the scientific

shown on the previous page). 2. a sum- srady of human consciousness, percep-
maty of traits and behavioral tendencies tions, and experience. 2. a movement in
that are believed to be typical of a partic- modern philosophy initiated by German
ular group or category, of individuals philosopher Edmund Husserl (1859-
(e.g., people with a particular disorder, 1938), who argued for an approach to
people employed in a particular profes- human knowledge in which both the tra-
sion). ditional concerns of philosophy and the
modern concern with scientific causa-
personality study the study of the dy- tion would be set aside in favor of careful
namic stracrare and processes that de- attention to the narare of immediate
termine individuals' emotional and conscious experience. The movement
behavioral adjustments to their environ- suggested that mental events should be
ment. Such research has focused on the sradied and described in their ovm terms,
structure of major trait constructs, on rather than in terms of their relationship
how traits affect actions and selection of to events in the body or in the external
social settings, and on the stabiUty of world. However, phenomenology should
processes. For example, a researcher be distinguished from introspection as it
might be interested in understanding is concerned with the relationship be-
how conscientiousness manifests itself tween acts of consciousness and the ob-
from adolescence to late adulthood. jects of such acts. Husserl's approach
personal p r o b a b i l i t y see SUBJECTIVE proved widely influential in psychology
PROBABILITY.
and the social sciences. phenomeno-
logical adj. phenomenologist n.
person t i m e 1. the sum of all of the
years, months, days, and hours that all p h i coefficient (symbol: ^) a measure
study participants have contributed to a of association for two dichotomous or
piece of research. Usually the partici- binaty RANDOM VARIABLES. The phi coef-
pants are an at-risk group for a certain ficient is the PRODUCT-MOMENT CORRE-
disease and this value is used to calculate LATION COEFFICIENT when both variables
an incidence rate without any adjust- are coded (0, 1). For example, the phi co-
ment for dropouts in a srady. When efficient could be used to examine the
years are the unit of interest, the value is relationship between gender (male [0]
called person-years at risk. 2. the total and female [1]) and left- (0) or right-
amount of researcher time contributed handedness (1). A TWO-BY-TWO TABLE
to a project. For example, if there are two could be constmcted to record the fre-
researchers who each devote 50 hours to quency of people with a 0 on both vari-
a study, the project involves 100 hours ables (i.e., left-handed males), a 1 on both
of person time. variables (i.e.,right-handedfemales), a 0
on the first variable and a 1 on the other
P G F abbreviation for PROBABILITY GEN- (i.e., right-handed males), and vice versa
ERATING FUNCTION. (i.e., left-handed females). The pairs of re-
sponses on the variables could then be
phase I study in CLINICAL TRIALS, an analyzed and a phi coefficient calculated
initial study in which a new drug or to determine whether any relationship
treatment that showed promise in other exists. The coefficient is interpreted in
settings is tested for basic safety on a the same way as other coefficients of cone-
small sample of patients. Should it pass lation. Also called fourfold point corre-
this test, subsequent phases of the test- lation coefficient. See also CRAMER'S V;
ing process will evaluate whether the
TETRACHORIC CORRELATION COEFFICIENT.
drug or treatment works and whether it
is better than the standard of care. P h i l l i p s - P e r r o n test see UNIT ROOT

259
phone survey

TEST. [Peter C. B. Phillips (1948- ), Brit- and meet at nodal points to form a con-
ish econometrician; Pierre Perron tinuous line. This is an approach that is
(1959- ), Canadian econometrician] vety useful for identifying abrapt changes
phone survey see TELEPHONE INTER-
or discontinuities in a process. For exam-
VIEW.
ple, piecewise regression might be ap-
propriate where a researcher is studying
physical determinism see STATISTI- a group of participants before treatment,
CAL DETERMINISM. during treatment, and after treatment.
physiological measure any of a set of Also called segmented regression.
instraments that convey precise infor- pie chart a graphic display in which a
mation about an individual's bodily circle is cut into wedges with the area of
functions, such as heart rate, skin con- each wedge being proportional to the
ductance, skin temperature, Cortisol percentage of cases in the category repre-
level, palmar sweat, and eye tracking. sented by that wedge. For example, a re-
Studies using these measures typically searcher might present the results of a
obtain measurements before and after survey on sources of psychology-related
the introduction of a stimulus condition literature used by the public in the man-
as a way to document an individual's re- ner shown below.
sponse to that stimulus. It generally works best when there are
p i n. (symbol: TC) 1. a ratio expressing the not many categories (with thin wedges)
circumference of a circle to its diameter, being shown. A downside of the graphic
given as 3.141592. 2. a symbol denoting is that it is not very efficient because it
the probability of success on a trial, used uses significant space to show the fre-
often in ODDS RATIO calculations as well quencies of a single variable.
as in PROBIT ANALYSIS and LOGIT M O D -
P i l l a i - B a r t l e t t trace a statistic used in
ELS.
MULTIVARIATE ANALYSES OF VARIANCE
piecewise regression a variant on or- and other procedures to determine the
dinary LEAST SQUARES REGRESSION In STATISTICAL SIGNIFICANCE of differ-
which a REGRESSION LINE consisting of ences among levels of the INDEPENDENT
several different lines isfittedto the data. VARIABLES while simultaneously taking
The several pieces have different slopes into account multiple DEPENDENT VARI-

Academic sources 6%

Popular magazines 10%

Advice columns 35%

Self-help books 19%

Internet 30%

pie chart
 PMP

ABLES. It is considered to be a robust sta- an;inert substance (placebo) instead of

tistic, even when assumptions of the active drug under investigation,
MULTIVARIATE NORMALITY, HOMOGENE- thus functioning as a C O N T R O L G R O U P
ITY OF VARIANCE, and COVARIANCE are against which to make comparisons re-
not entirely met. Also called P i l l a i ' s garding the effects of the active drug. See
trace. See also H O T E L L I N G ' S T R A C E ; also P L A C E B O EFFECT.
WILKS'S L A M B D A . [K. C. Sreedharan Pillai
p l a c e b o c o n t r o l l e d t r i a l a cUnical re-
(1920-1985), Indian statistician; M . S. search design that incorporates a PLA-
Bartlett (1910-2002), British statistician] CEBO CONTROL GROUP. Compare
pilot study a small, preliminaty srady A C T I V E C O N T R O L TRIAL.
designed to evaluate procedures and p l a c e b o e f f e c t a clinically significant
measurements in preparation for a sub- response to a therapeutically inert sub-
sequent, more detailed research project. stance or nonspecific treatment (PLA-
Although pilot studies are conducted to CEBO), deriving from the recipient's
reveal information about the viability of expectations or beliefs regarding the
a proposed project and implement nec- intervention. It is now recognized that
essary modifications, they may also pro- placebo effects accompany the adminis-
vide useful initial data on the topic of tration of any drug (active or inert) and
study and suggest avenues or offer impli- contribute to the therapeutic effective-
cations for future research. Also called ness of a specific treatment. For exam-
feasibility study; pilot experiment; ple, patients given a placebo to relieve
pilot research; pilot test. headaches may report statistically signif-
icant reductions in headaches in sradies
pivotal variable a quantity that is a
that compare them with-patients who
function of a statistic (e.g., sample mean,
receive no treatment at all. This term is
sample size) and a parameter (e.g., popula-
also more generally used in nonclinical
tion mean, variance) and is used in INFER-
studies to indicate any effect arising
ENTIAL TESTS based on a NORMAL DISTRI-
from participants' expectations regard-
BUTION. The z SCORE, T SCORE, and F
ing the study.
RATIO are all examples of pivotal vari-
ables. Also called pivot; pivotal quan- p l a c e b o g r o u p see P L A C E B O C O N T R O L
tity. GROUP.
p l a c e b o n. (pl. placebos) 1. a pharmaco- p l a n n e d c o m p a r i s o n see A PRIORI
logically inert substance, such as a sugar COMPARISON.
pill, that is often administered as a C O N - p l a t y k u r t i c adj. describing a distribu-
TROL in testing new drugs. Foriherly, tion of scores that is flatter than a NOR-
placebos were occasionally used as diag- M A L DISTRIBUTION, as in the general
nostic or psychotherapeutic agents, for depiction overleaf.
example, in relieving pain or inducing Thus, it has more scores at the ex-
sleep by suggestion, but the ethical im- tremes and fewer in the center. See also
plications of deceiving patients in such MESOKURTIC; LEPTOKURTIC.
fashion makes this practice problematic.
2. any medical or psychological inter- P L S abbreviation for partial least squares.
vention or treatment that is believed to See PARTIAL LEAST SQUARES REGRESSION.
have no effect, thus making it valuable pluralism n. the idea that any entity has
as a control condition against which to many aspects and that it may have a va-
compare the intervention or treatment riety of causes and meanings. plural-
of interest. See also P L A C E B O EFFECT. ist adj.
p l a c e b o c o n t r o l g r o u p a group of P M F abbreviation for PROBABILITY MASS
participants i n a study who receive FUNCTION.

261
PMLE

using a random process. The selected

cases may be used for research purposes
or to estimate the average quantity that
exists in a given area of the space.
p o i n t scale any scale for measuring
some constract or attribute in which
participants' responses to a series of mul-
tiple-choice questions are given numeri-
cal values (points). For example, a point
scale with the values 1 (poor) through 5
(excellent) might be used to assess an em-
ployee's job performance; See also
platykurtic LIKERT SCALE; SEMANTIC DIFFERENTIAL.
P M L E abbreviation for PENALIZED MAXI- Poisson d i s t r i b u t i o n a DISCRETE DIS-
M U M LIKELIHOOD ESTIMATION. TRIBUTION that generates the probabil-
ity of occurrence of rare events that are
point biserial correlation coeffi-
randomly distributed in time or space. It
cient (symbol: rpj,is; r^^,) a numerical
is a type of THEORETICAL DISTRIBUTION.
index reflecting the degree of relation-
Count variables, such as the number of
ship between two random variables, one
aggressive acts observed in a playground
CONTINUOUS and one dichotomous (bi-
or the number of times a person at-
nary). An example is the association be-
tempts to stop smoking, take this form.
tween the propensity to experience an
When a variable is distributed in this
emotion (measured using a scale) and
way, a researcher needs to consider spe-
gender (male or female). The coefficient
cialized statistical models. [Simeon D.
is interpreted in the same way as other
Poisson (1781-1840), French mathema-
C O R R E L A T I O N COEFFICIENTS. AlsO c a l l e d
tician]
point biserial; point biserial r.
Compare RANK BISERIAL CORRELATION Poisson process a STOCHASTIC PRO-
COEFFICIENT. CESS that governs the number of times
p o i n t estimate a single estimated nu- that specific data are observed or
merical value, determined from a sam- counted within a specified period of
ple, of a given population PARAMETER. time. In a study about communication
Point estimates can often be tested for among committed partners, for exam-
STATISTICAL SIGNIFICANCE if their STAN- ple, researchers might count the number
DARD ERRORS are also known. Compare of times that one member of the couple
INTERVAL ESTIMATE. interrupts the other within a 10-minute
problem-solving task. The distribution
p o i n t prevalence see PREVALENCE. of a Poisson process at time t is the POIS-
p o i n t process a random variable whose SON DISTRIBUTION. [Simeon D. Poisson]
values are the locations in space or time Poisson regression an analytic model
at which a given type of event occurs. that has as its dependent variable a
Point process analyses allow for model- count variable and any of a number of
ing the occurrence of such events, for ex- predictor variables; in other words it is
ample when and where earthquakes used when the outcome variable has a
occur or when and where bullying oc- POISSON DISTRIBUTION. For example, a
curs in a school setting. researcher may hypothesize that dyadic
p o i n t s a m p l i n g selecting units or cases adjustment, number of years as a couple,
from a region in space (e.g., on a com- and personality characteristics predict
puter screen, from a geographic region) the number of times that one member of

262
 polynomial regression

a couple interrapts the other during a egories. For example, the quality of a
problem-solving task. See also ZERO- training program could be described as
INFLATED POISSON REGRESSION. [Simeon polychotomous: poor, fair, good, or ex-
D. Poisson] cellent. Also called polytomous. See
also DICHOTOMOUS VARIABLE.
Poisson variable a study variable that
is described by a count distribution polychotomous variable a variable
within a particular time period. See POIS- having more than two possible catego-
SON DISTRIBUTION. [Simeon D. Poisson] ries, either ordered or unordered. For ex-
ample, college matriculation could be
policy analysis a collection of tech-
described as a polychotomous variable:
niques for synthesizing information (a)
freshman, sophomore, junior, or senior.
to specify alternative policy and pro-
Also called polytomous variable. See
gram choices in cost-benefit terms, (b)
to assess organizational goals in terms of also DICHOTOMOUS VARIABLE.
input and outcome, and (c) to provide p o l y n o m i a l n. a mathematical expres-
a guide for future decisions concerning sion consisting of multiple terms, each
research activities. Policy analysis is of which is the product of a constant (a)
a useful approach for examining the im- and a variable (x) raised to a whole num-
plications of different potential scenar- ber exponent:
ios. For example, an organization n-l
considering aflexibletime option for its + a_ix + a
employees might choose to engage in See also BINOMIAL.
this process before implementing the
change. p o l y n o m i a l contrast a comparison of
mean values for more than two different
policy research empirical studies con- levels or time points of an INDEPENDENT
ducted to guide the formulation of cor- VARIABLE to determine whether they
porate or public policies. For example, a follow a particular mathematical pat-
nonprofit organization might investi- tern, such as linear, quadratic, cubic, or
gate how to help students from lower in- quartic. For example, a researcher might
come urban areas gain access to higher have a specific prediction about how cli-
education. ents' average degree of motivation to at-
tend therapy will differ over the course
polychoric correlation coefficient
of treatment (e.g., will decrease, then in-
an index showing the degree of associa-
crease, and then decrease again), and he
tion between two variables that are
or she could use a polynomial contrast
scored as ordered categories but assumed
to be manifestations of underlying CON- to determine whether this is indeed the
TINUOUS RANDOM VARIABLES. In Other
case. See also REPEATED CONTRAST.
words, it is an alternative to the PROD- p o l y n o m i a l f u n c t i o n a mathematical
UCT-MOMENT CORRELATION COEFFI- function that relates one number, quan-
CIENT that is applied to ordered tity, or entity to a POLYNOMIAL. For ex-
POLYCHOTOMOUS VARIABLES. For exam- ample, y = 2x' + 4^" + x i s a polynomial
ple, a researcher might use a polychoric function.
correlation coefficient to assess the rela-
tionship between teacher ratings of stu- p o l y n o m i a l regression atypeof LIN-
EAR REGRESSION analysis in which the
dent interest in a topic (none, moderate,
substantial) and student self-reports of relationship between an explanatory or
INDEPENDENT VARIABLE and an outcome
interest.
or DEPENDENT VARIABLE Is modeled as a
polychotomous adj. describing a single POLYNOMIAL. Polynomial regression is
measurement or item scored with more appropriate for expressing higher order
than two unordered or ordered cat- NONLINEAR effects among variables (via

263
polynomial trend

quadratic or ciibic functions such as y, = characteristics of adults enrolled in dis-

Po + ?>\Xi + ^2^,^ + ^3X,^ + ...) whUe retain- tance learning classes, who are nested
ing all other aspects of linear ap- within families, which are further nested
proaches. within neighborhoods and communi-
ties, and so forth. A population-averaged
p o l y n o m i a l t r e n d in TIME-SERIES
model would provide information about
ANALYSIS, a systematic pattern in the
distance learners as a single entity,
data that may be characterized by a
whereas a SUBJECT-SPECIFIC MODEL
POLYNOMIAL.
would fociis on change in each distance
polytomous adj. see POLYCHOTOMOUS. learner individually, and a UNIT-
SPECIFIC MODEL would describe the dis-
polytomous variable see POLY-
tinct groups of learners (e.g., at the fam-
CHOTOMOUS VARIABLE.
ily level, neighborhood level).
pooled estimate a single estimated p o p u l a t i o n correlation coeffi-
value for a population PARAMETER (e.g., cient (symbol: p) an index expressing
a mean) obtained by averaging several the degree of association between two
independent estimates of that parame- continuously measured variables for a
ter. complete POPULATION of interest. For
p o o l i n g n. a procedure in which several example, a researcher could obtain in-
independent estimated values of a popu- come and education information for all
lation characteristic are averaged, with families in a town and calculate a popu-
or without WEIGHTS, to obtain a single lation correlation coefficient for the en-
value. For example, the pooled variance tire town. In contrast, the SAMPLE
is a single value for a variable's DISPER- CORRELATION COEFFICIENT indexes the
SION produced by combining several in- association for a specific subset of those
dependent estimates of that dispersion. cases (e.g., evety fourth family from a list
Consider a researcher examining corre- of all those in the town).
lations between children's educational p o p u l a t i o n d i s t r i b u t i o n the DISTRI-
performance and maternal depression BUTION of scores or responses on a vari-
who uses MULTIPLE IMPUTATION to fUl able for a complete POPULATION of
in missing data. He or she could calcu- interest. For example, a researcher could
late the value of the VARIANCE using the graph the income distribution of all
data set that exists at each stage of the families in a town; in contrast, the SAM-
imputation procedure, and then average PLE DISTRIBUTION would show the in-
each of those individual variances to ob- comes of only a specific subset of
tain a single overall value. families.
population n. a theoretically defined, p o p u l a t i o n mean (symbol: n) the av-
complete group of objects (people, ani- erage (MEAN) value on a variable for a
mals, institutions) from which a SAMPLE complete POPULATION of interest. In
is drawn to obtain empirical observa- many research settings this value is esti-
tions and to which results can be gener- mated using the SAMPLE MEAN but in sit-
alized. Also called universe. uations when information is known for
population-averaged model in the all the units of interest it can be calcu-
analysis of LONGITUDINAL DATA, a lated directly. For example, the mean
model that specifies a MARGINAL DISTRI- household income for an entire town
BUTION, or averaged pattern of re- may be determined by averaging the re-
sponses across cases, rather than a sponses from a survey returned by all of
the households in the town.
pattern conditional on a person-specific
or cluster-specific effect. For example,
consider a researcher investigating the
264
 positivism

p o p u l a t i o n p a r a m e t e r see P A R A M E - expect to find a positive correlation

TER. between srady hours and test perfor-
mance. Also called direct correlation.
p o p u l a t i o n s a m p l i n g see S A M P L I N G . Compare N E G A T I V E C O R R E L A T I O N .
population standard deviation p o s i t i v e d e f i n i t e describing a S Y M -
(symbol: o) a value indicating the DIS- M E T R I C A L M A T R I X whose E I G E N V A L U E S
PERSION of scores in a complete P O P U L A -
are all positive (i.e., > 0). In many model-
TION of interest, that is, how narrowly or ing approaches, if a matrix does not pos-
broadly the scores deviate from the sess this quality then the analysis cannot
MEAN. In many research settings the be conducted without model respecifi-
population standard deviation is esti- cation.
mated from the sample STANDARD DEVI-
ATION, but when information about the p o s i t i v e p r e d i c t i v e p o w e r a calcula-
full set of units is known it can be calcu- tion of the probability that someone
lated directly. See also P O P U L A T I O N identified as having a particular condi-
VARIANCE. . tion by a specific test does in fact have
that condition. It is the proportion of all
p o p u l a t i o n v a l i d i t y the degree to people identified as having the condi-
which study results from a sample can tion who were correctly identified. Thus,
be generalized to a larger target group of the positive predictive power of a new
interest (the POPULATION). For example, mammography technology would be
consider an educational researcher wish- determined by dividing the percentage
ing to characterize the academic motiva- of women whose results correctly show
tion of sradents. If he or she conducts a they have breast cancer (i.e., valid
srady of only urban students then positives) by the percentage of women
the investigator would need to demon- whose results are positives (i.e., both
strate that the findings apply beyond valid positives and false positives). See
the sample used in the study to the also SENSITIVITY. Compare NEGATIVE
broader set of all students as a whole.
PREDICTIVE POWER.
Population validity is a type of EXTER-
NAL VALIDITY. p o s i t i v e r e l a t i o n s h i p see DIRECT RE-
LATIONSHIP.
p o p u l a t i o n v a r i a n c e (symbol: o^) the
square of the P O P U L A T I O N STANDARD p o s i t i v e s e m i d e f i n i t e describing a
DEVIATION. It is a measure reflecting the SYMMETRICAL MATRIX whOSe EIGEN-
spread (DISPERSION) of scores in a com- VALUES can include zeros as well as posi-
plete POPULATION of interest (e.g., the tive numbers. See also POSITIVE DEFINITE.
test scores of all students at a school).
p o s i t i v e s k e w see SKEWNESS.
positive b i n o m i a l distribution a
p o s i t i v i s m n. a family of philosophical
theoretical set of discrete values that de-
positions holding that all meaningful
scribes the probability of different num-
propositions must be reducible to sen-
bers of successes occurring in a sequence
soty experience and observation, and
of BERNOULLI TRIALS, such that there is
thus that all genuine knowledge is to
zero probability of obtaining no suc-
be built on strict adherence to empirical
cesses and all other probabilities are
methods of verification. Its effect is to es-
scaled to add to one. Compare NEGATIVE
tablish science and the scientific method
BINOMIAL DISTRIBUTION. See BINOMIAL
as the model for all forms of valid in-
DISTRIBUTION.
quity and to dismiss the trath claims of
p o s i t i v e c o r r e l a t i o n a relationship religion, metaphysics, and speculative
between two variables in which both rise philosophy. Positivism, particularly L O G -
and fall together. For example, one would ICAL POSITIVISM, was extremely influen-

265
postdictive validity

tial in the early development of psy- comparison (or contrast); post hoc
chology and helped to form its commit- contrast; unplanned comparison
ment to empirical methods. Despite the (or contrast).
advent of POSTPOSITIVISM, it continues
to be a major force in contemporaty psy- post hoc power see RETROSPECTIVE
POWER.
chology. positivist adj.
postdictive v a l i d i t y see RETROSPEC- post hoc test a statistical procedure
TIVE VALIDITY. conducted on the basis of the findings
obtained from previous analyses. Most
posterior d i s t r i b u t i o n in BAYESIAN commonly, the phrase refers to compari-
approaches, an estimated distribution of sons of the mean values obtained on a
values for a population characteristic of variable by different study groups that
interest that is obtained by combining are made only after an overall ANALYSIS
empirical data with prior expectations OF VARIANCE has revealed a significant F
based on existing knowledge or opinion RATIO, indicating that there is some ef-
(the PRIOR DISTRIBUTION). fect or difference across the groups that
should be examined further. Also called
posterior p r o b a b i l i t y the probability follow-up test. See POST HOC COM-
that a certain event will occur given that PARISON.
other related events have been observed.
For example, assume that Basket A con- postpositivism n. 1. the general posi-
tains 10 white cards and 20 red cards, tion of U.S. psychology since the mid-
while Basket B contains 20 white cards 20th centuty, when it ceased to be domi-
and 10 red cards. If one selects a card nated by LOGICAL POSITIVISM, HYP
from a basket at random, without noting THETICO-DEDUCTIVE METHODS,
the color, the probability that it will OPERATIONALISM. Postpositivistic ps
be from Basket A is .5, since the only chology is a broader and more human
possibilities are A or B. If, however, endeavor, influenced by such develop-
the color of the card is noted then ments as social constructionism and
one must take this additional informa- PHENOMENOLOGY. 2. more generally,
tion into account in determining the approach to science and the philosophy
probability that the card came from Bas- of science that has moved away from a
ket A. For example, the probability that position of strict POSITIVISM. post-
a white card will be pulled from Basket A positivist adj., n. postpositivistic
would be revised downward fiom .5 adj.
according to BAYES THEOREM: Since
there are fewer white cards in A, a white posttest 1. n. an assessment carried out
card is more likely to be drawn from B after the application of some interven-
instead. tion, treatment, or other condition to
measure any changes that have oc-
post hoc comparison any examina- curred. Posttests often are used in re-
tion in which two or more quantities are search contexts, in conjunction with
compared after data have been collected PRETESTS, to isolate the effects of a vari-
and without prior plans to carty out able of interest. For example, in a study
the particular comparison. For example, examining whether a new therapy helps
after obtaining a significant F RATIO for a to alleviate depression, participants
data set, a researcher may perform post might receive the therapy and then
hoc comparisons to follow up on and complete a short symptom inventory,
help explain the initial findings. Differ- the results of which would be compared
ent statistical tests are required for post to those from an inventory taken prior
hoc comparisons than for A PRIORI COM- to the treatment. 2. vb. to administer a
PARISONS. Also called a posteriori posttest.
 practical significance

posttest-only control-group de- it is expressed by the equation y = ax ,

sign a research design in which an EX- where x and y are the variables and a and
PERIMENTAL GROUP and a CONTROL b ate numerical constants. Power func-
GROUP are compared on a POSTTEST tions have been used to characterize the
measure only. It is assumed that partici- scales relating perceived and physical in-
pants' results on any PRETEST adminis- tensity, for example, as well as to charac-
tered before the introduction of the terize the relationship between response
experimental manipulation would be es- speed and practice. 2. a formula relating
sentially equivalent across the groups different factors, such as sample size, EF-
due to the RANDOM ASSIGNMENT of indi- FECT SIZE, and SIGNIFICANCE LEVEL, to
viduals to conditions. A posttest-only the likelihood that use of a particular
control-group design is distinct from statistical procedure will lead to rejec-
a posttest-only design, in which all tion of the NULL HYPOTHESIS when it is
groups receive some treatment and there in fact false (see POWER). For example, a
is no neutral comparison. researcher may plan to use a specific sta-
tistical test to detect a medium-sized ef-
power n. 1. a measure of how effective a fect, evaluate the effect at the .001
statistical procedure is at identifying real significance level, and reach a desired
differences between populations: It is statistical power level of .80. Using a
the probability that use of the test will power function, the researcher could de-
lead to the NULL HYPOTHESIS of no effect termine the sample size needed under
being rejected when the ALTERNATIVE those conditions. Power functions may
HYPOTHESIS is true. For example, if a
be presented in tabular form or plotted
given statistical test has a power of .70, graphically as power curves.
then there is a 70% probability that its
use will result in the null hypothesis cor- power t r a n s f o r m a t i o n a class of
rectly being rejected as false, with a cor- TRANSFORMATIONS uscd to make the
responding 30% chance that its use will distribution of values on a variable more
lead to a TYPE II ERROR. Power ranges closely resemble that of a NORMAL DIS-
from 0 to 1, with values of .80 or above TRIBUTION and to have more stable
generally considered acceptable. Re- VARIANCE. In this approach, the ob-
searchers try to maximize power in their tained values are raised to some expo-
srady designs prior to data gathering. nential quantity (e.g., squared), which
Also called statistical power. 2. a preserves their original order. The BOX-
mathematical notation that indicates cox TRANSFORMATION is an example of
the number of times a quantity is multi- a power transformation.
plied by itself.
P-P plot abbreviation for probability-
power analysis the process of deter- probability plot. See PROBABILITY PLOT.
mining the number of cases or observa-
p r a c t i c a l significance the extent to
tions that a study would need to achieve
which a study result has meaningful ap-
a desired level of POWER with a certain
plications in real-world settings. An ex-
EFFECT SIZE and a certain SIGNIFICANCE
perimental result may lack STATISTICAL
LEVEL. This information is particularly
important because it enables a re- SIGNIFICANCE or show a small EFFECT
searcher to plan what resources will be SIZE and yet potentially be important
needed to enroll or select the desired nonetheless. For example, consider a
number of individuals. srady showing that the consumption
of baby aspirin helps prevent heart at-
power f u n c t i o n 1. a relationship in tacks. Even if the effect is small, the find-
which the values for one variable vaty ing may be of practical significance if it
according to the values of another vari- saves lives over time. Also caUed sub-
able raised to a power. In mathematics. stantive significance. See also CLINI-

267
practice effect

CAL SIGNIFICANCE; PSYCHOLOGICAL bad. See also INSTRUMENTALISM. ^prag

SIGNIFICANCE. mafist adj., n.
practice effect any change or improve- PRE abbreviation for PROPORTIONAL RE-
ment that results from practice or repeti- DUCTION OF ERROR.
tion of task items or activities. The
precision n. a measure of accuracy. In
practice effect is of particular concern
statistics, an estimate with a small STAN-
in experimentation involving WITHIN-
DARD ERROR is regarded as having a high
SUBJECTS DESIGNS, as participants' per-
degree of precision. precise adj.
formance on the variable of interest may
improve simply from repeating the ac- precision m a t r i x a SQUARE MATRIX that
tivity rather than from any study manip- is the RECIPROCAL of the COVARIANCE
ulation imposed by the researcher. MATRIX. Each element in a precision ma-
trix thus is a PARTIAL CORRELATION,
practice t r i a l the first of a series of op-
showing the relationship between two
portunities to respond to a test or an ex-
variables with the influence of all other
perimental setting, which is given to
variables removed.
participants to acquaint them with the
procedures used and is therefore not predicted value (symbol: y) in REG-
scored. The use of practice trials is be- RESSION ANALYSIS, the estimated value
lieved to ensure that the initial measure- of the outcome or DEPENDENT VARIABLE
ments of the study proper will provide (y) for a case in a data set, as calculated
more accurate assessments of partici- from the specified mathemafical model o
pants' performance. the relationship between predictor vari-
ables (i.e., from the REGRESSION EQUA-
pragmatic analysis 1. in a CLINICAL TION). The differences between predicted
TRIAL, an analysis that seeks to deter- values and observed data are known as RE
mine which treatments will be benefi-
SIDUALS. Also called Y hat.
cial to individuals in real-life settings,
rather than how or why a treatment is prediction n. an attempt to estimate,
effective, as in an EXPLANATORY ANALY- forecast, or project what will happen in
SIS. 2. a loose synonym of INTENTION- a particular case, generally on the basis
TO-TREAT ANALYSIS. of past instances or accepted principles.
In science, the use of prediction and ob-
pragmatic t r i a l see EXPLANATORY servation to test hypotheses is a corner-
TRIAL. stone of the empirical method. See also
pragmatism n. a philosophical posi- STATISTICAL PREDICTION. predict
tion holding that the truth value of a predictable adj. predictive adj.
proposition or a theory is to be found in
predictive criterion-related valid-
its practical consequences: If, for exam-
ity see PREDICTIVE VALIDITY.
ple, the hypothesis of God makes people
happy, then it may be considered true. predictive discriminant analysis
Although some forms of pragmatism a multivariate procedure that distin-
emphasize only the material conse- guishes between two or more categories
quences of an idea, more sophisticated of a future outcome on the basis of sev-
positions recognize conceptual and eral current assessments of predictor or
moral consequences. Arguably, all forms explanatory variables. More specifically,
of pragmatism tend toward RELATIVISM, the technique combines variables mea-
because they can provide no absolute sured at one point in time to produce
groundsonly empirical groundsfor the highest possible association with dif-
determining truth and no basis for ferent levels or categories of an outcome
judging whether the consequences in measured at a subsequent point in time.
question are to be considered good or For example, a researcher might measure

268
 pretest

daily amounts of exercise, cholesterol predictive value 1. the ability of a test

levels, and average stress and use this in- or scale to correctly classity items into
formation to predict which individuals mutually exclusive categories or states.
will have a diagnosis of serious, mild, or Predictive value' often is determined
no heart disease six months later. Com- using DECISION THEORY methods, such
pare DESCRIPTIVE DISCRIMINANT ANALY- as RECEIVER-OPERATING CHARACTERIS-
SIS. TIC CURVES. 2. the percentage of times
that the obtained value on a given test is
predictive e f f i c i e n c y 1. the number the trae value. Also called predictive
or proportion of correct predictions that power. See also NEGATIVE PREDICTIVE
can be made from a particular test or POWER; POSITIVE PREDICTIVE POWER.
model. 2. the gain in utility (explana-
toty information) obtained by adding predictor variable a variable used to
more predictor or independent variables estimate, forecast, or project future
to a REGRESSION EQUATION. For exam- events or circumstances. In personnel
ple, the accuracy of a sradent's predicted selection, for example, predictors such
college grade point average may be in- as qualifications, relevant work experi-
creased if one considers high school ence, and job-specific skiUs (e.g., com-
grade point average in addition to Scho- puter proficiency, ability to speak a
lastic Assessment Test score. particular language) may be used to esti-
mate an applicant's furare job perfor-
predictive research empirical re- mance. In REGRESSION ANALYSIS and
search concerned with forecasting fu- other models, predictor variables are in-
ture events or behavior: the assessment vestigated to assess the strength and di-
of variables at one point in time so as to rection of their association with an
predict a phenomenon assessed at a later outcome, criterion, or DEPENDENT VARI-
point in time. For example, a researcher ABLE. When more than one variable is
might collect high school data, such as included in the analysis, the relative
grades, extracurricular activities, teacher weights of each in determining the out-
evaluations, advanced courses taken, come are evaluated as well. This term
and standardized test scores, in order sometimes is used interchangeably with
to predict such college success measures INDEPENDENT VARIABLE. Also called
as grade point average at graduation, predictor.
awards received, and likelihood of
pursuing further education. Compare EX- preexperimental design a research
PLANATORY RESEARCH. design or study with no CONTROL
GROUP and no random assignment of
predictive v a l i d i t y evidence that a participants to experimental conditions.
test score or other measurement corre- Examples include the ONE-SHOT CASE
lates with a variable that can only be as- STUDY and the ONE-GROUP PRETEST-
sessed at some point after the test has POSTTEST DESIGN. Although such a de-
been administered or the measurement sign is of minimal value in establishing
made. For example, the predictive valid- causality, it may be used when it is not
ity of a test designed to predict the onset possible to identity an appropriate con-
of a disease would be strong if high test trol and circumstances prohibit ran-
scores were associated with individuals domization.
who later developed that disease. It is
one of threetypesof CRITERION VALID- pre-post design see PRETEST-POSTTEST
ITY. Also called predictive criterion- DESIGN.
related validity; prospective valid- pretest 1. n. an initial assessment de-
ity. See also CONCURRENT VALIDITY; signed to measure existing characteris-
RETROSPECTIVE VALIDITY. tics (e.g., knowledge, ability) before

269
pretest-posttest design

some intervention, condition, manipu- Pretest sensitization is important to con-

lation, or treatment is introduced. Pre- sider in the interpretation of research
tests often are given to research findings, as it may make a treatment ap-
participants before they take part in a pear more effective than it actually is.
study. For example, in a study examin- Also called pretest-treatment inter-
ing whether training helps math perfor- action.
mance, participants might be ad-
ministered a short math test to assess prevalence n. the total number or per-
their original knowledge prior to under- centage of cases (e.g., of a disease or dis-
going the training. See also POSTTEST. 2. order) existing in a given population,
n. a test administered before the main either at a given time (point prevalence)
study to ensure that participants under- or during a specified period (period
stand the instructions and procedures. prevalence). For example, health re-
See PILOT STUDY; PRACTICE TRIAL. 3. vb. searchers may want to investigate the
to administer a pretest. prevalence of a new disease in an area,
whereas education researchers may be
pretest-posttest design a research interested in the prevalence of bullying
design in which the same assessment or cheating among students of a certain
measures are given to participants both age group.
before and after they have received a
treatment or been exposed to a condi- prevention research research di-
tion so as to determine if there are any rected toward finding interventions to
changes that could be attributed to the reduce the likelihood of future pathol-
treatment. The A-B DESIGN and A-B-A-B ogy (e.g., cancer, substance abuse). Such
DESIGN are examples. A more complete
research often concentrates on individu-
version in which participants are ran- als or populations considered to be par-
domly assigned to a treatment group or ticularly at risk of developing a
a CONTROL GROUP is a pretest-posttest condition, disease, or disorder. It in-
control-group design: All individuals volves one or more of the following: (a)
are assessed at the beginning of the analyzing risk and protective factors and
study, the intervention is presented to assessing susceptibility; (b) identifying
the treatment group but not the control, markers for those at risk and developing
and then all individuals are measured screening methods; (c) developing and
again. The presence of the control group implementing interventions to promote
allows the researcher to identify any pre- health and prevent the disease; and (d)
existing disparities between the groups conducting prevention analysis, the
and thus more definitely attribute differ- methodological and statistical evalua-
ences between the pre- and posttest tion of the effects of such interventions.
scores to the treatment of interest. Also prevention t r i a l a clinical study in
called before-after design; pre-post which a new intervention to avert a par-
design. ticular disease or disorder is evaluated or
compared with an existing intervention.
pretest sensitization an effect in For example, a researcher might investi-
which the administration of a PRETEST gate whether daily engagement in
affects the subsequent responses of a cognitively challenging tasks (e.g., cross-
participant to experimental treatments. word puzzles) is more effective than
For example, if a researcher administers medication in preventing Alzheimer's
a mood measure prior to treatment con- disease. If so, standards of preventive
ditions, participants may realize that the care for Alzheimer's could be changed to
study is about mood and be more at- benefit older adults.
tuned to the mood induction occurring
as part of the experimental condition. p r i m a r y data 1. information cited in a

270
 principle of beneficence

srady that was gathered directly by the nical features. Also called principal-
researcher, from his or her own experi- components factor analysis.
ments or from first-hand observation.
principal components regression
Compare SECONDARY DATA. 2. original
a prediction model that uses a set of
experimental or observational data, that
uncorrelated variables obtained from a
is, RAW DATA.
PRINCIPAL COMPONENTS ANALYSIS aS
p r i m a r y f a c t o r see FIRST-ORDER FAC- predictor or INDEPENDENT VARIABLES.
TOR. The benefit of this approach is that the
original set of predictors in the model
p r i m a r y s a m p l i n g u n i t an item or may have been so highly interrelated as
case chosen in the initial stage of a to result in COLLINEARITY. The drawback
multistep process for selecting elements is that if the uncorrelated variables are
for study from a larger group. For exam- not interpretable then the problem of
ple, in a national srady of medical collinearity is not solved.
school students, the first major step
might be choosing a random pool of p r i n c i p a l f a c t o r analysis an ap-
medical schools from among those proach to identifying the dimensions
available throughout the countty; each underlying associations among a set of
medical school so identified is a primaty variables using a COVARIANCE MATRIX of
sampling unit. Compare SECONDARY estimated COMMUNALITIES as input.
Principal factor analysis assumes that all
SAMPLING UNIT; TERTIARY SAMPLING
variables have been measured with some
UNIT.
degree of error and requires that dimen-
principal-axis f a c t o r analysis in sions be extracted in a particular way.
EXPLORATORY FACTOR ANALYSIS, an ex- Specifically, the first dimension ex-
traction method in which the COEFFI- tracted must account for the maximum
CIENT OF MULTIPLE DETERMINATION of possible variance, having the highest
one variable with all other variables in squared correlation with the variables it
the system is used as the initial underlies; the second dimension must
COMMUNALITY estimate for that vari- account for the next maximal amount of
able. variance and be uncorrelated with the
previously extracted dimension; and so
p r i n c i p a l component a single linear forth. The researcher retains a certain
combination extracted as part of a PRIN- number of dimensions based on various
CIPAL COMPONENTS ANALYSIS. criteria, including interpretations of
FACTOR LOADINGS.
principal components analysis
(PCA) a data reduction approacfi in p r i n c i p l e o f beneficence in RE-
which a number of independent linear SEARCH ETHICS, the requirement of IN-
combinations of underlying explana- STITUTIONAL REVIEW BOARDS that
tory variables are identified for a larger studies "do good" with respect to the
set of original observed variables. PCA work being conducted, the benefits to
reproduces all of the information in the society at large, and the treatment of
original CORRELATION MATRIX and does participants. Thus, the researcher should
not assume that variables are measured maximize the possible benefits of each
with any degree of error. Thus, the result srady and consider its potential impact
is a new set of variables that are in the broadest sense. For example, in a
uncorrelated with each other and or- study of implicit attitudes a researcher
dered in terms of the percentage of the might note that understanding more
total variance for which they account. about the measurement of attitudes re-
The technique is similar in its aims to garding sensitive topics may lead to re-
FACTOR ANALYSIS but has different tech- duced societal prejudice.

271
principle of nonmaleficence

p r i n c i p l e o f nonmaleficence in RE- tation of a PROBABILITY DISTRIBUTION.

SEARCH ETHICS, the requirement of IN- Also called density curve.
STITUTIONAL REVIEW BOARDS that probability density function (PDF)
studies "do no harm" to participants. the mathematical representation of the
When a person considers taking part in a shape of a PROBABILITY DISTRIBUTION: It
study, there is an expectation that he or gives the likelihood that a CONTINUOUS
she will leave the study in a state that is VARIABLE of interest will assume any
no worse than when the srady began. specific value from the given range. Also
Where negative consequences are not called density function; probability
entirely avoidableas in an experiment function. See also PROBABILITY MASS
in which a participant is required to FUNCTION.
recall painful memories, for example
researchers have a duty to minimize the p r o b a b i l i t y d i s t r i b u t i o n a distribu-
impact of such consequences. tion describing the probability that a
random variable will take certain values.
p r i n c i p l e of parsimony see LAW OF The best known example is the bell-
PARSIMONY. shaped NORMAL DISTRIBUTION; Others
include the CHI-SQUARE DISTRIBUTION,
p r i o r d i s t r i b u t i o n a PROBABILITY DIS-
Student's T DISTRIBUTION, and the F DIS-
TRIBUTION of possible values for an un-
TRIBUTION. A probability distribution is
known population characteristic that is
a THEORETICAL DISTRIBUTION.
formulated before one obtains any cur-
rent data observations about the phe- probability generating function
nomenon of interest. It may be a (PGF) a formula giving the coefficients
subjective prior (based on a researcher's of a sequence of exponential terms de-
knowledge of the specific field) or an ob- scribing the likely values of a random
jective prior (based on evidence ob- variable. It is the summative sequence of
tained from other studies). In BAYESIAN the individual PROBABILITY MASS FUNC-
methods the prior distribution is com- TIONS for each possible value of that
bined with the LIKELIHOOD FUNCTION variable. See GENERATING FUNCTION. See
to yield the POSTERIOR DISTRIBUTION also CUMULANT GENERATING FUNCTION;
from which inferences are made. Also MOMENT GENERATING FUNCTION.
called prior; prior probability dis-
p r o b a b i l i t y judgment a decision
tribution. See also INFORMATIVE PRIOR.
made by an individual about the likeli-
p r i o r p r o b a b i l i t y in BAYESIAN meth- hood that an event will occur. Often
ods, the likelihood of a certain event oc- based on a personal feeling or estimate
curring as determined from current and thus highly subjective in nature,
accumulated knowledge about the phe- such decisions are a common area of
nomenon. For example, a researcher study, for example by researchers seek-
could calculate the prior probabihty of a ing to understand why some individuals
couple divorcing within five years of engage in risky health behaviors. See
marriage based on estimates obtained SUBJECTIVE PROBABILITY.
from census data and previously con- p r o b a b i l i t y level (p value) in statisti-
ducted studies. See also BAYES THEOREM; cal SIGNIFICANCE TESTING, the likeli-
POSTERIOR PROBABILITY. hood that the observed result would
p r o b a b i l i t y (symbol: p) n. the degree to have been obtained if the NULL HYPOTH-
which an event is likely to occur. See ESIS of no real effect were trae. Small p
values (conventionally, those less than
also CONDITIONAL PROBABILITY; PROBA-
.05 or .01) suggest that the chance of ex-
BILITY LEVEL. probabilistic adj.
perimental results mistakenly being at-
p r o b a b i l i t y curve a graphic represen- tributed to the independent variables

272
 probit transformation

present in the study (rather than to the p r o b a b i l i t y table 1. a chart showing

random factors actually responsible) is all possible outcomes of a siraation
small. Traditionally, the null hypothesis under specified conditions and their
is rejected if the value of p is no larger likelihoods of occurrence. For example,
than the SIGNIFICANCE LEVEL set for the one could create a probability table for
test. Also called probability value. the number of boys born into a family.
2. a display of CRITICAL VALUES for vari-
probability mass function (PMF) ous statistical tests (e.g., z TEST, T TEST, F
the mathematical expression that relates TEST, CHI-SQUARE TEST) at different SIG-
a given value to the probabUity that a NIFICANCE LEVELS and with different DE-
discrete RANDOM VARIABLE will take that GREES OF FREEDOM. Probability tables
value. It is thus the equivalent for a DIS- traditionally were presented as appen-
CRETE DISTRIBUTION ofthe PROBABILITY
dixes in the back of statistics textbooks
DENSITY FUNCTION for continuous vari-
but currently are widely available via
ables. computer software.
p r o b a b i l i t y m a t r i x see TRANSITION p r o b a b i l i t y theory the branch of
MATRIX. mathematics and statistics concerned
with the study of probabilistic phenom-
p r o b a b i l i t y plot a graphic that com-
ena.
pares probability values from two data
sets; typically, one set will be a series of p r o b a b i l i t y value see PROBABILITY
experimental findings and the other a LEVEL.
theoretical set. Good examples are the
probability-probability plot (P-P plot), probable error an estimated range of
depicting the CUMULATIVE PROBABILI- error for the value of a population char-
TIES of one data set against those of the acteristic as calculated from a sample of
other, and the QUANTILE-QUANTILE that population. The probable error of a
PLOT. mean, for example, is obtained by multi-
plying the STANDARD DEVIATION by
p r o b a b i l i t y ratio the likelihood that 0.6745.
a randomly selected unit will have a par-
ticular quality as derived from the p r o b i t n. probability unit: the inverse of
known or expected frequency of this the normal CUMULATIVE PROBABILITY
quality in the population. For example, DISTRIBUTION. It is used as the LINK
if a bag contains 300 red balls and 500 FUNCTION in PROBIT ANALYSIS and other
white balls, then the probabihty ratio of techniques for modeUng values of bi-
drawing a red ball is 300 out of 800 = 3:8. nary variables.
probit analysis a type of REGRESSION
p r o b a b i l i t y s a m p l i n g any process in ANALYSIS in which a dichotomous out-
which a sample of participants or cases come variable is related to any of a num-
are chosen from a larger group in such a ber of different predictor or independent
way that each item has a known (or cal- variables. The technique assumes that a
culable) likelihood of being included. latent process generated the observed bi-
This requires a well-defined POPULA- naty data and uses a LINK FUNCTION to
TION and an objective selection proce- generate a mathematical representation
dure, as in RANDOM SAMPLING. of relationships. Probit analysis and LO-
AdditionaUy, all members of the popula- GISTIC REGRESSION often produce vety
tion must have some (i.e., nonzero) similar results. Also called probit re-
chance of being selected, although this gression.
probability need not be the same for all
individuals. Compare NONPROBABILITY probit t r a n s f o r m a t i o n a method by
SAMPLING. which a nonlinear arrangement of val-

273
procedure

ues, such as those forming an S-shaped such as a need to clarity instructions.

curve, is converted into probability units Compare SELF-ADMINISTERED TEST.
(see PROBIT) through a formula that in-
volves addition, subtraction, multiplica- product-limit estimator see KAPLAN-
tion, or division with a constant. It is MEIER ESTIMATOR.
thus a type of LINEAR TRANSFORMATION.
product-moment correlation co-
procedure n. a specific sequence of e f f i c i e n t (symbol: r) an index of the
steps or actions delineating the manner degree of linear relationship between
in which a study is to be conducted or two variables. Devised by British statisti-
has been conducted. Procedural descrip- cian Kari Pearson (1857-1936), it is often
tions enable other researchers to under- known as the Pearson product-moment
stand the conditions of a study and correlation coefficient (Pearson's r) and
replicate it if desired. procedural adj. is one of the most commonly used SAM-
PLE CORRELATION COEFFICIENTS.
process research the study of the vari-
ous mechanisms that influence the out- product vector the result of multiply-
come of a treatment or intervention. A ing two matrices each of which com-
basic goal of such research is to identify prises a single column or row. Product
those therapeutic methods and pro- vectors are relevant to the calculation of
cesses that are most effective in bringing COVARIANCES and other statistics. Also
about positive change, as well as any in- called vector product. See also CROSS-
adequacies or other limitations. In other PRODUCT.
words, process research focuses on the
means by which participants benefit p r o f i l e n. see PERSONALITY PROFILE;
from receiving an intervention. Com- TEST PROFILE.
pare OUTCOME RESEARCH. p r o f i l e analysis a multivariate statisti-
Procrustes analysis an approach in cal technique that compares indepen-
which empirical data are superimposed dent groups of individuals across several
onto some fixed target matrix or other constructs or dimensions (e.g., of per-
stracrare of theoretical interest. The sonality) that are measured on the same
name derives from the robber in Greek scale in terms of their mean level or ele-
mythology who forced his victims to fit vation, their shape, and their scatter or
his bed by stretching them or cutting off variability. The data can be depicted
their Umbs. graphically with the scale name on the
>:-axis, the scale score on the y-axis, and
Procrustes r o t a t i o n a LINEAR TRANS- different lines reflecting the different
FORMATION of the points represented in groups being examined.
a MATRIX to best conform, in a least
squares sense (see LEAST SQUARES CRITE- p r o f i l e s i m i l a r i t y c o e f f i c i e n t any
RION), to the points in another target of several indices used in multivariate
matrix. Usually the points in the target analyses to describe the similarity in ele-
matrix represent some theoretical factor vation, shape, and scatter of sets of
structure or the results of a FACTOR scores from different groups of cases.
ANALYSIS on a different population. It Some measures, such as the PRODUCT-
is performed as part of a PROCRUSTES MOMENT CORRELATION COEFFICIENT,
describe the consistency of scores across
ANALYSIS. See also PROMAX ROTATION.
groups but not the mean level or eleva-
proctored test a test that is adminis- tion of the scores. Other indices, such as
tered with an observer present to moni- the INTRACLASS CORRELATION coeffi-
tor the security of the administration cient, describe whole score sets well but
and address any issues that might arise. are not able to distinguish extreme or av-
274
 propensity score

erage individual scores. Also called pro- well as sentence completion, word asso-
file proximity measure. ciation, and drawing tests. The use of
projective techniques has generated
prognostic variable in medical re-
considerable discussion among re-
search, a BASELINE CHARACTERISTIC Of
searchers, with opinions ranging from
an individual that may lead to a clinical
the belief that personality assessment is
diagnosis in the future. It is important
incomplete without data from at least
for researchers to define such character-
one or more of these procedures to the
istics early in a study to help customize
view that such techniques lack impor-
interventions to be as effective as possi-
tant psychometric features such as RELI-
ble for each person. For example, child-
ABILITY and VALIDITY. Also called
hood obesity is a prognostic variable for
projective method.
adult-onset diabetes, such that many in-
terventions focus on promoting healthy promax r o t a t i o n in EXPLORATORY
eating and exercise in youth. FACTOR ANALYSIS, atypeOf PROCRUSTES
ROTATION used to fit a solution to a sim-
p r o g r a m evaluation an appraisal pro-
pler structure or target. More specifi-
cess that contributes to decisions on
cally, it is a form of OBLIQUE ROTATION
installing, continuing, expanding, certi-
in which a researcher starts with a solu-
fying, or modifying social programs, de-
tion from a VARIMAX ROTATION and
pending on their effectiveness. Program
raises the FACTOR LOADINGS to a certain
evaluation also is used to obtain evi-
power to make them more extreme; he
dence to rally support or opposition for
or she then finds the rotation to the tar-
the organization providing services and
get using the LEAST SQUARES CRITERION.
to contribute to basic knowledge in the
A researcher subsequently interprets the
social and behavioral sciences about so-
loadings of the variables on the factors
cial interventions and social experimen-
and also a correlation matrix of the fac-
tation. See EVALUATION RESEARCH.
tors.
projection n. in linear algebra and
proof n. 1. the establishment of a propo-
EXPLORATORY FACTOR ANALYSIS, the
sition or theory as true, or the method
mapping of a set of points in multi-
by which it is so established. There is
dimensional space. project vb.
much debate as to whether propositions
projection pursuit in EXPLORATORY or theories can ever be truly proven. In
DATA ANALYSIS, a method for identify- logic and philosophy, even a valid argu-
ing the most interesting clusters in LIN- ment can be untrue if its first premise is
EAR COMBINATIONS of variables in a false. For example, it is a valid argument
multivariate data set. If a nonnormal to say that All trees are pines: I have a tree
cluster is identified, its corresponding in my garden: Therefore my tree is a pine. In
data are removed and a subsequent clus- empirical sciences such as psychology,
ter is obtained. The process continues both logical and methodological prob-
until no further clusters can be found. lems make it impossible to prove a the-
ory or hypothesis true. Disciplines that
projective technique any assessment
rely on empirical science must settle for
procedure that consists of a fixed series
some type of probabilism based on em-
of relatively ambiguous stimuli designed
pirical support of its theories and hy-
to elicit unique, sometimes highly idio-
potheses. See also FALSIFIABILITY. 2. in
syncratic, responses that reflect the per-
mathematics and logic, a sequence of
sonality, cognitive style, and other
steps formally establishing the truth of a
psychological characteristics of the re-
theorem or the validity of a proposition.
spondent. Examples of this type of pro-
cedure are the Rorschach Inkblot Test propensity score in an experimental
and the Thematic Apperception Test, as design lacking RANDOM ASSIGNMENT, a

275
prophecy formula

measure of the probability that a partici- searcher calculates an ODDS RATIO for
pant will be part of the treatment group. each category, which indicates the likeli-
It is a type of BALANCING SCORE used to hood of that specific outcome occurring
adjust for SAMPLING BIAS by matching according to the values of the predictors
differences among participants across and COVARIATES of interest. For exam-
conditions. Several different algorithms ple, one might use the proportional
exist for determining propensity scores. odds model to determine employees'
likely degree of job satisfaction (poor,
prophecy f o r m u l a see SPEARMAN-
fair, or good) according to such variables
B R O W N PROPHECY FORMULA.
as salary, work hours, length of com-
mute, work responsibilities, decision-
p r o p o r t i o n n. a decimal value that ex-
presses the size of a subset of cases of in-making authority, age, and sex.
terest relative to the set as a whole. For proportional reduction of error
example, if 10 students out of 40 stu- (PRE) an index of the extent to which
dents in a class complete their assign- the addition of one or more INDEPEN-
ments on time then the proportion of DENT VARIABLES In a REGRESSION EQUA-
students who are not late is 10/40 = .25. TION reduces inaccuracy in predicting
p r o p o r t i o n a l adj. having a constant scores on the DEPENDENT VARIABLE. It is
ratio between quantities, such that the based on a LOSS FUNCTION and generally
overall relationship does not change. ranges from 0 to 1 in value. The concept
For example, consider a researcher who may be appUed to GENERALIZED LINEAR
is examining the differences between MODELS and other statistical approaches
two treatments among males and fe- as well. Also called proportional re-
males as follows: 10 males in Treatment duction of prediction error.
A and 20 males in Treatment B, and 20 proportional sampling an approach
females in Treatment A and 40 females in which one draws cases for srady from
in Treatment B. Although there are un- certain groups (e.g., gender, race/ethnic-
equal numbers of people in each treat- ity) in the amounts that are observed in
ment condition, the ratio or proportion the larger population. For example, if a
between them remains the sametwice university has 60% female students and
as many females as males. In UNBAL- 40% male students, a researcher would
ANCED DESIGNS, the presence or absence obtain a sample comprising the same
of such proportional cell frequency is percentages or proportions, such as 120
critical to a researcher's choice of ana- females and 80 males in a 200-student
lytic strategy. subset. A significant drawback of this
proportional hazards assumption form of STRATIFIED SAMPLING is that
in cox REGRESSION ANALYSIS, the situa- small sample sizes may limit statistical
tion in which two individuals with simi- modeling and inference. Also called
lar COVARIATES have the same ratio of proportionate sampling.
estimated hazards over time. A variety of proportion of variance accounted
tests of proportionality exist for re- f o r the extent to which certain factors
searchers to evaluate whether this as- or INDEPENDENT VARIABLES in a pro-
sumption holds. posed model are associated with the out-
come or DEPENDENT VARIABLE of
p r o p o r t i o n a l hazards model see
interest. The total amount of possible
cox REGRESSION ANALYSIS. variance in an outcome is 1.00 or 100%,
p r o p o r t i o n a l odds model a type of hence the proportion of variance ac-
LOGISTIC REGRESSION used when the counted for indicates what fraction of
outcome variable has more than two or- that 100% is explained by the model.
dered response categories. The re- For example, in a REGRESSION EQUATION
 pseudorandom

predicting propensity to help someone protocol n. see RESEARCH PROTOCOL;

in distress, an individual's level of agree- VERBAL PROTOCOL.
ableness and confidence may explain
protocol v i o l a t i o n any intentional
.14, or 14%, of the variance in the pro-
deviation by a researcher or srady per-
pensity to help measure.
sonnel from the research procedures as
p r o p o r t i o n o f variance index any approved by an INSTITUTIONAL REVIEW
of several measures describing the BOARD. For example, an investigator
amount of variability in an outcome who begins participant recmitment be-
variable that is explained by the inde- fore the board has authorized the srady
pendent variables. In MULTIPLE REGRES- has committed a protocol violation.
SION, for example, the COEFFICIENT OF prototheory n. a working HYPOTHESIS
MULTIPLE DETERMINATION is a com- or starting set of assumptions about a
monly used proportion of variance phenomenon of interest. If empirical
index. The larger the value of this index, data from independent studies do not
the greater the amount of variance that support the prototheoty, it must be re-
is accounted for by the independent vised or rejected. See THEORY.
variables, with 1.00 (100%) being the
theoretical upper bound. proximate cause the most direct or
immediate cause of an event. For exam-
prosecutor's f a l l a c y see DEFENDER'S ple, the proximate cause of Smith's ag-
FALLACY. gression may be an insult, but the
ULTIMATE CAUSE may be Smith's early
prospective research research that chUdhood experiences.
starts with the present and follows par-
ticipants forward in time to examine p r o x i m i t y m a t r i x a general term de-
trends, predictions, and outcomes. Ex- noting either a SIMILARITY MATRIX or a
amples include randomized experi- DISSIMILARITY MATRIX.
ments and LONGITUDINAL DESIGNS. Also proxy variable a variable, b, used in
called forward-looking study; pro- place of another, a, when b and a are
spective study. Compare RETROSPEC- substantially correlated but scores are
TIVE RESEARCH. available only on variable b, often be-
cause of the difficulty or costs involved
prospective s a m p l i n g a sampling
in collecting data for variable a. In other
method in which cases are selected for
words, is a substituted measure of a.
inclusion in experiments or other re-
For example, rather than reporting the
search on the basis of their exposure to a
precise income levels of their parents or
risk factor. Participants are then fol-
guardians, survey respondents instead
lowed to see if the condition of interest
may be asked to indicate the highest
develops. For example, young children
level of education of each parent or
who were exposed to lead in their drink-
guardian; in this case, education is a
ing water and those who were not ex-
proxy for income. Also called surro-
posed to this risk factor could be
gate variable.
included in a study and then foUowed
through time to assess health problems pseudoinverse n. see GENERALIZED IN-
that emerge when they are adolescents. VERSE.
Compare RETROSPECTIVE SAMPLING.
pseudorandom adj. describing a pro-
prospective v a l i d i t y see PREDICTIVE cess that is close to being RANDOM but is
VALIDITY. not perfectly so. For example, so-called
RANDOM NUMBER GENERATORS are often
protected t test see FISHER LEAST SIG- described as pseudorandom because
NIFICANT DIFFERENCE TEST. their algorithms do not achieve the ideal

277
pseudoscience

of producing aflawlesslyrandom pro- C L I N I C A L SIGNIFICANCE; P R A C T I C A L SIG-

cess. Also called quasi-random. N I F I C A N C E ; STATISTICAL SIGNIFICANCE.

pseudoscience n. a system of theories psychological statistics the area

and methods that has some resemblance within psychology and the behavioral
to a genuine science but that cannot sciences that is concerned with research
be considered such. Examples range design and methodology, addressing is-
from astrology, numerology, and eso- sues of measurement, SAMPLING, data
teric magic to such modern phenomena collection, data analysis, and reporting
as Scientology. Various criteria for dis- of findings. See also QUANTITATIVE PSY-
tinguishing pseudosciences from true CHOLOGY.
sciences have been proposed, one of
psychological test any standardized
the most influential being that of
instrument, including scales and self-
FALSIFIABILITY. On this basis, certain ap-
report inventories, used to measure
proaches to psychology and psychoanal-
behavior or mental attributes, such as at-
ysis have sometimes been criticized as
titudes, emotional functioning, intelli-
pseudoscientific, as they involve theo-
gence and cognitive abilities (reasoning,
ries or other constructs that cannot be
comprehension, abstraction, etc.), apti-
directly or definitively tested by observa-
rades, values, interests, and personality
tion. pseudoscientific adj.
characteristics. For example, a researcher
pseudovalue n. an estimated PARAME- might use a psychological test of emo-
TER value used as a temporary place- tional intelligence to examine whether
holder in analytic approaches that some managers make better decisions in
require intensive iterative calculations, conflict situations than others. Also
such as BOOTSTRAPPING, the JACKKNIFE, called psychometric test.
and other RESAMPLING techniques. psychological testing see PSYCHO-
METRICS.
psychogram n. see TRAIT PROFILE.
psychometrician n. 1. an individual
psychological scale 1. a system of
with a theoretical knowledge of mea-
measurement for a cognitive, social,
surement techniques who is qualified to
emotional, or behavioral variable or
develop, evaluate, and improve psycho-
function, such as personality, intelli-
logical tests. 2. an individual who is
gence, attitudes, or beliefs. 2. any instru-
trained to administer psychological tests
ment that can be used to make such
and interpret their results, under the su-
a measurement. A psychological scale
pervision of a licensed psychologist.
may comprise a single item measured
Also called psychometrist.
using a variety of response formats (e.g.,
SEMANTIC DIFFERENTIAL, multiple choice, psychometric model any theoretical
checklist) or be a collection of items with and statistical framework describing
similar formats. Also called psycho- how respondents generate their answers
metric scale. See also PSYCHOLOGICAL to items on a scale or instrument and ex-
TEST. plaining associated sources of error,
which in turn provides a means to deter-
psychological significance the ex- mine the reliability of test scores across
tent to which an effect found in a re- different testing siraations. There are
search study is relevant to the attitudes, three major approaches: CLASSICAL TEST
cognitions, beliefs, and behavior of hu- THEORY, GENERALIZABILITY THEORY,
mans. For example, a finding that de- and ITEM RESPONSE THEORY.
scribes how individuals function in a
variety of different settings would have psychometric research studies in the
psychological significance. See also field of psychological measurement. Such

278
 P-technique factor analysis

research includes the development of ing physical stimulus properties to per-

new measures and appropriate methods ceived magnitude. For example, a re-
for their scoring, the establishment of spondent in a srady may have to
RELIABILITY and VALIDITY evidence for indicate the roughness of several differ-
measures, the examination of item and ent materials that vary in texture.
scale properties and their dimensions, Methods are often classified as direct or
and the evaluation of DIFFERENTIAL indirect, based on whether the observer
ITEM FUNCTIONING across subgroups. directly judges magnitude.
For example, psychometric research psychophysiological research em-
could be used to determine whether a pirical sradies, often conducted in a lab-
new scale is appropriately administered oratoty setting using functional
and scored in a specific subpopulation of magnetic resonance imaging and other
respondents. neuroscientific tools, that link an indi-
psychometrics n. the branch of psy- vidual's bodily responses (e.g., change in
chology concerned with the quantifica- heart rate, palmar sweat, eye blink) and
tion and measurement of human mental processes (e.g., memoty, cogni-
attributes, behavior, performance, and tive processing, brain function). For ex-
the like, as well as with the design, anal- ample, in addition to coUecting data
ysis, and improvement of the tests, ques- on several performance measures, a
tionnaires, and so on used in such psychophysiological researcher might
measurement. Also called psycho- examine Cortisol levels in adolescents
metric psychology; psychometry. with a diagnosis of conduct disorder and
compare them to levels in those without
psychometric scale see PSYCHOLOGI- the disorder.
CAL SCALE.
psychosocial effect any effect upon
psychometric scaling the creation of an individual's attirades and interac-
an instrament to measure a psychologi- tions that involves the influence of
cal concept through a process of analyz- a particular social environment. For ex-
ing responses to a set of test items or ample, a srady might show a psycho-
other stimuli. It involves identifying social effect in which lack of social
item properties, noting whether re- support lowers one's likelihood of seek-
sponses match theoretical formats, re- ing mental health treatment.
ducing the larger set of items into a
smaller number (e.g., through EXPLOR- P-technique f a c t o r analysis a
ATORY FACTOR ANALYSIS), and deter- method in EXPLORATORY FACTOR ANAL-
mining appropriate scoring methods. YSIS for understanding the major under-
THURSTONE SCALING Is an example. lying dimensions of variables for a given
person over time. In this approach, a sin-
psychometric test see PSYCHOLOGI- gle individual or unit is measured re-
CAL TEST. peatedly on a broad range of variables,
the variables are correlated over the se-
psychophysical research empirical
ries of occasions sampled, and the corre-
studies, often conducted in a laboratory
lation factor is analyzed. For example, a
setting, linking properties of a physical
set of memoty tests may be administered
stimulus to a sensoty response. For ex-
to a person at several points in his or her
ample, in a study of hearing, a partici- life; P-technique factor analysis could
pant might be requested to distinguish a then be used to identity dimensions of
number of different sounds by their change or patterns of covariation among
loudness. the variables across occasions. Compare
psychophysical scaling any of the Q-TECHNIQUE FACTOR ANALYSIS; R-
techniques used to constract scales relat- TECHNIQUE FACTOR ANALYSIS.

279
publication bias

p u b l i c a t i o n bias the tendency for pure research see BASIC RESEARCH.

study results that are published in jour- pure science see BASIC SCIENCE.
nals or other outlets to differ from study
results that are not published. In particu- purposive s a m p l i n g a SAMPLING
lar, published studies are more likely method that focuses on very specific
to show positive or statistically signifi- characteristics of the units or individuals
cant findings. Thus, when conducting a chosen. For example, a researcher inves-
META-ANALYSIS it Is important to gather tigating a specific type of amnesia may
the full range of available research, both select for study only those individuals
published and unpublished, to ensure who have specific lesions in their brains.
the analysis does not provide unrep- Although the final subset of cases is ex-
resentatively large EFFECT SIZES. See FILE- treme and not random, valuable infor-
DRAWER PROBLEM. mation nonetheless may be obtained
from their study. See NONPROBABILITY
p u b l i c a t i o n ethics the principles and SAMPLING.
standards associated with the process of
publishing the results of scientific re- p value n. see PROBABILITY LEVEL.
search or scholarly work in general. P y g m a l i o n effect a consequence or re-
These include such matters as giving the action in which the expectations of a
appropriate credit and authorship status leader or superior engender behavior
to those who have earned it; ensuring from followers or subordinates that is
that appropriate citations are given to consistent with these expectations: a
ideas, methodology, or findings from form of SELF-FULFILLING PROPHECY. For
another srady; not submitting the same example, raising a manager's expecta-
article to more than one journal simulta- tions regarding the performance of sub-
neously; and not submitting for ordinate employees has often been
republication results that have already found to enhance the performance of
been published elsewhere without indi- those employees. See UPWARD PYG-
cating that fact. MALION EFFECT.
 Qq
q 1. symbol for STUDENTIZED RANGE STA- the major dimensions or "types" of
TISTIC 2. symbol for the probability of people by identitying how they perceive
failure in a BINARY TRIAL: q=l-p. different variables. Instead of describing
how variables group together, as in R-
Q 1 . see COCHRAN Q TEST. 2. see YULE'S Q.
TECHNIQUE FACTOR ANALYSIS, thiS ap-
3. symbol for a specific QUARTILE (e.g., proach examines how people group to-
Q J , first quartile; Q3, third quartile). gether, with a correlation of people
QDA abbreviation for QUADRATIC DIS- across a broad range of variables. For ex-
CRIMINANT ANALYSIS. ample, types of students could be identi-
fied based on how participants describe
QI abbreviation for QUASI-INDEPENDENCE. themselves academically (e.g., "hard
Q methodology a set of research meth- working," "procrastinating"); individual
ods used to study how people think or descriptors of academic behavior (e.g.,
feel about a topic. It includes, among "Plans projects in advance of due dates")
other methods, Q-TECHNIQUE FACTOR would have FACTOR LOADINGS on each
ANALYSIS, which attempts to identity of the underlying person factors. Also
shared ways of thinking; and Q SORTS, in called inverse factor analysis; Q-
which participants are given statements technique factoring. Compare P-
about people and topics and asked to TECHNIQUE FACTOR ANALYSIS.
sort them into categories. Compare R
METHODOLOGY.
quadrant n. one of four divisions of a
two-dimensional x-y plot of psychologi-
Q-Q plot abbreviation for QUANTILE- cal variables, as shown in the generic
QUANTILE PLOT. depiction overleaf. The quadrants corre-
spond to (A) positive scores on both di-
Q sort a data-collection procedure, often mensions (upper right corner of plot),
used in personality measurement, in (B) negative scores on both dimensions
which a participant or independent (lower left corner of plot), and (C and D)
rater sorts a broad set of stimuli into cat- positive scores on one dimension and
egories using a specific instruction set. negative scores on the other dimension
The stimuli are often short descriptive (upper left corner and lower right cor-
statements (e.g., of personal traits) printed ner, respectively, of the plot).
on cards. Examples ofthe instruction set
are "describe yourself"; "describe this quadrant s a m p l i n g a method for se-
child"; and "describe your friend." In lecting units of analysis (e.g., partici-
the classic or structured Q sort, raters are pants, organizations) from different
consttained to use a predetermined num- areas of a space. The space is divided into
ber of stimuli in each categoty. Often, four sections and units are drawn from
general categories are used to reflect the each. In psychology, units may be
NORMAL DISTRIBUTION with raters as- drawn from psychological space, such as
signing fewer stimuli to the extreme cat- personality variables that are thought of
egories and many more stimuli to the as arrayed in two-dimensional space. See
middle categories. See Q METHODOLOGY. QUADRANT.
Q-technique f a c t o r analysis a type quadratic discriminant analysis
of FACTOR ANALYSIS used to understand (QDA) a method used to classify vari-

281
quadratic discriminant function

Quadrant C Quadrant A
(-,+) 2 {+.+)

-4 -3 -2 -1 1 2 3
H 1 1 1 -H 1 h

Quadrant B Quadrant D
(-, -) (+. -)

quadrant

ables into two or more groups. A set of used process in statistical procedures in-
PREDICTOR VARIABLES Is Combined opti- volving the LEAST SQUARES CRITERION
mally to maximize the association with or other principles for minimizing error;
a particular grouping outcome (e.g., dif- it sequentially determines the difference
ferent clinical diagnostic groups) with- between a provisional estimate and a
out the strong assumptions required of final estimate until some stopping point.
other, more simple linear versions of the See LOSS F U N C T I O N .
model (see DISCRIMINANT ANALYSIS).
These assumptions include MULTI- quadrature n. the numerical computa-
VARIATE NORMALITY and HOMOGENEITY tion of an integral of a function, either
OF V A R I A N C E .
by traditional rules or by an adaptive
process involving approximation. It is
quadratic discriminant function used extensively during estimation in
the set of optimal PREDICTOR VARIABLE modeling approaches, such as ITEM RE-
weights that emerge from QUADRATIC SPONSE THEORY and GENERALIZED LIN-
DISCRIMINANT ANALYSIS. T h e weights
EAR MODELS.
help the researcher understand and in-
terpret which of the predictor variables qualitative adj. referring to a variable,
in the analysis are particularly strong in study, or analysis that involves a method
accounting for membership in a particu- of inquity based on descriptive data
lar outcome group and which variables without the use of numbers. Qualitative
have little or no influence in this predic- methods and approaches focus on un-
tion. derstanding open-ended responses, such
as those found in written narratives,
quadratic form 1. a specific mathe-
interviews, focus groups, observation,
matical form that is central to MULTI-
and case studies (see QUALITATIVE RE-
VARIATE ANALYSIS. If X is a vector and A
SEARCH). Compare QUANTITATIVE.
a square matrix, then the quadratic form
is given by x'Ax. 2. a POLYNOMIAL of de-
qualitative analysis the investigation
gree two; for example, 4x^ + 2;i;y - 3y^ is a
of open-ended material and narratives
quadratic form of the variables x and y.
by researchers or raters who describe
quadratic loss f u n c t i o n a commonly dominant themes that emerge in the

282
 quality control angle chart

data. In many cases specialized com- years to death, time to relapse) but also
puter programs are used to identity these recognizes aspects of the treatment that
themes with researcher-provided search affect a patient's quality of life. For ex-
terms. A major component of describing ample, the analysis would identify a
the data is ttying to understand the rea- treatment condition (e.g., introduction
sons behind the observed themes. Com- of a new drag) that might lead to longer
pare QUANTITATIVE ANALYSIS. but much lower quality of life for the pa-
tient.
qualitative data information that is
not expressed numerically, such as de- q u a l i t y adjusted s u r v i v a l time in
scriptions of behavior, thoughts, atti- QUALITY ADJUSTED SURVIVAL ANALYSIS,
tudes, and experiences. If desired, a patient's days of survival in good
qualitative data can often be expressed health, that is, discounting days with
quantitatively through a CODING pro- side effects, sickness, or low productivity
cess. See QUANTITATIVE RESEARCH. Com- due to the treatment.
pare QUANTITATIVE DATA.
q u a l i t y assurance in health adminis-
qualitative observation a formal de- tration or other areas of service delivety,
scription of a phenomenon that takes a systematic process that is used to mon-
into account the context in which that itor and provide continuous improve-
phenomenon occurs but does not rely ment in the quality of health care
on numbers in the description. See services. It involves not only evaluating
QUALITATIVE RESEARCH. the services in terms of effectiveness, ap-
qualitative research a method of re- propriateness, and acceptability but also
search that produces descriptive (non- offering feedback and implementing so-
numerical) data, such as observations of lutions to correct any identified defi-
behavior or personal accounts of experi- ciencies and assessing the results.
ences. The goal of gathering this QUALI- quality control processes associated
TATIVE DATA is to examine how in-
with research, production, or services
dividuals can perceive the world from that are designed to reduce the number
different vantage points. A variety of of defective measurements and products.
techniques are subsumed under qualita-
tive research, including CONTENT ANAL- q u a l i t y c o n t r o l angle chart a picto-
YSES of narratives, in-depth INTERVIEWS, rial representation of a TIME SERIES in
FOCUS GROUPS, PARTICIPANT OBSERVA- which a change in the mean or variabil-
TION, and CASE STUDIES, often con- ity is indicated by a change in angular
ducted in naturalistic settings. Also direction. It shows the behavior of a sta-
called qualitative design; qualita- tistic (on the y-axis) across many differ-
tive inquiry; qualitative method; ent samples over time (on the ;t-axis). A
qualitative study. Compare QUANTI- researcher can see whether the statistic
TATIVE RESEARCH. settles to a particular point over time
and where extreme values of that statis-
qualitative variable a descriptive
tic are over samples. The figure usually
characteristic or attribute, that is, one includes the average of the statistic over
that cannot be numerically ordered. Ex- the samples and upper (UCL) and lower
amples are gender, eye color, and pre- (LCL) control limits shown as lines that
ferred sport. Compare QUANTITATIVE are two or three standard deviations
VARIABLE. from the mean.
quality adjusted survival analysis Consider the hypothetical example
in controlled clinical trials, a type of overleaf, which depicts mood change for
SURVIVAL ANALYSIS that not only pre- a group of 25 individuals undergoing 12
dicts time to an event (e.g., number of weeks of psychotherapy for depression.

283
quantification

UCL

baseline

LCL

10 15 20

Participant
quality control angle chart

q u a n t i f i c a t i o n n. the process of ex- summaries of MEANS and STANDARD DE-

pressing a concept or variable in numeri- VIATIONS of variablesand INFERENTIAL
cal form, which may aid in analysis and STATISTICSsuch as ANALYSIS OF VARI-
understanding. ANCE, REGRESSION ANALYSIS, and HIER-
ARCHICAL LINEAR MODELS. Compare
quantile n. a value in a series of values
QUALITATIVE ANALYSIS.
in ascending order below which a given
percentage of values lies; for example, a quantitative data information ex-
50% quantile (also called a MEDIAN) is pressed numerically, such as test scores
the point at which 50% of the values fall or measurements of length or width.
below that value (and 50% above). These data may or may not have a real
Othertypesof quantile are QUARTILES, zero but they have order and often equal
dividing the series into four equal-sized intervals. Compare QUALITATIVE DATA.
groups; DECILES, dividing it into 10
groups; and PERCENTILES (or centiles), quantitative psychology the srady
dividing it into 100 groups. Also called of methods and techniques for the mea-
fractile. surement of human attributes, the sta-
tistical and mathematical modeling of
quantile-quantile plot (Q-Q plot) a psychological processes, the design of
plot of the QUANTILES of One set of data research studies, and the analysis of psy-
against the quantiles of a second set, chological data. Researchers in this area
used to determine if the two data sets develop new methodologies and evalu-
follow a common distribution. A 45 ate existing methodologies under partic-
Une on the plot shows where a perfect ular conditions (e.g., with small samples).
match of distributions would be.
quantitative research a method of
quantitative adj. involving the use of a research that relies on measuring vari-
numerical measurement system to ana- ables using a numerical system, analyz-
lyze data. Compare QUALITATIVE. ing these measurements using any of
quantitative analysis the investiga- a variety of statistical models, and re-
tion of data empirically using numerical porting relationships and associations
variables. Quantitative analysis includes among the sradied variables. For exam-
both DESCRIPTIVE STATISTICSsuch as ple, these variables may be test scores or
 quasi-experimental design

measurements of reaction time. The goal quartiles: (Q3 + Qi - 2Q2)/(Q3 - Q,). The
of gathering this QUANTITATIVE DATA is value is negative (or positive) if few val-
to understand, describe, and predict the ues are at the negative (or positive) side
nature of a phenomenon, particularly of the distribution and is zero when the
through the development of models and data distribution is symmetric.
theories. Quantitative research tech-
niques include experiments and surveys. quartile deviation a measure of DIS-
Also called quantitative design; PERSION that is defined as the value half-
quantitative inquiry; quantitative way between the first and third
method; quantitative study. Com- QUARTILES (i.e., half the INTERQUARTILE
RANGE). Also called semi-interquartile
pare QUALITATIVE RESEARCH.
range.
quantitative variable a characteristic
or attribute that can be measured nu- q u a r t i m a x r o t a t i o n in FACTOR ANAL-
YSIS, an ORTHOGONAL ROTATION that
merically using a score obtained from
any of a variety of data sources. Exam- maximizes the variance across the rows
ples are age, height, and weight. Com- of the factor matrix by raising the load-
pare QUALITATIVE VARIABLE. ings to the fourth power; the effect is to
make large loadings especially large and
quartile n. one of the three values in a small loadings especially small. The ob-
series of values that divide it into equal- jective is to increase the interpretability
sized fourths. For example, the first (or of a factor solution by satisfying the SIM-
lower) quartile of a distribution is the PLE STRUCTURE ideal.
data value below which are the lowest
25% of scores, the second quartile is the quasi-experimental control group
data value below which are 50% of in a QUASI-EXPERIMENTAL DESIGN, any
scores, and the third (or upper) quartile group of participants who are assigned
is the data value below which are 75% of to a condition in which the treatment is
scores (or, conversely, above which are not introduced. These individuals will
25% of scores). These values provide in- receive the standard of care, receive a
formation to researchers about the rela- placebo, or participate in an activity that
tive spread of the distribution. See Q; involves them in the srady but does not
include the treatment in any way.
QUANTILE.
quartile coefficient of dispersion quasi-experimental design an ex-
a descriptive index that allows research- perimental design in which assignment
ers to compare the spread of two data of participants to an EXPERIMENTAL
distributions. It is computed by finding GROUP or to a CONTROL GROUP cannot
the difference between the first and be made at random for either practical or
third QUARTILES for each of the distribu- ethical reasons; this is usually the case in
tions and then comparing the values in FIELD RESEARCH. Assignment of partici-
a ratio: (Q, - Q3)/(Q] - Q3). For exam- pants to conditions is usually based on
ple, if the quartile coefficient of disper- self-selection (e.g., employees who have
sion is 10 for the first set and 4 for the chosen to work at a particular plant) or
second set, then the quartUe coefficient selection by an administrator (e.g., chil-
of dispersion is 2.5 times as great for the dren are assigned to particular class-
first set as for the second set (10/4 = 2.5). rooms by a superintendent of schools).
Such designs introduce a set of assump-
quartile coefficient o f skewness a tions or threats to INTERNAL VALIDITY
descriptive index of SKEWNESS in a data that must be acknowledged by the re-
set that is a ratio between a function of searcher when interpreting srady find-
the first, second, and third quartiles over ings. A study using this design is called
the difference between thefirstand third a quasi-experiment. Examples include

285
 quasi-experimental research

sradies that investigate the responses of to compression or stretching between

large groups to natural disasters or wide- scale points. See INTERVAL SCALE.
spread changes in social policy.
quasi-likelihood function a func-
quasi-experimental research re- tion used to obtain estimates for count
search in which the investigator cannot or binaty data that show more DISPER-
randomly assign units or participants to SION than the statistical model can han-
conditions, cannot generally control or dle. An advantage of quasi-likeUhood
manipulate the INDEPENDENT VARIABLE, models is that they permit an increased
and cannot limit the influence of extrane- flexibility in the data types and researc
ous variables, FIELD RESEARCH typically situations to which they may be applied.
takes the form of quasi-experimental re- A Umitation is that the function does
search. Also called nonexperimental not derive from a known population dis-
research. See QUASI-EXPERIMENTAL DE- tribution. Also called quasi-score esti-
SIGN. mating function.
quasi-F ratio in ANALYSIS OF VARIANCE
quasi-observation n. 1. the process of
and REGRESSION ANALYSIS, a substitute
collecting data about a person from a
for the F RATIO that can sometimes be
close source, rather than directly from
obtained when the denominators for an
the subject. An example is asking an in-
exact F ratio cannot be completed. The
dividual to report on the job satisfaction
quasi-f ratio is obtained by determining
of his or her partner. 2. the use of me-
the MEAN SQUARED ERROR using Certain
chanical means, such as video sur-
estimated components of variation when
veillance or audiotaping, to record
these cannot be calculated directly.
behaviors as a substitute for real-time
quasi-independence (Ql) n. in a CON- observation and questioning by a re-
TINGENCY TABLE, the situation in which searcher. In marketing research, an ex-
only a subset of entries or frequencies ample would be the use of surveillance
are independent or uninfluenced by one cameras to monitor shopper behavior in
another. Entries may not be independ- stores. This would cost less than paying a
ent for a variety of reasons: They may be trained researcher to observe and inter-
invalid, missing, or not counted in the view shoppers in sira. There is also the
n analysis. advantage that such data can be viewed,
stored, and analyzed at the researcher's
quasi-independent variable in ex-
convenience.
perimental design, personal attributes,
traits, or behaviors that are inseparable quasi-random adj. see PSEUDORAN-
from an individual and cannot reason- DOM.
ably be manipulated. These include gen-
der, age, and ethnicity. Such attributes quasi-random s a m p l i n g see SYSTEM-
may be modeled and treated as statisti- ATIC SAMPLING.
cally independent but are not subject to
RANDOM ASSIGNMENT, aS are INDEPEND- quasi-score estimating function
ENT VARIABLES. see QUASI-LIKELIHOOD FUNCTION.
quasi-interval scale a rating scale that questionnaire n. a set of questions or
classifies responses using ordered op- Other prompts used to obtain informa-
tions but lacks equal distances between tion from a respondent about a topic of
all scale points. For example, some re- interest, such as background characteris-
sponse items could show equal distances tics, attitudes, behaviors, personality,
between scale points, whereas for others ability, or other attributes. A question-
respondents could have a difficult time naire may be administered with pen and
differentiating among options, leading paper, in a face-to-face interview, or via
 quotient

interaction between the respondent and dividuals than are needed for a srady. It
a computer or website. also allows a researcher to obtain a de-
quick-and-dirty adj. describing a re- sired balance of sample sizes across
search design or data analysis that is ad- groups for statistical testing. Most com-
mitted to be informal and imperfect. puterized surveys include an automatic
The researcher who adopts such an ap- quota control function. See also QUOTA
proach recognizes that there is a more SAMPLING.
refined way to set up the experiment or quota s a m p l i n g a method of forming
to analyze the findings. In research, a a sample in which a prespecified number
quick-and-dirty approach is more infor- of individuals with specific background
mal than a PILOT STUDY; in data analy- characteristics, such as a particular age,
sis, it is regarded as a first peek at data race, ethnicity, sex, or education, are se-
from a specific analysis. lected for inclusion. Often, participants
q u i n t i l e n. one of four values in a score are recruited as they arrive; once the
distribution that divides it into five quota for a given demographic group is
equal parts. For example, the first filled, the researcher stops recraiting
quintile of a distribution is the value subjects from that group. A researcher
below which are the lowest 20% of who uses this approach can obtain a
scores. See QUANTILE. final srady sample that has the same
proportional characteristics as the target
quota c o n t r o l in survey methodology, population, enabling statistical testing
an approach that imposes a limit on the to be performed on a subset of cases that
number of respondents that are ob- is appropriately representative of the
tained either in the total sample or in larger group of interest. See DELIBERATE
substantively meaningful subgroups, SAMPLING. See also QUOTA CONTROL.
such as gender or ethnicity. The ap-
proach is efficient because it does not quotient n. the value that is the result of
overase resources, such as respondent dividing one number or quantity by an-
time or good will, by obtaining more in- other.

287
 Rr
r symbol for SAMPLE CORRELATION COEF-
FICIENT, which is typically in the form of
a PRODUCT-MOMENT CORRELATION CO-
EFFICIENT.
r^ symbol for COEFFICIENT OF DETERMI-
NATION.
R 1 . abbreviation for RESPONSE or respon-
dent. 2. symbol for MULTIPLE CORRELA-
TION COEFFICIENT.
symbol for COEFFICIENT OF MULTIPLE
DETERMINATION.
r a d i a l plot a SCATTERPLOT used in
META-ANALYSIS to assess Whether there
is HETEROGENEITY In the findings from
different studies, which could limit in-
terpretation of results. It depicts an EF-
FECT SIZE divided by its STANDARD
ERROR on the vertical y-axis and the
ratio of 1 divided by the standard error
on the horizontal x-axis, as in the hypo-
thetical example below.
Each point represents the strength of
the effect, relative to the standard error
N
w 0
U
T"
2 4
Precision (1/S)
radial
288
for each srady involved in the meta-
analysis. If an effect is large and the stan-
dard error is small, the point for that
study will appear in the upper right cor-
ner of the radial plot. A wide range of
points indicates that effect sizes and
standard errors vaty across studies (i.e.,
there is heterogeneity), which could
limit conclusions regarding the average
effect size in the meta-analysis and sug-
gests that variables potentially causing
this heterogeneity should be explored.
Also called Galbraith plot.
r a d i a n (rad) n. a unit of angular mea-
sure. It is calculated as the length of an
arc divided by its radius. 1 rad = 180/jt.
R^adj symbol for ADJUSTED R^.
r a n d o m adj. 1. without order or predict-
ability. 2. determined by chance alone,
as in RANDOM SAMPLING or a RANDOM
ERROR.
r a n d o m assignment in experimental
design, the assignment of participants or
95% confidence
limit
95% confidence
limit
T"
6
plot
 random error

units to the different conditions of an values or levels are drawn randomly

experiment entirely at random, so that from some larger (conceptual) popula-
each unit or participant has an equal tion of levels that could (in principle)
likelihood of being assigned to any par- have been selected. For example, a
ticular condition. In clinical trials this health researcher investigating the rela-
decreases the CONFOUNDING of the tionship between exercise and weight
treatment factor with other factors by may select a few levels of daily exercise
making the treatment and control for study (e.g., 0 hours, between 0 and 1
groups approximately comparable in all hour, between 1 and 2 hours, between 2
respects except for the treatment. Also and 3 hours) from a wide range of possi-
called randomization; random allo- ble options. Results involving a random
cation. See also RANDOMIZED GROUP effect can be generalized to values be-
DESIGN. yond those observed and modeled in the
study analysis. Also caUed random fac-
r a n d o m coefficient model see RAN- tor. Compare FIXED EFFECT.
DOM INTERCEPT MODEL.
random-effects analysis o f vari-
random-digit d i a l i n g (RDD) a sur- ance a statistical procedure in which
vey strategy in which participants are the variability of an outcome (typically a
contacted randomly by telephone. continuously measured variable) is ac-
Usually, the researcher selects the area counted for by several different factors
code(s) and then uses a computer pro- or predictors, each of which reflects a
gram to select the last digits at random. sampling of possible factor levels. The
This increases the likelihood of contact- focus in a random-effects analysis of
ing participants with unlisted numbers. variance is upon identitying differences
A concern with this strategy is that in the mean values obtained on an out-
nonresponders may differ from respond- come variable at the different levels of
ers in ways that introduce bias. The ap- the predictors sampled. Compare FIXED-
proach also assumes that potential E F F E C T S ANALYSIS OF VARIANCE.
respondents have telephones that are in
service. random-effects model any statistical
procedure or experimental design that
random digits a series of numbers gen- involves RANDOM EFFECTS. For example,
erated by a process that relies on chance a researcher wishing to investigate the
for each value, so that the occurrence of effects of temperarare on frequency of
any number in the list is unrelated to the aggressive behavior could not easily ex-
occurrence of any other. Such lists are amine each temperature value and so in-
often used in RANDOM SAMPLING. Po- stead examines a random sample of such
tential participants are each given a values and their effects. Although ran-
number and the sample is drawn by dom-effects models tend to be less pow-
choosing those with numbers in the ran- erful than FIXED-EFFECTS MODELS, they
domly generated list. This ensures that enable generalization to be made to lev-
selections are not driven by any known els of the independent variable not acm-
or unknown biases on the part of the re- ally employed in the study. Also called
searchers. Lists of random digits are now random model; variance compo-
produced mainly by computer program nents model. Compare MIXED-EFFECTS
(see RANDOM NUMBER GENERATOR). For- MODEL.
merly, they were often printed as tables
in the back of statistics textbooks. Also random error error that is due to chance
called random numbers. alone. Random errors are nonsystematic
and occur arbitrarily when unknown or
r a n d o m effect an INDEPENDENT VARI- unconttolled factors affect the variable
ABLE in an experimental design whose being measured or the process of measure-

289
random error variance

ment. Such errors are generally assumed tions components of VARIANCE at sev-
to form a NORMAL DISTRIBUTION around a eral levels. The similar random coeffi-
TRUE SCORE. AIso Called unbiased error; cient model (or random slope model)
unreUabiUty; unsystematic error; allows each data cluster to have its own
variable error. See also ABSOLUTE SLOPE, indicating the explanatory vari-
ERROR; CONSTANT ERROR. Compare SYS- able has a different effect upon each.
TEMATIC ERROR.
r a n d o m i z a t i o n n. see RANDOM AS-
r a n d o m error variance that part of SIGNMENT.
the variability of a test or other score randomization test an inferential ap-
that is not accounted for by other mod- proach that combines observed data
eled factors. It reflects RANDOM ERRORS across all participants and experimental
in measurement arising from the re- conditions and then randomly sorts the
spondent, the situation, evaluator idio- data into new samples. A test of STATIS-
syncrasies, the measure itself, or TICAL SIGNIFICANCE is performed, and
interactions among these factors. Differ- the value obtained is compared with
ent statistical frameworks handle this the value that was obtained when the
type of variation differently; some com- data were in their original form. This
bine it with SYSTEMATIC ERROR VARI- process is repeated many times, theoreti-
ANCE whereas others treat it separately. cally for all possible rearrangements
See ERROR VARIANCE. (permutations) of the data, although the
r a n d o m event an EVENT that is gener- sheer number of possible permutations
ated by a chance process and cannot be generally precludes this and a subset of
predicted from any other event. permutations (10,000 often is recom-
mended) is used instead. For example,
r a n d o m event generator (REG) see suppose there were 1,000 recalculated
RANDOM NUMBER GENERATOR. mean differences between the experi-
r a n d o m f a c t o r see RANDOM EFFECT. mental (E) and control (C) groups, de-
rived from 1,000 permutations. Com-
r a n d o m group design see RANDOM- parison of the 1,000 recalculated means
IZED GROUP DESIGN. to the original means provides an exact
probability of getting the original means.
r a n d o m intercept model a type of If 25% of the mean differences between
HIERARCHICAL LINEAR MODEL USed tO
the E and C group permutations differ
describe the correlations among obser- by as much or more than the difference
vations or scores within a cluster. For ex- between the original E and C means,
ample, a researcher examining the then the probability value for the ori-
average number of work hours and ginal mean difference is p = .25. A ran-
burnout among nurses at several differ- domization test is a NONPARAMETRIC
ent hospitals might use this approach to approach; that is, it does not make as-
account for dependence among certain sumptions about the distribution of the
variables, determining both the overall data. Also called permutation test.
score variation across all nurses and the
variation among the nurses in each of r a n d o m i z e d b l o c k design (RBD) an
the individual hospitals. When plotted, approach to assigning participants to
the data for each nurse grouping have treatment conditions in which mean-
a different INTERCEPT, thus providing ingful discrete strata within the sample
a more accurate and comprehensive de- (e.g., gender, experience) are used to
scription of relationships that corrects identity homogeneous subsamples; in-
for the underestimated STANDARD ER- dividuals from each subsample or
RORS obtained when other models are "block" are then assigned randomly to
applied to nested data and that parti- the different conditions. In this way.
 randomized-response technique

participants are initially matched on a randomized c o n t r o l l e d t r i a l see

"blocking variable" that the researcher RANDOMIZED CLINICAL TRIAL.
wishes to control. The acknowledgment
of heterogeneity within the sample en- randomized design any of various ex-
ables the researcher to reduce within- perimental designs in which individual
group variance and to use that informa- participants are assigned to different
tion when evaluating treatment effects. conditions (groups) using a purely
The variable on which participants are chance process, such as rolling a die. A
stratified is assessed prior to the srady. crucial assumption underlying random-
Also called randomized complete ized designs is that any systematic differ-
block design; treatment-by-blocks ences between treatment groups will be
design. due to the experimental conditions
themselves and not to any other unmea-
randomized block one-way analy- sured factors. Compare NONRANDOM-
sis of variance a model used to evalu- IZED DESIGN.
ate mean differences on an outcome
variable across three or more levels of a randomized f i e l d t r i a l a design that
factor, while allowing for an additional tests the effectiveness of a social inter-
variable used in assigning participants to vention in settings outside the labora-
conditions (e.g., gender). It is assumed toty and incorporates RANDOM ASSIGN-
that the additional variable will not in- MENT of participants to different study
teract with the main factor of interest. conditions. Such designs trade the con-
Also called one-way blocked analy- trol of the laboratory for the chance to
sis of variance. See ONE-WAY ANALYSIS observe the acraal behavior of partici-
pants in their everyday settings.
OF VARIANCE; RANDOMIZED BLOCK DE-
SIGN. randomized group design an experi-
mental design that involves the use of a
r a n d o m i z e d c l i n i c a l t r i a l (RCT) an
purely chance process, such as the toss
experimental design in which patients
of a coin, to assign participants to the
are randomly assigned to either a group
different study conditions.
that will receive an experimental treat-
ment, such as a new drug, or to one that randomized-response technique
will receive a comparison treatment, (RRT) a procedure for reducing SOCIAL
standard-of-care treatment, or a PLA- DESIRABILITY bias when measuring sen-
CEBO. The RANDOM ASSIGNMENT occurs sitive attitudes (e.g., racial attitudes) or
after recruitment and assessment of eli- behaviors (e.g., drug use, eating behav-
gibility but before the intervention. ior) at an aggregate group level. Respon-
There may be multiple experimental dents are presented with a pair of
and comparison groups, but each pa- questions that have dichotomous re-
tient is assigned to one group only. Also sponse options (e.g., agree or disagree,
called randomized controlled clini- yes or no), one question being the target
cal trial; randomized controlled question (sensitive question) and the
trial. Compare NONRANDOMIZED CLIN- other an innocuous filler question. They
ICAL TRIAL. are instructed to roll a die (or use a simi-
lar randomization procedure) to deter-
r a n d o m i z e d complete block de- mine which question they should
sign see RANDOMIZED BLOCK DESIGN. answer and to conceal the result of this
roll from the interviewer; they then pro-
randomized consent design see
vide the answer to that question but do
ZELEN'S DESIGN.
not tell the interviewer which one it is.
randomized c o n t r o l l e d c l i n i c a l The ambiguity regarding which ques-
t r i a l see RANDOMIZED CLINICAL TRIAL. fion has been answered is assumed to re-

291
random model

duce participants' concerns about the may be analyzed according to their

social desirability of their answers. De- probability but not otherwise predicted.
spite the fact that the interviewer does See STOCHASTIC.
not know which question each person
r a n d o m response set the tendency
has answered, PROBABILITY THEORY can
for certain research participants to an-
be used to estimate the distribution of
swer study questions in a way that is or
responses to the target question in the
appears to be based on a chance process.
population.
This can occur when participants are not
r a n d o m model see RANDOM-EFFECTS taking the study seriously (e.g., in a rush,
MODEL. tired, distracted, experiencing low moti-
vation). If this is detected, a researcher
r a n d o m n u m b e r generator (RNG) a may decide to discard these responses
device or system used to produce a ran- because they were generated by a chance
dom output of numbers. Such RANDOM process only and not for any reasons
DIGITS have various experimental uses, that are related to the constructs under
including the RANDOM ASSIGNMENT of study.
participants to treatment conditions,
thereby taking the decision of assign- r a n d o m s a m p l i n g a process for select-
ment out of the hands of the researcher. ing a SAMPLE of srady participants from
Most current RNGs operate by computer a larger potential group of eligible indi-
program; strictly speaking, such pro- viduals, such that each person has the
grams produce a "pseudorandom" out- same fixed probability of being included
put because the algorithms that they use in the sample and some chance proce-
rely on a nonrandom system. Also called dure is used to determine who specifi-
random event generator (REG). cally is chosen. A group selected in this
way is known as a random sample. The
r a n d o m numbers see RANDOM DIG- main value of this form of PROBABILITY
ITS. SAMPLING is its positive impact on
r a n d o m number table a table of RAN- GENERALIZABILITY and EXTERNAL VALID-
DOM DIGITS. Historically, before the ITY.
widespread use of computers to generate r a n d o m selection any procedure for
random values, such tables were often sampUng a set of participants or units
printed as an appendix in statistics text- from a larger set that relies on the use of
books. A researcher requiring random a chance process to minimize risk of re-
numbers for a srady would close his or searcher bias, either conscious or uncon-
her eyes and point to a start place on the scious.
page. Next, with eyes open, he or she
would use a predetermined way of mov- r a n d o m series a series of values that are
ing through the table to select values produced by a chance process, such as
(e.g., skip every other value). The re- a flip of a coin. The WALD-WOLFOWITZ
searcher could then be assured that a TEST is an inferential test for evaluating
chance process characterized the selec- whether a string of numbers is indeed
tion of values for the srady. demonstrating this chance process. See
RANDOM DIGITS; RANDOM NUMBER TABLE.
r a n d o m observation any observation
r a n d o m set a set of items (e.g., mea-
that results from a chance process (such
sures, experimental stimuli) that are
as a flip of a coin), is uncontrolled, or is
chosen by a chance process, such as a
not part of a schedule or pattern of orga-
flip of a coin. For example, to save time
nized observation.
and reduce participant burden, a re-
r a n d o m process a process that relies searcher might ask participants to com-
on chance alone, such that outcomes plete a subset of possible items rather
 rank correlation coefficient

than the entire set; the subset might be crepency between the low and high
chosen for each participant using a pro- scores, the range is generally perceived
cess determined by chance. as less informative than other measures
of dispersion, such as the STANDARD DE-
r a n d o m slope model see RANDOM IN-
VIATION.
TERCEPT MODEL.
r a n d o m variable a variable that takes range restriction see RESTRICTION OF
on different values according to a chance RANGE.
process. These values cannot be predicted r a n k 1. n. a relative position along an or-
with certainty and are assumed to vaty dered continuum. See RANK ORDER. 2.
across sradies; however, their frequency vb. to arrange items in a graded order,
can be described in terms of probability. such as from highest to lowest value. In a
Also called stochastic variable. Com- peer nomination study, for example, a
pare FIXED VARIABLE. child might be asked to order individu-
r a n d o m v a r i a t i o n differences in a DE- als in a class from most disruptive to least
PENDENT VARIABLE that are due to disruptive. See also ORDINAL DATA. 3. n.
chance, rather than to the factors being the maximum number of linearly inde-
studied. Causes of random variation in pendent row vectors or column vectors
test results may include respondent in a CORRELATION MATRIX: These values
factors, such as health, motivation, at- are always equal. For example, a 10 x 10
tention, concentration, and fatigue; sit- correlation matrix of personality scores
uational factors, such as room tempera- might have a rank of 6, indicating that
ture, noise, and working environment; there are not 10 independent pieces of
or respondent-by-siraation factors, such information present in the 10 scores.
as a respondent not being prepared for r a n k biserial correlation coeffi-
the specific rating task. Researchers try cient an index of association between a
to estimate the extent to which these DICHOTOMOUS VARIABLE and an ORDI-
factors may be involved in the srady (see NAL VARIABLE. Its interpretation is the
RANDOM ERROR VARIANCE) to under-
same as for other standardized measures
stand the true impact of the factors of association. For example, a researcher
being assessed. See also CHANCE DIFFER- might relate experimental condition
ENCE.
(experimental vs. control group) to an
r a n d o m w a l k a series of values plotted ordinal measure of task performance.
over time that reflects the workings of a Compare POINT BISERIAL CORRELATION
chance process, such that each value has COEFFICIENT.
an unpredictable relation to the preced-
r a n k correlation coefficient a nu-
ing value and the series has no definable
merical index reflecting the degree of re-
pattern. The concept is used in many re-
lationship between two variables that
search settings, including simulation
have each been arranged in ascending
sradies and models of price movements
or descending order of magnitude (i.e.,
on the stock market. Compare MARKOV
ranked). It does not reflect the associa-
CHAIN. See also DRUNKARD'S WALK; STO-
tion between the actual values of the
CHASTIC
variables but rather that between their
range n. a measure of DISPERSION ob- relative position in the distribution. For
tained by subtracting the lowest score example, placement in a marathon race
from the highest score in a distribution. could be correlated with the runners'
For example, if the highest score on heights but in this case the two vari-
a test is 100 and the lowest score is ablesrace outcome and heightwould
10, then the range is (100 - 10) = 90 take the form first place, second place,
points. Because it describes a raw dis- and so on; and tallest, next tallest, and

293
rank-flifference correlation

so on, respectively (rather than actual group differences or associations be-

times run in the race and specific heights tween ranked variables. Such tests make
in feet and inches). Among the most use of a RANK CORRELATION COEFFI-
commonly used such indexes are the CIENT, such as the SPEARMAN CORRELA-
SPEARMAN CORRELATION COEFFICIENT TION COEFFICIENT or KENDALL'S TAU.
and KENDALL'S TAU. Also called rank-
r a n k regression a type of REGRESSION
order correlation coefficient.
ANALYSIS in which independent vari-
rank-difference correlation a spe- ables are used to predict the rank (as op-
cific approach to calculating the SPEAR- posed to the acraal value) of a
MAN CORRELATION COEFFICIENT, which dependent or response variable. This ap-
involves computing discrepancies be- proach is particularly useful for distribu-
tween the relative positions of an indi- tions of errors that lack NORMALITY in
vidual on two variables. their end values. The relative efficiency
of a rank regression as compared to a
r a n k e d data see ORDINAL DATA. LEAST SQUARES REGRESSION Is more than
r a n k e d d i s t r i b u t i o n a set of values on 95%. See NONPARAMETRIC REGRESSION.
a variable sorted in magnitude from low-
rank-sum test any NONPARAMETRIC
est to highest. The entries in a CUMULA-
TEST that involves combining the data
TIVE FREQUENCY table are an example of
points from two or more samples in a sin-
a ranked distribution.
gle data set and ranking these values in
r a n k i n g experiment a srady in which ascending order. See MANN-WHITNEY U
the researcher asks participants to make TEST; WILCOXON-MANN-WHITNEY TEST;
a series of comparisons among stimuli WILCOXON RANK-SUM TEST.
(such as pictures, words, or emotions) so
r a n k t r a n s f o r m a t i o n a class of
that the stimuli can be ordered on some
TRANSFORMATION in which a partici-
dimension of interest (e.g., size, prefer- pant's score on a variable is replaced by
ence, cost, importance). For example, the rank position of the score relative to
a researcher conducting a marketing the other scores in the data set. For ex-
study might ask respondents about their ample, an instructor might modify an
preferences among certain products by original distribution of exam scores for a
ranking these from most likely to buy to class into a listing that is ordered from
least likely to buy, or by comparing two highest to lowest. Rank transformations
products at a time and indicating which serve as the basis for a wide variety of
one of the two would be preferred.
NONPARAMETRIC TESTS.
r a n k order the arrangement of a series Rasch model in ITEM RESPONSE THE-
of items (e.g., scores, individuals) in order ORY, a model in which only one parame-
of magnitude. ter, item difficulty, is specified. This is
rank-order data see ORDINAL DATA. thought to be a parsimonious way to de-
scribe the relation between an item re-
rank-order method a procedure in sponse and an underlying dimension
which a participant sorts various srady
and is thus preferred in some cases. Also
stimuli (e.g., cards, pictures, words,
called one-parameter model. See
people) from highest to lowest on a di-
TWO-PARAMETER MODEL; THREE-PARA-
mension of interest. See RANKING EXPER-
METER MODEL, [proposed in 1960 by
IMENT.
Georg Rasch (1901-1980), Danish statis-
rank-order scale see ORDINAL SCALE. tician]
rank-order statistic test any NON- ratee n. an individual who is being rated
PARAMETRIC TEST that allows research- on a particular dimension. Compare
ers to evaluate hypotheses related to RATER.
 rationalism

rate o f change the amount of change characterized by a particular attribute

in a variable per unit time divided by the that a respondent judges to be present
value of the variable before the change. (often scored as 1) versus absent (often
If a score rises from 20 to 30 in unit time, scored as 0). The number of items can be
for example, the rate of change is (30 - summed to reflect the extent that the at-
20)/20 = 10/20 = 0.5. tribute is present for the respondent. For
example, a daily hassles checklist might
rate of response see RESPONSE RATE.
list "flat tire," "visit from mother-in-
rater n. a judge or evaluator who assesses law," or "difficult work project due,"
a person or other unit on a characteristic and a sum of this list might reflect the
of interest. To understand bullying on amount of stress that a person is experi-
the playground, for example, a re- encing.
searcher might ask three independent
and trained individuals to rate a specific ratio n. the quotient of two numbers,
child's behavior during school recess, that is, one number divided by the other
usually on a particular scale (e.g., type of number.
interactions in which the child is en- r a t i o data numerical values that indi-
gaged). The consistency of the judg- cate magnitude and have a true, mean-
ments made by these three individuals ingful zero point. Ratio data represent
could then be evaluated (see INTER- exact quantities of the variables under
RATER RELIABILITY). Compare RATEE. consideration, and when arranged con-
secutively they have equal differences
rater r e l i a b i l i t y see INTERRATER RELI- among adjacent values (regardless of the
ABILITY.
specific values selected) that correspond
r a t i n g n. a score assigned to a person or to genuine differences between the
object on a numerical scale (e.g., 1 to 5) physical quantities being measured. In-
or a verbal scale (e.g., very good to very come provides an example: The differ-
poor). See RATING SCALE. ence between an income of $40,000 and
$50,000 is the same as the difference be-
r a t i n g error an incorrectly assigned tween $110,000 and $120,000, and an
RATING. Error of this kind may be owing income of $0 indicates a complete and
to several types of bias, including END-
genuine absence of earnings. Ratio data
AVERSION BIAS, HALO EFFECTS, leniency
are continuous in nature (i.e., able to
effects, and primacy effects. For exam-
take on any of an infinite variety of
ple, if a rater always describes the first of
amounts) and of the highest MEASURE-
two targets presented in more favorable
MENT LEVEL, surpassing INTERVAL DATA,
terms, then a primacy effect may ac-
ORDINAL DATA, and NOMINAL DATA in
count for these assigned responses.
precision and complexity.
r a t i n g scale an instrament that is used
r a t i o n a l i s m n. 1. any philosophical po-
to assign scores to persons or items along
sition holding that (a) it is possible to
some numerical dimension, such as
agreement with an attirade statement or obtain knowledge of reality by reason
frequency of occurrence. Rating scales alone, unsupported by experience, and
can be classified according to the num- (b) all human knowledge can be brought
ber of points along the dimension that is within a single deductive system. This
being assessed (e.g., a 5-point scale, 7- confidence in reason is central to classi-
point scale) and the way in which the re- cal Greek philosophy, notably in its mis-
sponse labels are ordered along the di- trust of sensory experience as a source of
mension. See BIPOLAR RATING SCALE; truth and the preeminent role it gives to
LIKERT SCALE; SEMANTIC DIFFERENTIAL. reason in epistemology. However, the
term "rationalist" is chiefly applied to
r a t i n g scale checklist a list of items thinkers in the Continental philosophi-

295
rational number

cal tradition initiated by French philoso- TRANSFORMATIONS, and quantitative or

pher Rene Descartes (1596-1650), most qualitative analysis. For example, a sur-
notably Dutch Jewish philosopher vey may ask respondents to enter their
Barach Spinoza (1632-1677) and Ger- annual income in dollars: The figures
man philosopher Gottfried Wilhelm supplied by respondents would be the
Leibniz (1646-1716). Rationalism is usu- raw data. For the purposes of analysis,
ally contrasted with EMPIRICISM and however, the researcher may prefer to
POSITIVISM, which hold that knowledge use data that have been cleaned to ac-
comes from or must be validated by sen- count for improbable entries or individ-
sory experience. In psychology, psy- uals who prefer not to answer and then
choanalytical approaches, humanistic recoded to create a smaller set of income
psychology, and some strains of cogni- categories.
tive theoty are heavily influenced by ra-
tionalism. 2. in general language, any raw score a participant's score on a test
position that relies on reason and evi- before it is converted to other units or
dence rather than on faith, intuition, another form or subjected to quantita-
custom, prejudice, or other sources of tive or qualitative analysis. For example,
conviction. rationalist adj., n. a score may be transformed into a per-
centage (e.g., 45 correct answers out of
r a t i o n a l number any value that can 50 = 90%) or into a standardized metric
be expressed as the ratio between two in- such as a z SCORE (mean of 0; standard
tegers (e.g., 2/3 or 8/1). A ratio with zero deviation of 1) or a T SCORE (mean of 50
as the denominator would not be a ratio- standard deviation of 10). Also called
nal number. unstandardized score.
ratio scale a measurement scale having r a w score p a r t i a l regression coef-
a true zero (i.e., zero on the scale indi- f i c i e n t in REGRESSION ANALYSIS, the
cates an absence of the measured attrib- average oi: expected change in the DE-
ute) and a constant ratio of values. Thus, PENDENT VARIABLE for each increase in
on a ratio scale an increase from 3 to 4 the INDEPENDENT VARIABLE, with all
(for example) is the same as an increase other independent variables in the
from 7 to 8. The existence of a true zero model held constant. It is expressed in
point is what distinguishes a ratio scale the units of the variable being measured,
from an INTERVAL SCALE. whereas a PARTIAL REGRESSION COEFFI-
ratio score 1. formerly, in the early days CIENT is derived from data that have un-
of the aptitude testing of children, a dergone STANDARDIZATION.
score expressed as a ratio of the child's Rayleigh d i s t r i b u t i o n a special type
mental age to his or her chronological of CONTINUOUS DISTRIBUTION often
age multiplied by 100. Thus, a child who used in the analysis of data having both
was 8 years old but solved problems that speed and direction components. For ex-
normally could be solved only by chil- ample, in psychology it may be applied
dren of 10 years old had a ratio score or in the assessment of human auditory
ratio IQ of 120 [(12/10) x 100]. 2. any processing, [fohn William Strutt, Lord
score that is expressed as a ratio of one Rayleigh (1842-1919), British physicist]
value to another.
Rayleigh test a test for SIGNIFICANCE of
ratio variable a variable that is mea-
directional data that form a circular pat-
sured with a RATIO SCALE (e.g., height or
tern (e.g., movements of captive animals
weight). See RATIO DATA.
away from a point of release into the
r a w data the original measurements on wild). The test involves calculating the
a variable as collected by the researcher, mean of a number of circular means; the
prior to data cleaning, RECODING. results are the mean angle, the length of
 recall data

the mean vector, and a CONFIDENCE IN- reactivity n. the condition in which a
TERVAL around the mean angle. [Lord participant being observed is changed
Rayleigh] in some way by the act of observation.
Within an experimental setting reactiv-
r^ symbol for BISERIAL CORRELATION CO- ity is viewed as a threat to INTERNAL VA-
EFFICIENT.
LIDITY because the change in behavior is
RBD abbreviation for RANDOMIZED not due to the experimental manipula-
BLOCK DESIGN. tion. See also REACTIVE MEASURE.
'bis symbol for BISERIAL CORRELATION reactivity effect any of several specific
COEFFICIENT. ways in which a target's performance
may change as a result of being observed
Rc symbol for CANONICAL CORRELATION
or receiving increased attention within
COEFFICIENT, A further subscript may be
a research setting. Examples include EX-
added to show which canonical variates
PERIMENTER EFFECTS, the HAWTHORNE
are being correlated. EFFECT, the PYGMALION EFFECT, and
R correlation see MULTIPLE CORRELA- SELF-FULFILLING PROPHECY effects.
TION COEFFICIENT.
real l i m i t the lower or upper bound of a
RCT abbreviation for RANDOMIZED CLIN- given value for a continuous variable
ICAL TRIAL. measured on a RATIO SCALE. For exam-
ple, a test score of 95 has the lower real
R D D 1. abbreviation for RANDOM-DIGIT
limit of 94.5 and the upper real limit of
DIALING. 2. abbreviation for REGRES-
95.5 since any value within that range
SION-DISCONTINUITY DESIGN.
will equal 95 when rounded to a whole
reaction time (RT) the time that number.
elapses between the onset or presenta-
real-world setting conditions for data
tion of a stimulus and occurrence of a
collection that closely resemble condi-
specific response to that stimulus. There
tions in the natural flow of life (e.g., ob-
are severaltypes,including simple reac-
serving workers at their workplace rather
tion time (single stimulus and single re-
than having them report about what oc-
sponse) and choice reaction time (two or
curs at work). A srady conducted in a
more stimuU, each with, a different re-
real-world setting is often viewed as
sponse). Reaction time can be used to as-
more ecologically valid and gen-
sess various psychological constructs. To
eralizable than one conducted in a labo-
assess negative affect, for example, a re-
ratoty setting.
searcher might measure the time be-
tween presentation of various words recall bias the type of BIAS that often
with emotional connotations and a par- occurs when an individual reports about
ticipant's indication that the word was a past behavior or event. Although such
either "positive" or "negative." Also retrospective reporting may have accu-
called response latency; response rate fearares, it also tends to include in-
time. accurately remembered aspects, such as
a systematic undercount or overcount of
reactive measure a measure that alters
the ftequency with which a certain be-
the response under investigation. For ex-
havior occurred. This type of distortion
ample, if participants are aware of being
is discussed in the literatures associated
observed, their reactions may be influ-
with survey methodology and eyewit-
enced more by the observer and the fact
ness testimony.
of being observed than by the stimulus
object or situation to which they are os- recall data responses to questions that
tensibly responding. Compare UNOB- ask what a participant can remember
TRUSIVE MEASURE. about something. For example, a partici-

297
receiver-operating characteristic curve

pant may be asked to read a passage and l/x and that of 1/x is x. So, for example,
later list specific features of it from mem- 1/4 is the reciprocal of 4.
ory. Another type of memory task might
involve asking the participant whether reciprocal relationship 1. a correla-
or not specific items have been pre- tion between two variables such that the
sented before (the recognition method). value on one variable is the RECIPROCAL
of the value on the other. For example, if
receiver-operating characteristic a researcher is studying the average time
curve (ROC curve) in a detection, dis- taken to complete a task, then tasks
crimination, or recognition task, the re- completed per unit time (e.g., two per
lationship between the proportion of hour) has a reciprocal relationship with
correct "yes" responses (hit rate) and the unit time taken per task (0.5 hours). 2.
proportion of incorrect "yes" responses the situation in which two variables can
(false alarm rate). This is plotted on a mutually influence one another, that is,
graph to show an individual's sensitivity each can be both a cause and an effect.
on the particular task: The axes are hit
and false alarm rates, points are marked reciprocal transformation a TRANS-
to denote the different rates obtained FORMATION of raw data that involves (a)
under different condifions, and the points replacing the original data units with
are connected to form a smooth area (see their RECIPROCALS and (b) analyzing
iUustration below). the modified data. It can be used with
For example, an ROC curve may be used nonzero data and is commonly used
to indicate how well a person detects a when distributions have SKEWNESS or
specific tone in the presence of noise. A clear OUTLIERS. Unlike other transfor-
single quantitative INDEX of performance mations, a reciprocal transformation
may be calculatedfromthe curve as well. changes the order of the original data.
Also called isosensitivity function. For example, if the original unit of a
study variable is time, a researcher might
reciprocal n. the number that when transform the raw data to produce an
multiplied by another number gives a re- analysis of rate. Also called inverse
sult of 1. The reciprocal of x is therefore transformation.

1 1 1 r
0.0 0.2 0.4 0.6 0.8 1.0
False Alarm Rate

receiver-Operating characteristic curve

298
 redundancy coefficient

recoding n. a manipulation of an origi- CHAID) to identity predictor variables

nal variable in a data set so that it can be that differentiate high from low values
used in a different way in furare analysis on a response variable or outcome of in-
(e.g., reverse keying items, collapsing terest. A researcher can identify through
many categories into just a few catego- this systematic search which predictors
ries). See also TRANSFORMATION. re- from a larger set are most associated with
code vb. different mean splits on an outcome.
The statistical significance of each split
recruitment bias see SAMPLING BIAS.
is described using an F TEST and can be
rectangular axes two coordinate lines corrected for MULTIPLE COMPARISONS.
(the ;it-axis and y-axis) that cross at the The analysis generates a tree, with the
origin (the point at which x and y are outcome at the top and branches show-
both 0), are at right angles, have a single ing various splits based on the different
unit length, and create a two- predictors and their association with the
dimensional plane. See CARTESIAN CO- outcome.
ORDINATE SYSTEM.
reduced m a x i m u m l i k e l i h o o d see
rectangular d i s t r i b u t i o n see UNI- RESTRICTED M A X I M U M LIKELIHOOD.
FORM DISTRIBUTION.
reduced model in the GENERAL LINEAR
rectilinear adj. moving in or formed by MODEL, a model that has fewer parame-
straight lines. For example, a rectilinear ters than the most highly parameterized
polygon is a figure whose edges all meet model in a set of models to be compared.
at right angles and a rectilinear path re- Usually, the smaller model is said to be
fers to motion or progress in a straight "nested" within the larger, more highly
line. In memory research, pictures or parameterized model.
photographs presented as stimuli are re-
reductionism n. the strategy of ex-
ferred to as rectilinear views.
plaining or accounting for some phe-
recursive adj. describing a rule or proce- nomenon or constract, A, by claiming
dure that is applied repeatedly for a fi- that, when properly understood, it can
nite number of times, with the output of be shown to be some other phenome-
each application becoming the input to non or construct, B, where B is seen to be
the next. For example, a recursive algo- simpler, more basic, or more fundamen-
rithm might be used to help identity the tal. The term is mainly applied to those
set of predictors that relate maximally to positions that attempt to understand
a specific outcome variable. human culrare, society, or psychology
in terms of animal behavior, physical
recursive model a set of relationships
laws, or biological phenomena.
in which the effects flow in one direc-
tion only and there are no feedback redundancy analysis a multivariate
loops such that effects are sometimes statistical model for examining the de-
also causes. In STRUCTURAL EQUATION gree to which one set of variables or
MODELING, for example, a recursive scores may maximally relate to a second
model in which independent variables such set. A researcher derives a number
lead to dependent variables without of CANONICAL VARIATES of interest from
feedback loops is generally more easily the first set of variables and examines
estimated. Compare NONRECURSIVE M O - the relation between these variates and
DEL. the variance in the second set. Redun-
dancy analysis is an alternative to CA-
recursive partitioning regression NONICAL CORRELATION ANALYSIS.
a CLASSIFICATION TREE Strategy that uses
a systematic algorithm (the CHI-SQUARE redundancy coefficient in MULTI-
AUTOMATIC INTERACTION DETECTOR Or VARIATE ANALYSIS, an asymmetric index

299
reference axis

showing the proportion of variance in ison reference population. 2. the target

the outcome or dependent variables that population itselL
is accounted for by a set of predictor or
independent variables. It can be differ- reference p r i o r a set of values based on
entiated from the CANONICAL CORRELA- expectation or beUef that is used as a
TION COEFFICIENT, which assumes that
standard starting point in certain multi-
the two sets of variables are symmetric. variate problems, especially in the
The redundancy coefficient is highly af- BAYESIAN tradifion. See PRIOR DISTRIBU-
fected by the unit of measurement; TION.
therefore, common metrics need to be "effect size S y m b o l f o r EFFECT-SIZE C O R R E -
established before it is applied. LATION.

reference axis in FACTOR ANALYSIS, any r e f l e x i v i t y n. 1. a bidirectional relation-

of a set of AXES that create a coordinate ship of cause and effect. 2. in QUALITA-
frame depicting the spatial configuration TIVE RESEARCH, the self-referential
of the dimensions underlying the rela- quality of a study in which the re-
tionships among a group of variables. searcher reflects on the assumptions be-
Thisframecan be rotated to improve the hind the study and especially the
interpretability of the configuration, influence of his or her own motives, his-
using either ORTHOGONAL or OBLIQUE tory, and biases on its conduct. See
methods. EPISTEMOLOGICAL REFLEXIVITY.
reference database a database of bib- refusal rate the proportion of poten-
liographic information in a particular tially eligible respondents for a survey or
field of study. Among other uses, such study who have been successfuUy con-
a resource may be invaluable for devel- tacted but will choose not to participate
oping a comprehensive review of the lit- for a variety of reasons (e.g., survey takes
eramre in a particular field or for too much time, respondent lacks inter-
supporting an empirical synthesis of est in topic or surveys in general, respon-
sradies in a meta-analysis. dent tires of answering questions and
reference d i s t r i b u t i o n see THEORETI- breaks off the interview midway). The
CAL DISTRIBUTION. refusal rate must be taken into account
when calculating the likely RESPONSE
reference i n t e r v a l a range of expected RATE for a survey. Also called non-
values regarding behavior, performance, response rate.
or clinical levels (e.g., blood pressure).
The limits of the interval are generally REG abbreviation for random event gen-
established through a systematic pro- erator. See RANDOM NUMBER GENERA-
cess, such as examination of past empiri- TOR.
cal studies. Also called normal range; region o f acceptance see ACCEP-
reference range. See also NORM.
TANCE REGION.
reference p o p u l a t i o n 1. a subset of a region o f rejection see CRITICAL RE-
TARGET POPULATION that serves as a GION.
standard against which research find-
ings are evaluated. For example, con- regressed change-score analysis a
sider an investigator examining the REGRESSION ANALYSIS in which DIFFER-
effectiveness of eating disorder preven- ENCE SCORES (scores based on two or
tion programs at four-year colleges and more measurements over time) are used
universities in the United States. In such as the outcome variables. For example, in
a situation, the portion of educational a PRETEST-POSTTEST DESIGN a researcher
institutions that have not implemented could "remove" the pretest score from
any program would serve as the compar- the posttest score to create a difference

300
 regression-discontinuity design

score: This could then be used as the out- with a unit change in a specific
come variable in a regression analysis. independent (predictor) variable on the
dependent (outcome) variable, given
regression analysis any of several sta- the relationship of that predictor to
tistical techniques that are used to de- other independent variables already in
scribe, explain, or predict (or all three) the model. This value may be standard-
the variance of an outcome or DEPEND- ized (see STANDARDIZATION) with a vari-
ENT VARIABLE using scores on one or ance equal to 1 (in which case it is called
more predictor or INDEPENDENT VARI- a BETA COEFFICIENT), or it may be un-
ABLES. Regression analysis is a subset of standardized and expressed in the units
the GENERAL LINEAR MODEL. It yields a of the outcome variable being measured
REGRESSION EQUATION as well as an (in which case it is called a B coefficient).
index of the relationship (R or r) be- Also called regression weight. See
tween the dependent and independent also PARTIAL REGRESSION COEFFICIENT.
variables. In addition, the regression
weights obtained for the various inde- regression constant the value of a re-
pendent variables provide information sponse or DEPENDENT VARIABLE in a RE-
about their relative predictive contribu- GRESSION EQUATION when its associated
tion to the outcome. For example, a re- predictor or INDEPENDENT VARIABLES
gression analysis could show the extents equal zero (i.e., are at baseline levels).
to which first-year grades in college (out- Graphically, this is equivalent to the Y-
come) are predicted by such factors as INTERCEPT, or the point at which the RE-
standardized test scores, courses taken in GRESSION LINE crosses the y-axis.
high school, letters of recommendation,
and particular extracurricular activities. regression curve see REGRESSION LINE.
Also called slope analysis. regression diagnostics a set of graph-
regression a r t i f a c t an experimental ical and numerical techniques routinely
finding that has been distorted by ex- used by researchers to check for VIOLA-
treme measurements and the associated TIONS OF ASSUMPTIONS in the applica-
influence of REGRESSION TOWARD THE
tion of REGRESSION ANALYSIS tO
MEAN. particular data sets. For example, one as-
sumes that the relationship between the
regression c a l i b r a t i o n 1. an ap- INDEPENDENT VARIABLES and the DE-
proach in which a researcher uses spe- PENDENT VARIABLE is linear, that the
cific values of a DEPENDENT VARIABLE to variables have been measured accu-
determine the associated values of an IN- rately, and that any prediction errors re-
DEPENDENT VARIABLE when the rela- sulting from the REGRESSION EQUATION
tionship between the two variables is are independent and normaUy distrib-
already known. For example, a re- uted with equal variance and a mean of
searcher sradying university faculty sala- zero. If the data do not possess such
ries might choose a very high salaty characteristics, the analysis may not be
(dependent variable) to see how many appropriate and thus its results may not
years it takes for a new faculty member be valid. See DIAGNOSTICS; RESIDUAL
at that instiration to achieve such a rate ANALYSIS.
of pay (independent variable). 2. a
method of adjusting POINT ESTIMATES regression-discontinuity design
and INTERVAL ESTIMATES for COEFFI- (RDD) a type of QUASI-EXPERIMENTAL
CIENTS in a prediction model to account DESIGN in which a specific threshold
for known MEASUREMENT ERROR. value or CUTOFF SCORE is used to assign
participants to treatment conditions.
regression coefficient in a REGRES- Theoretically, individuals near the
SION ANALYSIS, the WEIGHT associated threshold value are comparable and

301
regression effect

only differ on the basis of their treat- called regression formula; regres-
ment assignment, which enables a re- sion model.
searcher to estimate treatment effects.
For example, a researcher might use a re- regression estimate an EXPECTED
gression-discontinuity design to investi- VALUE for an outcome or DEPENDENT
gate worker performance, assigning VARIABLE that is calculated from a RE-
employees who work more than a cer- GRESSION EQUATION. For a typical uni-
tain number of hours to receive a reward variate linear model, in which multiple
while those below that threshold num- INDEPENDENT VARIABLES predict a single
ber do not. The analysis of such a design outcome, this value is obtained using
involves examining the REGRESSION LEAST SQUARES REGRESSION.
LINES for those receiving the treatment
(e.g., receiving a reward) versus those regression f o r m u l a see REGRESSION
not receiving the treatment (e.g., no re- EQUATION.
ward). A continuous straight line for the
two groups indicates no effect of reward regression i m p u t a t i o n see IMPUTA-
on performance, whereas any break or TION.
jump (discontinuity) in the line across
the groups indicates a treatment effect. regression line a straight or curved line
fitting a set of data points and usually
regression effect see REGRESSION TO- obtained by a LEAST SQUARES REGRESSION.
WARD THE MEAN. A graphic representation of the REGRES-
SION EQUATION expressing the hypothe-
regression equation the mathematical sized relation between an outcome or
expression of the relationship between DEPENDENT VARIABLE and One or more
a dependent (outcome or response) vari- predictors or INDEPENDENT VARIABLES, a
able and one or more independent (pre- regression line summarizes how well the
dictor) variables that results from proposed model acraally fits the sample
conducting a REGRESSION ANALYSIS. It data obtained.
often takes the form y = a + bx + e, in Data points that do not fall exactly on
which y is the dependent variable, x the line indicate deviations in model fit,
is the independent variable, a is the IN- as in the hypothetical example below
TERCEPT, b is the REGRESSION COEFFI- showing parent ratings of aggressive be-
CIENT, and e is the ERROR TERM. Also havior in their children as a function of

15-
o
regression line

O)
O)
< 5-

n\II\\ I I I I
0 1 2 3 4 5 6 7 8 9 10
Amount of TV Violence
regression line
 regression tree

amount of violence watched on televi- SQUARES and calculate an F RATIO. Also

sion. called explained sum of squares.
Compare ERROR SUM OF SQUARES.
regression m o d e l see REGRESSION
EQUATION. regression t h r o u g h the o r i g i n an
approach to REGRESSION ANALYSIS in
regression o f x o n y a method that
which the REGRESSION CONSTANT is re-
"reverses" the typical model of REGRES-
moved from the REGRESSION EQUATION
SION ANALYSIS by treating the INDE-
and the Y-INTERCEPT is zero rather than
PENDENT VARIABLES (;ifs) as DEPENDENT
nonzero as it is in ordinaty LEAST
VARIABLES (ys). In other words, one uses
SQUARES REGRESSION. Thus, when the
a REGRESSION EQUATION to estimate an
value of one or more predictor or inde-
X score from one or more y scores. Re-
pendent variables (xs) equals zero, the
gression of ;t on y typically is conducted
mean value of an outcome or'dependent
to better understand possible MULTI-
variable (y) also equals zero (i.e., the RE-
COLLINEARITY among the independent
GRESSION LINE goes through the origin).
variables. Also called inverse predic-
Use of regression through the origin is
tion. Compare REGRESSION OF Y ON X.
justified only in certain circumstances,
regression o f y on x a typical REGRES- such as when the dependent variable
SION ANALYSIS, in which one predicts has a true baseline value of zero (e.g.,
values of a DEPENDENT VARIABLE, y, zero dollars, zero cigarettes smoked, zero
from values of one or more INDEPEND- words recalled). Also called no-inter-
ENT VARIABLES, X, USing a REGRESSION cept model; regression without in-
EQUATION. Compare REGRESSION OF X tercept.
ON Y.
regression t o w a r d the mean the
regression plane the plane in three- tendency for extremely high or ex-
dimensional space created when graph- tremely low scores to become more
ing data points from a REGRESSION moderate (i.e., closer to the MEAN) upon
EQUATION in which more than one pre- retesting over time. In experimental
dictor or INDEPENDENT VARIABLE is asso- sradies this tendency threatens INTER-
ciated with an outcome or DEPENDENT NAL VALIDITY in that shifts of scores may
VARIABLE. The regression plane is flat be for reasons unrelated to srady manip-
and slices through EUCLIDEAN SPACE ulations or treatments. For example, re-
when the relationship between the vari- gardless of the interventions a researcher
ables is linear but becomes curved and is investigating to improve mathematics
flowing for nonlinear relationships. A performance (e.g., extra study sessions,
REGRESSION LINE is used for siraations providing positive or negative reinforce-
involving a single independent variable. ment), low scoring students will tend to
perform slightly better on the next math
regression sum o f squares (symbol:
exam, while high scoring students will
SSfegression) 3 number indicating the'
tend to perform slightly worse, RANDOM
amount of variance in a DEPENDENT
ASSIGNMENT to treatment and control
VARIABLE that can be explained by the
conditions may be used to minimize the
variance in one or more associated INDE- influence of regression toward the mean
PENDENT VARIABLES. It thus describes
upon experimental results. Also called
how well a particular model fits the ob- regression effect. See REGRESSION AR-
served data. For example, in LINEAR RE- TIFACT; THREATS TO VALIDITY.
GRESSION it is used to calculate a
COEFFICIENT OF DETERMINATION Or a regression tree a diagram displaying a
COEFFICIENT OF MULTIPLE DETERMINA- set of conditions and their associations
TION, and in ANALYSIS OF VARIANCE it is with a particular outcome variable as de-
used to determine the total SUM OF termined via-a REGRESSION ANALYSIS.

303
regression weight

An initial node represents the outcome significant differences between inde-

of interest (e.g., annual income under pendent groups in a ONE-WAY ANALYSIS
$15,000), from which branches extend OF VARIANCE of a BALANCED DESIGN. It iS
to additional nodes according to the val- one of many such MULTIPLE COMPARI-
ues of a studied predictor variable (e.g., SON TESTS. Also called Ryan's method.
level of education). Further branches ex- [Thomas A. Ryan Jr., U.S. statistician;. Is-
tend from these subnodes based on addi- rael Einot, Israeli statistician; K. Ruben
tional predictors (e.g., area of residence, Gabriel (1929-2003), German-born U.S.
type of employment), with the process statistician; Roy E. Welsch, U.S. statisti-
continuing until no more predictors are cian]
available in the data set or until a prede-
termined number of nodes is obtained. reinforcer effect a situation in which
Regression trees provide a convenient, one variable strengthens the relation-
visually appealing method for examin- ship between two other variables. For ex-
ing large amounts of data. See also CART ample, if performance on a free recall
task is enhanced when participants are
ANALYSIS. in a positive mood when studying words
regression weight see REGRESSION for the memory task, positive mood
COEFFICIENT. has demonstrated a reinforcer effect on
study recall. See also INTERACTION EF-
regression w i t h o u t intercept see FECT; MEDIATOR. Compare SUPPRESSOR
REGRESSION THROUGH THE ORIGIN. EFFECT.
regress o n to determine the extent to Reinsch spline an approach that fits a
which a given DEPENDENT VARIABLE (y) curve to TIME-SERIES DATA while mini-
can be explained or predicted by a num- mizing noise around the data points. It
ber of INDEPENDENT VARIABLES (XS). For is a useful nonparametric strategy for
example, a researcher may be interested SMOOTHING a REGRESSION. LINE repre-
in learning how scores on a measure of senting change over time, which typi-
relationship commitment vary as a func- cally contains many bends or turns. Also
tion of age, relationship staras, time in called Demmler-Reinsch spline. See
relationship, and shared experiences. SPLINE FUNCTION. [A. Demmler; Chris-
That is, the researcher may regress y on tian H. Reinsch]
X. See REGRESSION ANALYSIS.
reinterviewing n. collecting LONGITU-
regressor variable see INDEPENDENT DINAL DATA by interviewing the same
VARIABLE. . participants at several different points in
regularized discriminant analysis time. For example, a researcher investi-
gating familial interactions might con-
a procedure for classitying individuals or
duct multiple home visits to obtain
units into discrete groups based on a
numerous details about the parent-child
set of variables under study that repre-
relationship and how it evolves over
sents a compromise between linear
time. Potential problems with this ap-
DISCRIMINANT ANALYSIS and QUA- proach include participant REACTIVITY,
DRATIC DISCRIMINANT ANALYSIS. In gen- participant ATTRITION, and decreased
eral terms, it involves the adjustment of INTERNAL VALIDITY. See also RETEST RE-
two PARAMETERS to yield different com- LIABILITY.
binations or ways of classifying the
units, with the overall goal being to min- rejection error see TYPE I ERROR.
imize the risk of misclassification.
rejection method a technique that
REGWQ test Ryan-Einot-Gabriel- uses an algorithm to generate and select
Welsch multiple range test: a procedure random values for a study (see RANDOM
used to evaluate if there are statistically DIGITS). Values are automatically in-
304
 relative frequency distribution

eluded in or excluded from the srady to the measured value. For example, if
sample depending on whether they fall the value of a particular characteristic in
within a particular range. Also called ac- the population is known to be .7 but a
ceptance-rejection method; rejec- recently developed statistical procedure
tion sampling. estimated a value of .5, the relative error
rejection region see CRITICAL REGION. would be (.7- .5)/.7 = .2/.7 = .286. A rela-
tive error may be converted to a percent
rejection value see CRITICAL VALUE. error by multiplying by 100thus, for
related-measures design see WITHIN- the example given, the percent error is
SUBJECTS DESIGN. .286 X 100, or 28.6%.
related samples see DEPENDENT SAM- relative frequency the frequency of a
PLES. type or categoty of event expressed as a
proportion of the total frequency of all
related-samples design see WITHIN-
types or categories. For example, if 47 out
SUBJECTS DESIGN.
of 100 participants answered "yes" to a
related-samples t test see DEPEN- particular question on a survey, the rela-
DENT-SAMPLES T TEST. tive frequency of "yes" responses would
equal .47 (i.e., the number of "yes" re-
r e l a t i o n a l research see CORRE-
sponses divided by the total number
LATIONAL RESEARCH.
of "yes," "no," and "I don't know" re-
relationship n. an association or con- sponses).
nection between objects, events, vari-
ables, or other phenomena. Research relative frequency d i s t r i b u t i o n a
often involves the srady of associations tabular display of the number of obser-
between and among variables. See also vations at each level of a variable com-
CORRELATION. pared to the total number of observations
obtained. For example, a researcher asks
relative efficiency 1. for two tests (A 200 participants to describe their level
and B) of the same hypothesis operating of extraversion on a 5-point scale and
at the same SIGNIFICANCE LEVEL, the obtains the following results: 32 individ-
ratio of the number of cases needed by uals indicated they are extremely intro-
Test A compared to the number needed verted; 24 individuals indicated they are
by Test B for each to have the same sta- somewhat introverted; SO individuals
tistical POWER. The relative efficiency indicated they are neither introverted
value enables a researcher to determine nor extraverted; 38 individuals indi-
whether there is a preferred statistical cated they are somewhat extraverted;
approach for evaluating a particular and 56 individuals indicated they are ex-
phenomenon. For example, when hy- tremely extraverted. These values can be
pothesis testing involves NORMAL DIS- expressed as both proportionsthe
TRIBUTIONS, parametric inferential sta-
number of observations per level di-
tistics have a more favorable ratio than vided by all observations (i.e., .16, .12,
nonparametric inferential statistics. 2. .25, .19, and .28)and as percentages
for two parameter ESTIMATES (A and B), the number of observations per level di-
a value reflecting the ratio of the STAN- vided by all observations multiplied by
DARD ERROR of Estimate A compared to 100 (i.e., 16%, 12%, 25%, 19%, and
the standard error of Estimate B. See REL- 28%). Thus, in the corresponding rela-
ATIVELY EFFICIENT ESTIMATOR.
tive frequency distribution overleaf, the
relative error in measurement, the first column lists the different extra-
ratio of the ABSOLUTE ERROR (i.e., the version levels, the second column lists
positive difference between an exact the number of people at each of the dif-
measured value and the estimated value) ferent extraversion levels, the third col-

305
relative frequency polygon

Relative frequency Relative

Extraversion level Frequency ( f ) (proportion) frequency (%)

Extremely introverted 32 .16 16

Somewhat introverted 24 .12 12

Neither Introverted nor

50 .25 25
extraverted

Somewhat extraverted 38 .19 19

Extremely extraverted 56 .28 28

relative frequency distribution

umn lists each value in the second the horizontal ;t-axis with the interval
column as a proportion of the total, and midpoints from lowest to highest; (c) la-
the fourth column lists each value in the beling the vertical y-axis with the range
second column as a percentage of the of proportional score values; (d) plotting
total. the specific proportion associated with
This type of table is useful in identify- each midpoint; and (e) connecting aU of
ing which scores or values are most the plotted points with lines. A hypo-
likely to occur at which variable level, thetical example is given below.
and it provides an organized display of Also called relative frequency
data that could also be graphed in a REL- curve; relative frequency diagram;
ATIVE FREQUENCY POLYGON. AISO Called relative frequency function; rela-
relative distribution; relative fre- tive frequency graph. See FRE-
quency table. See FREQUENCY DISTRI- QUENCY POLYGON.
BUTION.
relatively efficient estimator given
two ESTIMATORS (approximated popula-
relative frequency p o l y g o n a graph-
tion values) for a given model, the esti-
ical representation of a RELATIVE FRE-
mator that has the smallest variance in
QUENCY DISTRIBUTION, conveying the
its sampling distribution and thus pro-
number of individuals responding in a
vides a more precise estimate of the pa-
particular manner on a given variable as
rameter. See RELATIVE EFFICIENCY.
compared to the total number of people
responding. For a CONTINUOUS VARI- relative measurement a context-
ABLE, the polygon is constructed by (a) specific assessment approach in which
creating intervals of scores and identify- the value of an individual score depends
ing each interval midpoint; (b) labeling on its comparison to other scores within

Extremely Somewhat Neither Somewhat Extremely

introverted introverted introverted nor extraverted extraverted
extraverted

Extraversion Level
relative frequency polygon

306
 reliability diagonal

a group or subgroup. For example, one diction. In ethics, relativism is the claim
might evaluate beauty by assigning a that there are no moral absolutes. Thus,
person a number that reflects his or her judgments of right and wrong are based
attractiveness compared with others of on local culrare and tradition, on per-
the same age, the same occupation, and sonal preferences, or on artificial princi-
so forth. Comparison of scores within ples. Standards of conduct vaty
subgroups can sometimes be useful but enormously across individuals, cultures,
also may produce bias. Compare ABSO- and historical periods, and it is impossi-
LUTE MEASUREMENT. ble to arbitrate among them or to pro-
duce universal ethical principles because
relative risk see RISK RATIO.
there can be no means of knowing that
relative standing the position of a these are trae. In this way, relativism in
particular score in the context of a distri- epistemology and relativism in ethics
bution of scores. There are several ways are related. relativist adj.
to describe relative standing, including
PERCENTILES (e.g., deciles, quartiles), relevance-sensitivity tradeoff in
STANDARDIZED SCORES (e.g., T SCORES, Z research, the balance strack between
SCORES), and staras as an OUTLIER. measuring dependent variables accu-
Graphical approaches such as BOX-AND- rately and specifically and obtaining re-
WHISKER PLOTS and FREQUENCY POLY- sults with applicability to contexts
GONS alSO can provide valuable informa- beyond the original study. The rele-
vance-sensitivity tradeoff should be
tion about relative standing.
considered by a researcher when design-
relative survival a value indicating ing a study.
the number of people with a specific dis-
ease who are alive compared to the num- r e l i a b i l i t y n. the trustworthiness or
ber of surviving people who are disease consistency of a measure, that is, the de-
free but otherwise have the same charac- gree to which a test or other measure-
teristics (age, sex, etc.). Relative survival ment instrument is free of RANDOM
often is calculated for several different ERROR, yielding the same results across
time points, such as one year since a sig- multiple applications to the same sam-
nificant milestone (e.g., diagnosis, treat- ple. See ALTERNATE-FORMS RELIABILITY;
ment, remission), five years since the INTERNAL CONSISTENCY; RELIABILITY
milestone, and 10 years since the mile- COEFFICIENT; RETEST RELIABILITY.
stone. reliability coefficient (symbol: r^x)
relative v a r i a b i l i t y an assessment of an index describing the consistency of
the DISPERSION within a sample, typi- scores across contexts (e.g., different
cally expressed as a COEFFICIENT OF times, items, or raters). Its value, ranging
VARIATION or a DIVERSITY INDEX. from 0 to 1, provides an estimate of the
amount of obtained score variance that
relativism n. any position that chal- is due to TRUE VARIANCE rather than to
lenges the reaUty of absolute standards error. The larger this coefficient, the
of truth or value. In EPISTEMOLOGY, rel- more confident a researcher may be that
ativism is the assertion that there are no scores obtained at different times under
absolute grounds for truth or knowledge similar conditions with the same partici-
claims. Thus, what is considered trae de- pants wUl be alike. Typically, reUabUity
pends solely on individual judgments coefficients are considered to be accept-
and local conditions of culture, reflect- able if they are above .80. Also called co-
ing individual and coUective experience. efficient of reliability; index of
Such relativism challenges the validity reliability; reliability index.
of science except as a catalog of experi-
ence and a basis for ad hoc empirical pre- r e l i a b i l i t y diagonal the MAIN DIAGO-

307
reliability index

NAL in a CORRELATION MATRIX where ORY, GENERALIZABILITY THEORY, and

the values of 1 that are typically given ITEM RESPONSE THEORY. Each considcts
along this diagonal are replaced by RELI- error differently and thus it is important
ABILITY COEFFICIENTS indicating the for a researcher to specify which concep-
consistency of each measure used. This tualization underlies his or her study.
allows a researcher to evaluate the reli-
ability of each measure quickly and eas- remitted-disorder study an empiri-
ily. cal investigation of patients with a disor-
der whose condition has improved.
r e l i a b i l i t y index see RELIABILITY CO- Because the trajectories of certain disor-
EFFICIENT. ders (e.g., schizophrenia, bipolar disor-
der, chronic pain, cancer) include
r e l i a b i l i t y o f components the aver-
periods of relapse and abatement, re-
age RELIABILITY of scoies Obtained from
searchers seek to identify the conditions
the component subsets that make up a
under which each occurs so as to predict
test. Components are identified by
these transitions in severity more accu-
grouping a large set of items into clusters
rately and develop better treatments.
with similar content through rational
or intuitive approaches or through the R E M L abbreviation for RESTRICTED M A X -
use of an empirical method (e.g., FAC- I M U M LIKELIHOOD.
TOR ANALYSIS, PRINCIPAL COMPONENTS
ANALYSIS). The INTERNAL CONSISTENCY remote cause see ULTIMATE CAUSE.
of each of the component sets is deter- remote effect any outcome that results
mined and then used to estimate the indirectly from a cause, such as some
overall reliability of the larger test. For manipulation in a system or experi-
example, a researcher might factor ana- ment. For example, a remote effect of a
lyze a 20-item social support question- head trauma might be a reduced sense of
naire, identify three content subscales, taste. See also ULTIMATE CAUSE. Com-
calculate the internal consistency of pare LOCAL EFFECT.
scores for those three subscales, and then
use that information to determine the removing harmful consequences
test's overall reliability. The concept is a an ethical principle requiring research-
form of COMPOSITE RELIABILITY specific ers to ensure that participants in DECEP-
to testing contexts, although the two TION RESEARCH or other potentially
terms nonetheless are often used inter- detrimental practices leave a study in
changeably. See SPEARMAN-BROWN PRO- the same emotional state as when they
PHECY FORMULA. arrived. Investigators thus are obligated
to alleviate any feelings of alienation, re-
r e l i a b i l i t y o f composites see COM- sentment, negativity, and so forth by
POSITE RELIABILITY. minimizing srady risks before the study
r e l i a b i l i t y s a m p l i n g a form of ACCEP- begins and providing an in-depth DE-
TANCE SAMPLING in which samples of, BRIEFING after the srady is complete. For
say, a consumer product are inspected example, at the conclusion of a study in
to determine their acceptabihty against which participants were induced into
quality specifications at some future negative moods, the experimenters
date. This is often a test of the product's would need to take steps to induce a pos-
future life. itive mood in participants and explain
the reasons for the methods used in the
r e l i a b i l i t y theory any pf various con- study. See also FREEDOM FROM HARM.
ceptualizations about why scores on a
test or performance task are consistent reparameterization n. the process of
across contexts. Three prominent reli- redefining the PARAMETERS necessary
ability theories are CLASSICAL TEST THE- for the complete specificafion of a model,

308
 representational validity

usually for the purpose of removing recorded for an individual research

technical difficulties in an analytic solu- participant. For example, in a study ex-
tion that stem from the original para- amining political attitudes a researcher
meterization. For example, in STRUC- may collect approval ratings for a candi-
TURAL EQUATION MODELING 3 re- date at four time points before the elec-
searcher may decide to add another tion. Thus, the variable of candidate
pathway between a MANIFEST VARIABLE approval would be a repeated factor hav-
and a LATENT VARIABLE because of a new ing four levels corresponding to the four
insight gained from theoty or an empiri- different points at which ratings were
cal result. obtained. See also REPEATED MEASURES
DATA; WITHIN-SUBJECTS DESIGN.
repeatability n. the degree to which
specific research sradies obtain similar repeated measures analysis of
results when they are conducted again. variance see WITHIN-SUBJECTS ANALY-
Study or measurement conditions (e.g., SIS OF VARIANCE.
instructions, assessments, setting) must repeated measures data in a WITHIN-
be identical on both occasions. See REP- SUBJECTS DESIGN, scores generated from
LICATION; REPRODUCIBILITY.
observing a sample of individuals multi-
repeated contrast in a WITHIN- ple times (e.g., across experimental tri-
SUBJECTS ANALYSIS OF VARIANCE, a com- als, therapy sessions, grades in school)
parison of means conducted across dif- on a given outcome measure. These mul-
ferent levels of the INDEPENDENT tiple assessments can be analyzed using
VARIABLE. For example, assume a re- REPEATED CONTRASTS as part of ap-
searcher is interested in how commit- proaches such as WITHIN-SUBJECTS ANAL-
ment to therapy changes over the course YSIS OF VARIANCE and LATENT GROWTH
of the process. He or she could assess CURVE ANALYSIS.
participant commitment at the start of repeated measures design see
therapy (Month 1) and at the beginning WITHIN-SUBJECTS DESIGN.
of each of the four months thereafter,
and then evaluate how the mean com- repeated measures t test see DE-
mitment scores at Months 2, 3, 4, and 5 PENDENT-SAMPLES T TEST.
differ from the mean baseline commit- replacement s a m p l i n g see SAMPLING
ment score obtained at Month 1. There WITH REPLACEMENT.
are several types of repeated contrasts
available, such as SIMPLE COMPARISONS replication n. the repetition of an origi-
and POLYNOMIAL CONTRASTS. nal experiment or research study to ver-
ity or bolster confidence in its results. In
repeated event an episode or occur- exact replication (or literal replication),
rence that happens at multiple points in a researcher uses procedures that are
time. For example, a study designed to identical to the original experiment
prevent falls in older adults might docu- or duplicated as closely as possible.
ment each time participants experience In modified replication, a researcher in-
a fall as well as all of the features of the corporates alternative procedures and
fall (e.g., surface conditions, location, additional conditions. In conceptual
time of day), SURVIVAL ANALYSIS may be replication, a researcher introduces dif-
used to examine such recurrent data and ferent techniques and manipulations to
model the episode as a function of any gain theoretical information. See also
number of independent variables (e.g., BALANCED REPLICATION.
age, health staras, treatment condition).
representational v a l i d i t y the extent
repeated f a c t o r an INDEPENDENT to which a SIMULATION accurately repli-
VARIABLE for which multiple scores are cates the real-world situation that it is

309
representative conclusion

intended to represent (external repre- characteristics and constitution in cor-

sentational validity) and functions in rect proportions and allowing for
the intended manner (intemal repre- GENERALIZABILITY.
sentational validity).
representative s a m p l i n g the selec-
representative conclusion a single tion of srady units (e.g., participants,
finding from an empirical study that homes, schools) from a larger group
represents the entire set of results ob- (population) in an unbiased way, such
tained. For example, researchers might that the sample obtained accurately re-
summarize a set of findings from a flects the total population. For example,
multiyear study of women in midlife a researcher conducting a study of uni-
with the statement, "Women who expe- versity admissions would need to ensure
rience high role conflict in their work he or she used a representative random
change jobs more often and show lower sample of schoolsin other words, each
resilience." In many research settings it school would have an equal probability
is not possible to derive a representative of being chosen for inclusion, and the
conclusion because there is some uncer- group as a whole would provide an ap-
tainty in the findings or the number of propriate mix of different school charac-
findings is too small. teristics (e.g., private and public, student
body size, cost, proportion of students
representative design an experimen- admitted, geographic location).
tal design that includes processes and
variables that might be found outside reproduced correlation in FACTOR
the laboratory setting. In representative ANALYSIS, the correlation between the
designs, the participants, situations, LATENT VARIABLES Or factors that have
constructs, and assessments are sampled been extracted and the MANIFEST VARI-
in a way that permits generalization be- ABLES. Reproduced correlations are cal-
yond the specific research setting, and culated for all extracted factors and then
background variables are intentionally displayed in a tabular format called a re-
not controlled so that research results produced correlation matrix. This is
will apply more realistically to the real compared to the original CORRELATION
world. See ECOLOGICAL VALIDITY. MATRIX to determine whether their val-
ues are similar: If so, the proposed fac-
representative measure an instru- tors are a good fit for the obtained data.
ment, assessment, or procedure that ad-
equately reflects a broader group of r e p r o d u c i b i l i t y n. the extent to which
available options, having characteristics a study produces the same findings
or features that enable it to serve as an when it is conducted again by a different
appropriate indicator for the larger set. independent researcher. A given re-
For example, a representative measure of search finding is thought to be stronger
general job satisfaction would be appli- when it can be both repeated and repro-
cable to multiple types of jobs and set- duced. See REPEATABILITY; REPLICA-
tings and systematically cover a fuU TION.
range of associated topics, such as satis-
faction with pay, with coworkers, with resampling n. an analytic method in
the work itself, with supervision, with which a researcher repeatedly chooses
company policies, and so forth. subgroups of observations from a larger
overall data set in order to estimate vari-
representativeness n. the correspon- ous characteristics of that larger set. The
dence between a sample and the larger smaller subsets of observations may be
population from which it is drawn such drawn through random SAMPLING WITH
that the sample accurately reflects its REPLACEMENT Or wlthout replacement,
population, reproducing the essential with either strategy having implications
 research program

for the different analytic approaches r e s e a r c h e r b i a s any unintended errors

that may be used.
r e s e a r c h n. the systematic effort to dis-
cover or confirm facts, to investigate a
new problem or topic, or to describe
events and understand relationships
among variables, most often by scien-
tific methods of observation and experi-
mentation. Research is essential to
science in contributing to the accumula- r e s e a r c h e t h i c s the values, principles,
tion of generalizable knowledge.
r e s e a r c h d e s i g n a strategic plan of the their professional lives, including the
procedures to be followed during a study
in order to reach valid conclusions, with
particular consideration given to partici-
pant selection and assignment to condi-
tions, data collection, and data analysis.
Research designs may take a variety
of forms, including not only experi-
ments but also quasi-experiments (see r e s e a r c h h y p o t h e s i s a statement de-
QUASI-EXPERIMENTAL RESEARCH), OB-
SERVATIONAL STUDIES, LONGITUDINAL
DESIGNS, surveys, focus groups, and
other nonexperimental methods. See
also EXPERIMENTAL DESIGN.
r e s e a r c h d i a r y an investigator's docu-
mentation of the activities undertaken
during a study, including its overall de-
participants' income will increase with
sign and conceptualization, sampling
and measurement procedures, data col-
lection and analysis, and reporting of
findings as well as reflections, notes, and
observations of a more personal nature.
Research diaries may be reviewed to un-
derstand the nuances of a project and r e s e a r c h m e t h o d a procedure for the
often provide ideas that form the basis of
future studies.
r e s e a r c h e r n. the investigator who is
conducting a study or experiment. This
person may be working on the project
independently or be part of a larger col-
laborative team. A researcher's involve-
ment in the study may range from
designing the research procedure to col-
lecting and analyzing data or reporting r e s e a r c h p r o g r a m a set of planned, in-
findings. In the context of a traditional
experimental design, a researcher often
is referred to more specifically as an EX-
PERIMENTER.
in the research process or the interpreta-
tion of its results that are attributable to
an investigator's expectancies or precon-
ceived beliefs. The term essentially is
synonymous with EXPERIMENTER BIAS
but is applied to all types of investigative
projects rather than to experimental de-
signs only.
and standards that guide the conduct of
individual researchers in several areas of
design and implementation of studies
and the reporting of findings. For exam-
ple, research ethics stipulate that studies
involving data collection from human
participants must be evaluated by INSTI-
TUTIONAL REVIEW BOARDS.
scribing the investigator's expectation
about the pattern of data that may result
from a given study.' By stating specific
expectations before the data are col-
lected, the investigator makes a commit-
ment about the direction (e.g.. Method
A will yield higher final exam scores
than Method B) and magnirade (e.g.,
more education) of potential relation-
ships based on the shidy's theoretical
framework and related prior studies. See
also ALTERNATIVE HYPOTHESIS; NULL
HYPOTHESIS.
formulation and evaluation of hypothe-
ses that is intended to reveal relation-
ships between variables and provide
an understanding of the phenomenon
under investigation. In psychology, this
generally involves empirical testing and
takes the form of the SCIENTIFIC M E T H O D .
See also Q U A L I T A T I V E R E S E A R C H ; Q U A N -
TITATIVE R E S E A R C H .
terrelated empirical sradies that an in-
vestigator conducts, usually on a general
topic of interest. For example, a re-
searcher interested, in the effectiveness

311
research protocol

of elementary-school teachers may residual n. in REGRESSION ANALYSIS, the

srady several different classrooms using difference between the value of an em-
a different method in each, such as in- pirical observation and the value of that
school visits, online surveys, observa- observation as predicted by a model.
tions of students, and interviews with Analysis of residuals aUows a researcher
teachers, parents, and peers. to judge the fit or appropriateness of the
research protocol the complete de- model for the data.
scription of one's outline or plan for residual analysis a diagnostic review
conducting a srady. It should be as de- of the discrepancies between specific ob-
tailed as possible, including such ele- servations and their respective values
ments as the RESEARCH HYPOTHESIS to predicted by a model (see RESIDUAL).
be addressed and the rationale for doing Summarizing these discrepancies in sev-
so; the materials and resources that will eral different ways can help a researcher
be required; the timeline or duration; identify problems in the application of a
the precise sampling, measurement, and model to a particular data set. For exam-
analysis procedures that will be used; a ple, a residual analysis might show large
discussion of any ethical considerations; discrepancies for one group but not an-
a description of strengths and limita- other, suggesting the. model is not ap-
tions; and so forth. propriate for that specific group. See
research r i s k the potential costs to par- REGRESSION DIAGNOSTICS.
ticipants or to society at large associated
with a particular study, which must be residual degrees o f freedom the
clearly specified to an INSTITUTIONAL total number of cases in a REGRESSION
REVIEW BOARD by an investigator prior ANALYSIS minus the number of groups
to conducting the research. Examples of or conditions into which they have been
potential risk to participants include placed (i.e., the number of PARAMETERS
embarrassment at being involved in DE- being estimated). It is equivalent to the
CEPTION RESEARCH, Side effects from a WITHIN-GROUPS DEGREES OF FREEDOM
treatment, time and resource commit- in an ANALYSIS OF VARIANCE.
ments, and emotional upset (e.g., from residual error see ERROR VARIANCE.
exposure to unpleasant or painful mem-
ories). An investigator must discuss how residiial m a t r i x a MATRIX displaying
the costs will be minimized to the fullest the discrepancies between observed data
extent possible and how the study bene- and the corresponding values predicted
fits outweigh the costs. by a researcher-specified model. For ex-
research strategy see RESEARCH DE- ample, in FACTOR ANALYSIS 3 residual
matrix reflects what remains after fac-
SIGN.
tors or dimensions are removed from
research synthesis the systernatic use a CORRELATION MATRIX or COVARIANCE
of established data-gathering methods MATRIX. Large values in the matrix show
and statistical approaches to evaluate a where the model did not flt the data well
body of empirical literature on a topic. and can provide clues about its MIS-
An investigator must be careful to in- SPECIFICATION. Researchers often prefer
clude both published and unpublished to standardize the values in a residual
sradies, to document the methods that matrix so that they are directly interpret-
each study uses, to provide detail about able, with values greater than 1.96 con-
the findings and obtained EFFECT SIZES, sidered statistically significant at the p <
and to summarize commonalities and .05 level (see PROBABILITY LEVEL) and
account for differences across studies. values greater than 2.58 considered sta-
See META-ANALYSIS. tistically significant at the p < .01 level.

312
 response acquiescence

Variable 1 2 3 4 5 6 7 8 9 10 II 12 13 14 15 16
1. JobSall 0.20
2. JobSal2 1.04 1.89
3. JobSat3 -0.49 0.87 0.33
4. HlConl -0.09 0.56 0.09 0.02
5. HlCon2 -1.81 -0.04 -0.35 6.39 0.01
6. HlConS 0.41 0.54 0.81 -3.00 -2.98 0.02
7. Psych] -1.25 -0.69 -0.51 -8.55 -4,73 -SJ< 0.70
8. Psycii2 -2.30 -1.41 -1.21 -8.46 -4.57 -9.05 0.67 0.35
9. Psych3 -1.38 -0.60 -0.56 -7.93 -^.77 -8.79 0.86 0.34 0.43
10. HelSall 1.50 1.49 1.92 -1.45 -1.79 1.88 -8.19 -8.00 -8.29 0.04
11. HclSat2 0.05 0.22 0.26 -2.82 -0.55 2.10 -6.47 -6.53 -5.93 1.79 0.02
12. HCISBL; 1.72 0.29 0.54 0.51 -2.16 3.77 - 9 J 6 -9.37 -8.62 -0.86 -1.02 0.02
13. JwithI -3.04 -1-31 -3.58 1.44 1.25 1.40 0.45 2.12 0.75 1.01 0.15 -0 27 0.00
14. Jwilh2 -5.54 -5.67 -5.49 -0.16 0.09 -0.20 4.43 4.78 4.54 -1.05 -0 65 -1.00 -0.02 0.00
15. Jwilh3 -5.74 -6.01 -5.66 1.69 0.50 1.09 299 3.96 2.61 0.63 -1.16 -0.44 1.14 -0.88 0.00
16. Wwilhl -O30 1.34 0 02 -3.36 -2.81 -3.93 5.99 6.19 6.15 -5.63 -4.26 -4.80 2.30 1.96 2.33 0.00
17. Wwith2 -2.60 -1 68 -2.20 -3.37 -2.10 -4.71 7.52 7.00 6.70 -4.75 -3.99 -1.51 2.98 4.41 3.51 0.00 0 00
18. Wwilh3 -0.30 1.81 1.23 -4 15 -3.21 -4.37 6.12 6.72 5.99 -5.50 -4.29 -5.15 2.03 2.26 1.76 0.00 -0.40

residual matrix

as indicated by boldface in the example tion that is less likely to be influenced

above. by the presence of OUTLIERS (extreme
JobSatl, JobSat2, and JobSat3 are the scores) in the sample upon which it is
three indicators for the latent constract based. See also ROBUST ESTIMATOR.
job satisfaction. Each of the subsequent
respondent n. a study participant who
sets of three variables are the three indi-
is interviewed as part of a research de-
cators for the latent constructs health
sign or who completes a survey or ques-
conditions, psychological conditions,
tionnaire.
health satisfaction, job withdrawal, and
work withdrawal, respectively. Also respondent v a l i d a t i o n see MEMBER
called residual correlation matrix. CHECK.
residual m a x i m u m l i k e l i h o o d see responders-versus-nonresponders
RESTRICTED M A X I M U M LIKELIHOOD. analysis an empirical review of re-
residual mean square see MEAN search data to understand potential dif-
SQUARED ERROR.
ferences between those who choose to
participate in a survey and those who
residual p a t h coefficient in PATH choose not to participate. For example, a
ANALYSIS, a value reflecting the amount responders-versus-nonresponders analy-
of error in an equation linking measured sis might show that people who com-
variables to an outcome of interest. In plete a questionnaire about mood
other words, the residual path coeffi- exhibit fewer symptoms of depression
cient indicates the degree of variance in than those who opt not to do so. See AT-
a DEPENDENT VARIABLE that is not ac- TRITION. See also MISSING VALUES PRO-
counted for by the INDEPENDENT VARI- CEDURE.
ABLES included in the model.
response n. any glandular, muscular,
residual score see ERROR SCORE.
neural, or other reaction to a stimulus. A
residual sum o f squares see ERROR response is a clearly defined, measurable
SUM OF SQUARES. unit of behavior discussed in terms of its
result (e.g., pressing a lever, indicating
residual t e r m see ERROR TERM. yes VS. no on a survey item) or its physical
residual variance see ERROR VARI- characteristics (e.g., raising an arm, shar-
ANCE. ing a toy).
resistant estimator an ESTIMATOR for response acquiescence see YEA-
an unknown characteristic in a popula- SAYING.

313
response bias

r e s p o n s e b i a s the tendency for a study the SOCIAL DESIRABILITY response set.

participant to give one answer ortypeof See also RESPONSE BIAS; RESPONSE STYLE.
answer more than others, regardless of r e s p o n s e s t y l e a RESPONSE SET arising
the stimulus condition. There are several from dispositional factors that appear
different types of response bias, includ- across contexts and over time rather
ing the HALO EFFECT, NAY-SAYING, and than from situational factors.
Y E A - S A Y I N G . See a l s o R E S P O N S E SET; R E -
SPONSE S T Y L E . response surface methodology
(RSM) a set of procedures used to model
r e s p o n s e d e v i a t i o n 1. the difference and analyze the relationships between
between an individual's score and the one or more outcome or DEPENDENT
average value for the total set of scores. VARIABLES and multiple predictor or IN-
2. see NAY-SAYING. D E P E N D E N T VARIABLES posited in a par-
ticular REGRESSION EQUATION. The
r e s p o n s e e f f e c t the influence of some
analysis involves graphing a function in
attribute of the M E A S U R E M E N T S C A L E or
three-dimensional space (i.e., creating a
administration context on a partici-
response surface), examining the con-
pant's answers to survey or interview
tours of the resulting plot, and using the
items. For example, the order in which
information so obtained to optimize
response options are presented may af-
and refine the model by adding or delet-
fect how a participant will answer, as
ing predictors, identitying OUTLIERS,
might the inclusion of a middle or neu-
and considering NONLINEAR terms. Con-
tral point on an agreement scale or
sider the following example.
whether a survey is conducted in person,
via the Internet, or over the telephone.

r e s p o n s e - f e a t u r e a n a l y s i s any of
several methods of evaluating L O N G I T U -
DINAL DATA that involve calculating the
same S U M M A R Y STATISTICS, such as the
MEAN, STANDARD DEVIATION, SLOPE,
and AREA UNDER THE CURVE, for a data
set at different time points.
r e s p o n s e l a t e n c y see R E A C T I O N T I M E .

r e s p o n s e r a t e the number of individu-

als who complete an interview, answer a
survey, or join a research srady com-
This response surface depicts team av-
pared to the number who were con-
erage perception of climate for organiza-
tacted to participate, often expressed as a
tional support, manager perception of
P E R C E N T A G E . Compare REFUSAL RATE.
team climate, and levels of team positive
r e s p o n s e scale any of various types of affect for employees in different branches
instrument provided to a respondent to of three savings banks in the same geo-
express an answer to an item. Examples graphical region. Also called response
of different response scales include surface analysis.
F I X E D - A L T E R N A T I V E Q U E S T I O N S , LIKERT
r e s p o n s e t i m e see R E A C T I O N T I M E .
SCALES, VISUAL ANALOGUE SCALES, and
SEMANTIC DIFFERENTIALS. response variable see DEPENDENT
VARIABLE.
r e s p o n s e set a tendency to answer ques-
tions in a systematic manner that is un- restricted m a x i m u m likelihood
related to their content. An example is (REML) a technique for estimating the

314
 retrospective sampling

PARAMETERS of a distribution, often used exposure to secondhand tobacco smoke

in HIERARCHICAL LINEAR MODELS, might use existing survey data to divide
GENERALIZABILITY THEORY, and VARI- a sample into groups or cohorts based on
ANCE COMPONENTS ANALYSIS. It iS a reported exposure in their childhood
form of MAXIMUM LIKELIHOOD that uses homes (no smoking, one person
less information than other related ap- smoked, more than one person smoked)
proaches (i.e., it does not use all of the before examining current reported
available observations in a vety large data health problems. Because the research
set) to produce coefficients vfith desirable relies on reports of past occurrences, in-
characterisfics (e.g., UNBIASED esfimates). accuracies may arise due to poor recall.
Also called residual maximum like- See also RETROSPECTIVE RESEARCH.
lihood; reduced maximum likeli- retrospective power an assessment of
hood. the probability that a study will detect
restriction o f range the limitation a statistically significant effect that is
via sampling, measurement procedures, conducted after the study has been com-
or other aspects of experimental de- pleted. Researchers sometimes use this
signof the.full range of total possible approach to provide evidence of
scores that may be obtained to only a whether a study had a sufficient likeli-
narrow portion of that total. For exam- hood of detecting an effect at a certain p
ple, in a study of the grade point aver- value (see PROBABILITY LEVEL), given the
ages of university students, restriction of sample size used and the EFFECT SIZE ob-
range occurs if only sradents from the tained. Generally, the more appropriate
dean's list are included. Range restric- strategy is to conduct a POWER ANALYSIS
tion on a particular variable may lead to prior to coUecting study data so as to de-
such negative effects as failing to ob- termine if the planned research design is
serve or improperly characterizing a re- sufficient to produce significant results
lationship between the variables of at a certain statistical level. Also called
interest. post hoc power.

retest 1. n. the administration of a test, retrospective research observa-

assessment instrument, or other mea- tional, nonexperimental research that
surement procedure to the same partici- tries to explain the present in terms of
pants at a point in time subsequent to past events; that is, research that starts
the original administration (e.g., two with the present and follows partici-
weeks later, one year later). 2. vb. to pants backward in time. For example, an
readminister such a test, assessment in- investigator may select a group of indi-
strument, or measurement procedure. viduals who exhibit a particular prob-
lematic symptom and then srady them
retest reliability a measure of the con- to determine if they had been exposed to
sistency of results on a test or other as- a risk factor of interest. Also called ret-
sessment instrument over time, given as rospective study. Compare PROSPEC-
the correlation of scores between the TIVE RESEARCH.
first and second administrations. It pro-
vides an estimate of the stability of the retrospective s a m p l i n g a technique
construct being evaluated. Also called in which participants or cases from the
test-retest reliability. general population are selected for in-
clusion in experiinents or other research
retrospective cohort study research based on their previous exposure to a
that compares outcomes for groups of risk factor or the completion of some
individuals who differ on a single identi- particular process. Participants are then
fied characteristic that occurred in the examined in the present to see if a par-
past. For example, a researcher studying ticular condition or state exists, often in

315
retrospective validity

comparison to others who were not ex- LEAST SQUARES REGRESSION designed to
posed to the risk or did not complete the remedy problems that arise from
particular process. Compare PROSPEC- MULTICOLLINEARITY. It involves modi-
TIVE SAMPLING. fying the MAIN DIAGONAL Of the CORRE-
LATION MATRIX before calculating
retrospective v a l i d i t y the extent to
the coefficients, thus eliminating the as-
which an instrament that purports to
sociations among the INDEPENDENT
measure a particular behavior or phe-
VARIABLES. A researcher can use the in-
nomenon of interest can be shown to
formation obtained from this analysis to
correlate with past behaviors or occur-
determine whether certain independent
rences that demonstrate this behavior.
variables should be removed from the
For example, a researcher evaluating
final model. Also called damped re-
a new measure of accident proneness
gression.
might administer it to a sample of
respondents and then check the individ-
r i g h t censoring inability of a re-
uals' archived medical records to deter-
searcher to document when all partici-
mine if higher test scores correlate with
pants have reached a target event (e.g.,
the number of actual treated incidents.
achieving a specific milestone, experi-
It is one of several types of CRITERION
encing a relapse) at the conclusion of the
VALIDITY. Also called postdictive va-
study period. This may occur for one of
lidity. See also CONCURRENT VALIDITY;
three reasons: A participant may never
PREDICTIVE VALIDITY. experience the target event or experi-
reversal design an experimental de- ence it after the observation period has
sign, generally used when only a single ended; a participant may experience a
group is being sradied, that attempts competing event that prevents him or
to counteract the confounding effects her from experiencing the target event;
(see CONFOUND) of sequence, order, and or a participant may be lost to ATTRI-
treatment by alternating baseline condi- TION. Compare LEFT CENSORING.
tions with treatment conditions. Exam-
ples include the A-B-A DESIGN, A-B-A-B r i g i d r o t a t i o n see ORTHOGONAL ROTA-
DESIGN, and other similar combina- TION.
tions. See also ALTERNATING TREAT-
MENTS DESIGN. risk level the probability of making a
TYPE I ERROR that One is willing to accept
reverse counterbalancing a specific in null hypothesis SIGNIFICANCE TEST-
procedure for ordering stimulus materi- ING.
als in a research study that involves ad-
ministering' one order (A-B-C) for one risk ratio, the comparison of one
half of the participants and the opposite group's probability of experiencing an
order (C-B-A) for the other half of the event (e.g., being diagnosed with lung
participants. Reverse counterbalancing cancer) to a second group's probability
is used to minimize any potential influ- of experiencing that event. It is often
ence of presentation upon results, so as used to describe health status following
to ensure it is the stimuli themselves exposure to some stimulus (e.g., lead in
that are producing any effect that may water) or clinical intervention. A value
be seen. greater than 1 indicates the group under
study has a higher probability than the
r h o c o r r e l a t i o n (symbol: p) 1. see POP- control group of experiencing the event;
ULATION CORRELATION COEFFICIENT. 2. a value less than 1 indicates the group
see SPEARMAN CORRELATION COEFFI- under srady has a lower probability of
CIENT. experiencing the event; and a value of
ridge regression a variant on ordinaty exactly 1 indicates the two groups are

316
 rootogram

equally Ukely to experience the event. procedure to produce valid estimated

Also called relative risk. values for a population characteristic
(PARAMETER) despite violations of the
risk set in SURVIVAL ANALYSIS, the num-
ASSUMPTIONS upon which the tech-
ber of individuals or cases who are at risk
nique is based.
of experiencing a specific event (e.g.,
first drink, birth of a child, promotion) ROC curve abbreviation for RECEIVER-
during a particular time period. OPERATING CHARACTERISTIC CURVE.

R methodology a collection of various role-playing research a design in

methods used to evaluate individuals' which participants are asked to assume a
scores on a set of objective measures, specific character in a defined situation
such as inteUigence tests. Examples in- and to behave as if that character were
clude FACTOR ANALYSIS and PRINCIPAL really theirs and the situation were ac-
COMPONENTS ANALYSIS. Compare Q tual. It is a type of SIMULATION, often
METHODOLOGY. used in therapeutic and organizational
contexts, that enables researchers to as-
RMS abbreviation for ROOT MEAN sess how individuals think, feel, and act
SQUARE. under certain circumstances.
RMSE abbreviation for ROOT-MEAN- r o l l i n g average see MOVING AVERAGE.
SQUARE E R R O R .
root mean square (RMS) the square
RMSR abbreviation for ROOT-MEAN- root of the sum of the squares of a set of
SQUARE RESIDUAL. values divided by the number of values.
RNG abbreviation for RANDOM NUMBER For a set of values x^, X2, ... x, the root
GENERATOR. mean square value is
Robinson matrix in TIME-SERIES
ANALYSIS, a SYMMETRICAL MATRIX con- and describes the average size of the val-
taining CORRELATIONS of eariy time ues in the set. In the physical sciences
points with later time points. Correla- the term is used as a synonym for STAN-
tions are arranged such that those clos- DARD DEVIATION under certain circum-
est to the MAIN DIAGONAL are the largest stances.
and the ones furthest away are the small-
est. [W. S. Robinson] root-mean-square deviation the
square root of a MEAN-SQUARE DEVIA-
robust estimator an ESTIMATOR for an TION.
unknown characteristic in a population
that is less likely to be influenced by VIO- root-mean-square error (RMSE) the
LATIONS OF ASSUMPTIONS about the square root of a MEAN SQUARED ERROR.
sample data upon which it is based. For root-mean-square residual (RMSR)
example, rather than using a mean to the square root of a mean-square resid-
describe CENTRAL TENDENCY in a distri- ual (see MEAN SQUARED ERROR). A stan-
bution with OUTLIERS, a researcher dardized version of this value commonly
might choose a MEDIAN, a TRIMMED is used to assess model fit in STRUC-
MEAN, or a WINSORIZED MEAN. Each op- TURAL EQUATION MODELING.
tion reflects a different way of handling
the outliers, such as downweighting rootogram n. a HISTOGRAM modified to
them, replacing them with a new value, show the response or outcome variable
or ignoring them. Also called sturdy on the horizontal ;t-axis and the square
statistic. See also RESISTANT ESTIMA- root of the response frequencies (or
TOR. counts) on the vertical y-axis, as in the
hypothetical example overleaf.
robustness n. the ability of a statistical Alternatively, the square root of RELA-

317
Rosenbaum's test of unidimensionality

160

o o o o o o o o o o o o o o o o o o o o o
o o o o o o o o o o o o o o o o o o o o
o o o o o o o o o o o o o o o o o o o o
T-" CM'm in CD r^' cd a i o T-" CN in co r^' co'

Debt (dollars)

rootogram

TIVE FREQUENCIES, CUMULATIVE FRE- ample, it may be used to maximize load-

QUENCIES, or CUMULATIVE RELATIVE ings on certain factors while minimizing
FREQUENCIES may be given along the y- these loadings on other factors, thereby
axis. See also H A N G I N G R O O T O G R A M . showing SIMPLE STRUCTURE. See
OBLIQUE ROTATION; ORTHOGONAL RO-
R o s e n b a u m ' s test o f u n i d i m e n -
TATION; PROCRUSTES ROTATION; QUAR-
s i o n a l i t y a statistical procedure to eval-
TIMAX ROTATION; VARIMAX ROTATION.
uate whether the associations among a
set of items or variables are generally de- rotational adj.
scribed by a single dominant factor r o t a t i o n s a m p l i n g a technique used
rather than by several factors (i.e., are when conducting surveys in which after
unidimensional instead of multidimen- specified time periods some proportion
sional). It is a N O N P A R A M E T R I C method of respondents or cases is replaced witti
that assesses conditional associations new units. Rotation sampling helps to
within C O N T I N G E N C Y TABLES. Also reduce the burden on respondents and
called H o l l a n d - R o s e n b a u m test o f allows for better prediction from past
u n i d i m e n s i o n a l i t y . [Paul W. Holland samples.
and Paul R. Rosenbaum, U.S. statisti- r o w m a r g i n a l a summaty of the values
cians] across each horizontal row of cells in a
R o s e n t h a l e f f e c t the situation i n table. For a table containing frequency
which an investigator's expectations counts it is a sum of the number of
about the outcome of a given study un- counts in each row, whereas for a table
wittingly affect the acraal study out- of averages it is the mean value of data
come. A researcher may use BLINDS to observations across each row. Compare
prevent the Rosenthal effect from occur- COLUMN MARGINAL.
ring and biasing study results. This term r o w s u m o f s q u a r e s 1. in an ANALYSIS
is often used synonymously with EXPER- OF VARIANCE for two independent vari-
IMENTER EXPECTANCY EFFECT and SELF- ables a and b, the amount of V A R I A N C E
FULFILLING PROPHECY. See also D E M A N D that is associated with either a or b, as
CHARACTERISTICS. [Robert Rosenthal derived from the values given in the rele-
(1933- ), U.S. psychologist] vant row of the data table. It is obtained
r o t a t i o n n. in statistics, movement by determining the average of all obser-
around the origin in a multidimensional vations i n the row, calculating how
space. Rotation is commonly used in much each score deviates from that av-
FACTOR ANALYSIS to enhance inter- erage, multiplying the resulting value
pretabUity of a factor solution. For ex- by itself, and adding-it to the similarly

318
 run-in period

obtained values for all other cases in r t o z t r a n s f o r m a t i o n see FISHER'S R

the row. The calculated quantities for TO Z TRANSFORMATION.
each row are then used to compute
the BETWEEN-GROUPS SUM OF SQUARES, R u b i n ' s c a u s a l m o d e l a model used
which in rarn is used to compute an F to estimate the magnitude of an inter-
RATIO. See also T O T A L S U M OF SQUARES. vention's effect relative to a comparison
Compare C O L U M N S U M OF SQUARES. 2. condition. Used primarily in economics,
in E X P L O R A T O R Y F A C T O R ANALYSIS, the medicine, and public health research,
variance accounted for by each variable the model unrealisticaUy assumes that
across factors (see C O M M O N A L I T Y A N A L - participants experience multiple condi-
YSIS). It is determined by raising the FAC- tions simultaneously (e.g., the same par-
TOR LOADINGS for each variable in a ticipant was in a treatment and control
factor loading matrix to the second condition at the same time) in order to
power and then totaling the values. envision all possible potential outcomes
of the intervention. [Donald B. Rubin
r o w v e c t o r a data matrix with a single (1943- ), U.S. statistician]
row of values. In other words, it has the
dimensions 1 x c, where 1 denotes the Rudas-Clogg-Lindsay index of f i t
single row and c refers to the number of a quantitative measure of how well a sta-
columns. Compare C O L U M N V E C T O R . tistical model corresponds to data in a
CONTINGENCY TABLE that exhibit a FI-
Roy's greatest characteristic root
NITE MIXTURE DISTRIBUTION. It rcliCS On
see GREATEST CHARACTERISTIC ROOT.
the idea that for part of the population
R o y ' s l a r g e s t - r o o t test see GREATEST- the model is true and for part it is not
CHARACTERISTIC-ROOT TEST. true. The measure, which does not de-
pend on sample size as does the C H I -
rpb symbol for P O I N T BISERIAL C O R R E L A - SQUARE GOODNESS-OF-FiT TEST, ranges
T I O N COEFFICIENT. in value from 0 to 1, with the latter indi-
rpbis symbol for P O I N T BISERIAL C O R R E L A - cating good fit. [Tamas Rudas, Hungar-
T I O N COEFFICIENT. ian mathematician; Clifford C. Clogg
(1950-1995), U.S. sociologist; Brace G.
RRT abbreviation for RANDOMIZED- Lindsay (1947- ), U.S. mathematician]
RESPONSE T E C H N I Q U E .
r u g p l o t n. a H I S T O G R A M or SCATTERPLOT
rs symbol for S P E A R M A N CORRELATION
of data for one or more groups of a CATE-
COEFFICIENT.
GORICAL VARIABLE that has been modi-
R S M abbreviation for RESPONSE SURFACE fied to include each group's distribution
METHODOLOGY. of values on another C O N T I N U O U S VARI-
ABLE, as in the generic example overleaf.
R T abbreviation for R E A C T I O N T I M E . It is SO named because the short hori-
R - t e c h n i q u e f a c t o r a n a l y s i s a type zontal lines along the x-axis showing the
of FACTOR ANALYSIS performed on a density of observations at different
C O R R E L A T I O N M A T R I X : It examines asso- points along the continuously measured
ciations between variable measurements variable resemble a rug in cross section.
in order to understand how the variables
themselves group together and are re- r u n - i n p e r i o d in an experiment or SIM-
lated. Compare P - T E C H N I Q U E FACTOR
U L A T I O N , an initial period of time that is
ANALYSIS; Q-TECHNIQUE FACTOR ANALY- included to allow the phenomenon of
SIS. interest to stabilize before any treatment
or manipulation is administered. A ran-
rjet symbol for T E T R A C H O R I C C O R R E L A - in period helps a researcher identify
TION COEFFICIENT. problems with the protocol or treatment.

319
r u n n i n g average

100-1

80-

^ 60-

.a
.2
^ 40-

20-

liiiimini |iiiiiiI iiiiiji

iiiiiiii ill I I
1
2 4 6 8 10
Variable x

rugplot

r u n n i n g average see MOVING AVER- runs test see WALD-WOLFOWITZ TEST.

Ryan's m e t h o d see REGWQ TEST.
r u n n i n g medians see MOVING MEDI- R^yy- symbol for CROSS-VALIDATED MUL-
ANS. : TIPLE CORRELATION.

320
 Ss
s symbol for SAMPLE VARIANCE. observations in this subset, such as the
mean, median, or STANDARD DEVIATION,
S abbreviation for a SUBJECT (i.e., individ- are called STATISTICS.
ual or case) in a srady. People partaking
in research are now more commonly re- sample correlation coefficient
ferred to as PARTICIPANTS. (symbol: r) an index of the degree of as-
sociation between two variables based
St. Petersburg paradox a paradox on the data in a studied subset (sample)
often cited in PROBABILITY THEORY and of cases from a larger group of interest. It
DECISION THEORY, which shows how a is a variant of the PRODUCT-MOMENT
decision criterion that only takes an EX- CORRELATION COEFFICIENT, such that
PECTED VALUE into account may result
the same symbol is used for both statis-
in a recommended course of action that tics. See also POPULATION CORRELATION
no rational person would follow. The
COEFFICIENT.
paradox involves a theoretical game
called the St. Petersburg Lottety, which sample d i s t r i b u t i o n the DISTRIBU-
is played by flipping a coin until it comes TION of scores in a particular subset
up tails, and the total number offlips,n, (sample) drawn from the wider popula-
determines the prize, which equals $2 x n. tion. Of interest is the general shape of
Thus, if the coin comes up tails the first the distribution, reflecting the fre-
time, the prize is $2 and the game ends. quency of particular scores (its KUR-
If the coin comes up heads the first time, TOSIS, SKEWNESS, etc.). Compared to the
it isflippedagain. If it comes up tails the theoretical POPULATION DISTRIBUTION,
second time, the prize is $4 and the the distribution of scores in a sample is
game ends. If it comes up heads the sec- often jagged and not smooth.
ond time, it isflippedagain, and so forth.
The expected value of the game is the sample mean (symbol: X, M) the arith-
sum of the potential payoffs; since the metic average (MEAN) of a set of scores
payoff of each possible consequence is from cases or observations in a subset
$2/2 = $ 1, and since there are potentially from a wider population. Because each
score contributes equally to this index of
an infinite number of payoffs, the sum is
CENTRAL TENDENCY, it can be affected
an infinite number of dollars. In theoty,
greatly by OUTLIERS.'Many widely used
because the amount one could win is in-
STATISTICAL TESTS are based on the com-
finite, a rational gambler should be will- parison of sample means.
ing to pay any finite amount to play (i.e.,
any price of entry is smaller than the ex- sample of convenience see CONVE-
pected value of the game). In practice, NIENCE SAMPLING.
however, this would clearly not be the
case. sample overlap the situation in which
two or more subsets drawn from a popu-
sample n. a subset of a POPULATION of lation feature items in common (i.e., the
interest that is selected for study with same individual or observation features
the aim of making inferences to the pop- in more than one sample). A researcher
ulation. It is important to ensure that a can design a selection strategy in which
sample is representative of the larger the probability of finding the same indi-
population. Characteristics that describe viduals across two sradies is maximized.

321
sample reliability

thereby reducing costs associated with QUOTA SAMPLING. Each approach has a
data collection. Alternatively, this prob- different potential of obtaining a sample
ability could be minimized, thereby en- appropriately representative of the POP-
suring independence across studies and ULATION under study.
reducing respondent burden.
s a m p l i n g bias a systematic and direc-
sample r e l i a b i l i t y the degree to which tional error involved in the choice of
a SAMPLE is representative of the POPU- units, cases, or participants from a larger
LATION from which it is drawn. It is typi- group into a research study. This is a po-
cally indexed by the STANDARD ERROR tential problem whenever the researcher
OF THE MEAN. has latitude in selecting individual units
for the sample. Selection bias can pose
sample size the number of observations a threat to the INTERNAL VALIDITY of a
(cases, individuals, units) included in study if there is a possibility that pre-
the sample to be sradied. This is usually existing differences arising from the
denoted N (for the study as a whole) or n sampling process may interact with the
(for subgroups from the study). variable of interest. Similarly, if the pro-
sample space the collection of all possi- cedure used to choose participants tends
ble outcomes of an experiment of chance. to favor specially motivated individuals
For example, for a toss of a single coin or people from a certain segment of soci-
the sample space is heads and tails, ety, there would be a threat to the srady's
whereas for a toss of two coins the sam- EXTERNAL VALIDITY (I.e., Inferences to a
ple space is heads-heads, heads-tails, larger population would not be viable).
taUs-tails, and tails-heads. Selection bias is associated with a lack
of RANDOM SAMPLING and with non-
sample standard deviation (sym- random assignment to conditions. Also
bol: 5) see STANDARD DEVIATION. called selection bias; recruitment
bias. See also SELF-SELECTION BIAS.
sample variance (symbol: 5^) the dis-
persion of scores within a group selected sampling design the specific ap-
for study, as opposed to the POPULATION proach, method, or strategy that a re-
VARIANCE. It is calculated by determin- searcher decides to use to select a sample
ing each score's difference from the aver- from the larger population. Formulating
age for the set, squaring and summing a design involves determining the na-
these differences, and then dividing by ture of the target population, a suitable
the total number of scores minus one. SAMPLING FRAME for drawing the cases,
the desired sample size, whether ran-
s a m p l i n g n. the process of selecting a
dom or nonrandom selection will be
limited number of units from a larger set
used, and whether there are any impor-
for a study. The term most often refers
tant variables on which to stratity selec-
to the selection of respondents, obser- tion. Also called sampling plan.
vations, or cases for inclusion in ex-
periments, surveys, interviews, or other s a m p l i n g d i s t r i b u t i o n the distribu-
research. However, sampling can also in- tion of a statistic, such as the mean,
volve selecting theoretical constructs to over repeated samples drawn from a
study, selecting measurement instru- population, SIMULATION studies allow
ments from a broad set of potential op- researchers to specify known population
tions, or selecting time points at which information, conduct a very large num-
to observe individuals or cases. There are ber of repeated draws on the population,
various different methods of selecting and buUd an empirical distribution of
participants for a study, including SIM- the statistic based on these draws (e.g., t,
PLE RANDOM SAMPLING, STRATIFIED F, or distributions). For example, the
SAMPLING, CONVENIENCE SAMPLING, and means calculated from samples of 100

322
 sampling variability

observations, repeatedly and randomly cords in light of the inferences that will
drawn from the population, yield a sam- be made: Some relevant questions would
pling distribution for the mean. Knowl- be whether part-time as well as full-time
edge about the distribution of a statistic students are included in the institution's
allows researchers to say when a finding files and whether students currently on
from a sample is unusual (e.g., statisti- leave are included. The key point is that
cally significant) and when it would any sampling frame should be represen-
be expected from the statistic's known tafive of the target population as a whole.
behavior, thus enabUng the sampling
s a m p l i n g i n t e r v a l in SAMPLING DE-
distribution of a statistic to be used in
SIGNS that involve the researcher select-
testing hypotheses about variables and
ing every th case from a list or table,
their relationships. See also INFERENTIAL
the value of n. The chosen interval will
TEST. usually depend on the nature of the
s a m p l i n g error the predictable MAR- SAMPLING FRAME. For example, if the re-
GIN OF ERROR that occurs in sradies that searcher has access to an alphabetized
draw samples of cases or observations list, he or she might want to ensure that
from a larger POPULATION: It indicates all parts of the alphabet are represented
the possible variance between the trae and select a value to maximize that goal.
value of a parameter in the population Alternatively, for a study conducted in
and the estimate of that value made from the field that involves the researcher ap-
the sample data. The larger the sample, proaching individuals in an office build-
the smaller the sampling error (if the ing, the value could be selected based on
entire population was sampled, there the average number of people who are
would be no error in the sample esti- likely to be available and eligible for se-
mate). Large national surveys, such as lection.
those reporting political attitudes, state s a m p l i n g p l a n see SAMPLING DESIGN.
the sampling error along with their find-
ings. For example; a survey finding that s a m p l i n g theory the body of princi-
65% prefer a particular policy with a ples underlying the drawing of SAMPLES
margin of error of 3% means that the that accurately represent the population
true figure could be anywhere between from which they are taken and to which
62% and 68%. inferences wUl be made. This conceptu-
alization of the sampling process pro-
s a m p l i n g f r a m e the specific source vides guidance to researchers about which
used in drawing a subset of cases or indi- SAMPLING DESIGN to choose in the par-
viduals from the larger POPULATION. In ticular circumstances and how best to
many cases the samplingframewill be a account for subsets of cases that are not
complete list of all the elements in a well represented (or are overrepresented)
population (e.g., the electoral register). in the population (e.g., by using WEIGHTS).
In other cases this will be impossible or
s a m p l i n g u n i t any of the elements se-
the issue may be less straightforward
than it appears. In a study of college sra- lected from a population to make up a
dents' aspirations for life after gradua- sample. For instance, if classrooms are
tion, for example, the researchers might selected at random from the population,
decide to use a listing of students then the classroom, not the individual
obtained from the institution's office student, is the sampling unit. See UNIT
of evaluation and assessment and ran- OFANALYSIS.
domly select from this list evety fifth s a m p l i n g v a r i a b i l i t y the extent to
student. In this example, however, the which the value of a statistic differs across
researcher would need to be assured of a series of samples, such that there is
the completeness of the institution's re- some degree of uncertainty involved in

323
sampling without replacement

making inferences to the larger POPULA- nitude and not direction. Compare VEC-
TION. See SAMPLING ERROR. TOR. 2. n. in MATRIX ALGEBRA, a quan-
tity that can multiply a vector in a vector
s a m p l i n g w i t h o u t replacement a
space to produce another vector. 3. adj.
SAMPLING technique that involves se-
describing a variable that can be repre-
lecting an item from the larger set and
sented by positions on a scale. 4. adj. de-
removing it from the general pool; thus,
scribing a matrix in which the entries in
this particular case cannot be redrawn.
the MAIN DIAGONAL are equal and all
Compare SAMPLING WITH REPLACEMENT.
other entries are zero.
s a m p l i n g w i t h replacement a SAM-
PLING technique in which each item se- scalar analysis the evaluation of data
lected from the larger set is returned to obtained from a test or measure using an
the general pool so that it may subse- ITEM RESPONSE THEORY model, often a
quently be redrawn. This means that a RASCH MODEL.
particular case may be drawn more than scale n. 1. a system for ordering test re-
once for a given sample. Compare SAM- sponses in a progressive series, so as to
PLING WITHOUT REPLACEMENT. measure a trait, ability, attirade, or the
saturated model a model that fits the like. For example, an agreement scale
data perfectly because it has as many es- used on an attitude survey might have
timated parameters as there are values to seven response options ranging from
be fitted. In STRUCTURAL EQUATION strongly disagree (1) to strongly agree (7),
MODELING this type of model is called a with neither disagree nor agree (4) as the
JUST-IDENTIFIED MODEL: It poses a prob- middle point. See LIKERT SCALE; RATING
lem as a researcher cannot evaluate fit, SCALE. 2. a sequence of ordered values
as would be the case with the preferred used as a reference in measuring a physi-
OVERIDENTIFIED MODEL (or oversatu- cal property (e.g., weight, temperature).
rated model), in which there are fewer See I N T E R V A L S C A L E ; R A T I O S C A L E . See
estimated parameters than DEGREES OF also MEASUREMENT LEVEL. 3. more gen-
FREEDOM. Relatedly, the underidentified erally, any test or other assessment in-
model (or unsaturated model) has more strament as a whole.
parameters than can be estimated from scale attenuation the situation in
the available data. In FACTOR ANALYSIS, which the response format on a measure
the problem of saturation occurs when includes too few options to reflect a re-
the number of factors that will be esti- spondent's actual behavior, opinion, or
mated is the same as the number of vari- belief. For example, if an item asks about
ables in the data set. average time spent watching television
saturated test in EXPLORATORY FAC- each day and a scale ranging from none
TOR ANALYSIS, a scale that is highly cor- (1) to 7 hour or more (5) is used, the upper
related with one of the factors being end of the scale is likely to be selected by
measured. Such a test helps define the the great majority of respondents, some
meaning of the factor. of whom will watch many hours per
day; this results in little ability to differ-
saturation n. see SATURATED MODEL. entiate the sample in terms of how
scalability n. the ability of an item on a many hours are actually watched. See
test or scale to elicit responses that repre- CEILING EFFECT; FLOOR EFFECT.
sent identifiable positions on an ordered
scale development the process of cre-
progression of scores or values describ-
ating a new INSTRUMENT for measuring
ing an underlying constract, such as an
an unobserved or latent constract, such
ability or attitude. See GUTTMAN SCALE.
as depression, sociability, or fourth-
scalar 1. n. a quantity having only mag- grade mathematics ability. The process
324
 scatterplot matrix

includes defining the construct and test tendency for a set of scores for an indi-
specifications, generating items and RE- vidual to vaty from the average of scores.
SPONSE SCALES, piloting the items in a
scatter analysis 1. a study of the rela-
large sample, conducting analyses to
tionship between two variables carried
fine-tune the measure, and then re-
out using a SCATTERPLOT. 2. any study
administering the refined measure to de-
of SCATTER, especially in intelligence
velop NORMS (if applicable) and to assess
tesfing and PROFILE ANALYSIS.
aspects of RELIABILITY and VALIDITY.
scatterplot n. a graphical representa-
scale homogeneity see INTERNAL CON- tion of the relationship between two
SISTENCY. continuously measured variables in
which one variable is arrayed on each
scale parameter see DISPERSION PA- axis and a dot or other symbol is placed
RAMETER.
at each point where the values of the
scale value a number that represents an variables intersect. For instance, the hy-
individual's position on an underlying pothetical example below depicts the
construct (trait, ability, attitude, etc.), as weight and average daily calorie con-
measured by an item on a SCALE or the sumption for a sample of individuals.
scale as a whole. The overall pattern of dots provides an
indication of the extent to which there
scaling n. the process of constructing an is a LINEAR RELATIONSHIP between vari-
instrament to measure, assess, and order ables. A round mass of points shows no
some quantity or characteristic (e.g., linear relation, an elliptical cloud of
height, weight, happiness, empathy). A points with a positive slope shows a pos-
researcher must evaluate how a given itive relation, and an eUiptical cloud
CONSTRUCT of interest should be mea- with a negative slope shows a negative
sured and how optimally to obtain scores relation, OUTLIER points are also clearly
for individuals on these measures. See visible. Also called dot plot; scatter-
SCALE DEVELOPMENT. gram. See also REGRESSION LINE.
scalogram n. see GUTTMAN SCALE. scatterplot m a t r i x a matrix in which
all variables being studied are arrayed as
scatter n. 1. the tendency of data points both rows and columns, so that each
to diverge from each other. An example cell reflects the relationship between
is the variation in scores across a set of two variables and the values along the
tests on the same individual; another is MAIN DIAGONAL reflect the distribution
the variation among test takers on a par- of each variable. For example, the
ticular test. See also DISPERSION; VARI- scatterplot matrix overleaf shows re-
ANCE. 2. in PROFILE ANALYSIS, the spondent age, respondent gender, and

3,500 1
3,000
a re 2,500
2,000

I " 1,500
1,000
500
0 I 1 1 1 I

100 200 300 400 500

Weight (lbs)
scatterplot

325
scedasticity

stringent MULTIPLE COMPARISON TESTS

because it is conservative in its identi-
fication of statistically significant mean
differences between groups. [Henty
Scheffe (1907-1977), U.S. mathemati-
cian]
schematization n. the act or process of
reducing very rich information to a sim-
pler scheme or outline, often with the
help of graphical tools.
Schemper's measures a set of indices
that measure the amount of EXPLAINED
VARIANCE in cox REGRESSION ANALYSIS
or other SURVIVAL MODELS in which
the time to an event is predicted by a
AGE SEX INCOME CHILDREN TOGETHER
set of independent variables. [Michael
number of years in current relationship Schemper, Austrian biostatistician]
for a survey of counseling clients. All Schoenfeld residual in a cox REGRES-
pairwise displays (e.g., in a study with 8 SION ANALYSIS of time to an event, the
variables, Variables 1 with 2, Variables 1 RESIDUAL for each individual case for
with 3, ... Variables 7 with 8) are shown each predictor in the model. These resid-
so that the researcher can easily and effi- uals are not defined for CENSORED DATA.
ciently see the associations among vari- Also called score residual. [David A.
ables and other characteristics of the Schoenfeld, U.S. biostatistician]
data.
science n. the systematic srady of strac-
scedasticity n. the disfribution of ERROR ture and behavior in the physical, natu-
TERMS in a set of random variables. The ral, and social worlds, involving the
pattern of errors may be due to chance generation, investigation, and testing of
and have constant variance (homosce- HYPOTHESES, the accumulation of data,
dasticity), or there may be some distinct and the formulation of general laws and
pattern, such as a clustering of greater theories. There are several major
error with certain points on the inde- branches, including physical, biological,
pendent variable (heteroscedasticity). and social sciences. The subdisciplines
REGRESSION ANALYSIS generally assumes of psychology are themselves divided
homoscedasticity (homogeneity of vari- among the different branches. For exam-
ance). ple, neuroscience and the study of the
biological bases for behavior can be
Scheffe test a POST HOC TEST used after
classed with the natural sciences, whereas
a researcher obtains a significant F RATIO social psychology and many aspects of
in an ANALYSIS OF VARIANCE that has clinical psychology can be seen as be-
more than two levels (i.e., more than longing to the social sciences. See SCIEN-
two conditions of an independent vari- TIFIC EXPLANATION; SCIENTIFIC METHOD.
able that are being examined for differ-
ences among their mean values). It allows scientific explanation an account of
for the testing of all possible contrasts an event, behavior, or thought that is
(weighted comparisons of any number couched in terms of an established set of
of means) while controlling the proba- scientific principles, facts, and assump-
bility of a TYPE I ERROR for the set of con- tions. Typical forms of explanation may
trasts at a prespecified level. The Scheffe be reductionistic, analyzing phenomena
test is considered to be one of the most into components and describing how

326
 screening sample

they combine to produce the phenome- to test results or other measurable

non; ontogenic, relating the phenome- responses. 2. vb. to assign scores to re-
non to a universal set of developmental sponses using some predetermined crite-
stages; empiricistic, describing a phe- ria.
nomenon in terms of the conditions
that have been observed to produce it; or score equating the process of ensuring
metaphoric or categorical, identitying a that results from one version or adminis-
phenomenon as similar in some impor- tration of a test have the same distribu-
tant respects to other phenomena al- tion as those obtained from another
ready understood. Such an explanation version or administration, so that the in-
stated systematically is generally known terpretation of the results can be as fair
as a THEORY. as possible to all test takers. Also called
observed-score equating.
scientific hypothesis see HYPOTHE-
SIS; RESEARCH HYPOTHESIS. score residual see SCHOENFELD RESID-
UAL.
scientific method a set of procedures,
guidelines, assumptions, and attirades scorer r e l i a b i l i t y see INTERRATER RE-
required for the organized and system- LIABILITY.
atic collection, interpretation, and veri- scoring n. the application of an answer
fication of data and the discovety of key to a test or survey for the purpose of
reproducible evidence, enabling laws obtaining a value (SCORE) that reflects
and principles to be stated or modified. an individual's position on an under-
See also SCIENCE; SCIENTIFIC EXPLANA- lying construct. The answer key would
TION.
typically indicate individual scores for
scientific notation a compact way of the different responses on each item
reporting vety large or very small num- (e.g., from 1 to 7), such that the values
bers in which the reported number is per item can then be summed or aver-
represented by a value multiplied by 10 aged to obtain a composite score. Some
raised to either a positive or negative instruments may have multiple sub-
number. For example, p = .0025 would scales yielding separate values.
be expressed as 2.5 x 10~^.
screener n. see SCREENING TEST.
scientific theory see THEORY. See also
screening experiment a preliminaty
SCIENCE; SCIENTIFIC EXPLANATION; SCI-
srady in which a large number of factors
ENTIFIC METHOD.
are examined with the aim of filtering
scientism n. an uncritical commitment out those that show little association
to a particular view of science and scien- with the studied outcome. This allows
tific methods that leads its adherents to researchers to narrow the field of poten-
dismiss all other approaches as intellec- tially important factors before devoting
tually invalid. The term is mainly used resources to further study. See PILOT
by those who criticize the assumptions STUDY.
of Western science as arrogant or flawed,
who maintain that scientific methods screening sample in survey methodol-
are inappropriate in certain fields or ogy, a large group of individuals who are
incapable of apprehending certain asked a preliminary set of questions in
kinds of truth, or who reject the implica- order to identity a much narrower subset
tion that all philosophical questions of the population with an attribute of in-
will one day reduce to scientific ques- terest. For example, a nationwide survey
tions. scientistic adj. of businesses might allow researchers to
identity and then study in greater depth
score 1. n. a quantitative value assigned those companies that were started by

327
screening test

business owners over 60 years of age. See ently sized rocks (screes) that form at the
also SCREENING TEST. foot of a mountain. Also caUed Cat-
tell's scree test.
screening test any testing procedure
designed to separate out people or items SD symbol for STANDARD DEVIATION.
with a given characteristic or property.
Screening tests aretypicallyused to dis- SDT abbreviation for SIGNAL DETECTION
tinguish people who have a disease, dis- THEORY.
order, or predisease conditionfromthose SE symbol for STANDARD ERROR.
who do not; they may be used, for exam-
ple, in primary health care settings at seasonal adjustment in TIME-SERIES
intake to identity people who are de- ANALYSIS, the removal of that compo-
pressed and need further clinical atten- nent of the variance that is associated
tion. Screening tests are designed to be with a systematic pattern occurring at
broadly sensitive, and subsequent highly regular intervals over time. For example,
specific or focused testing is often re- if highway accidents occur about 5%
quired to confirm the results. They are more in the winter than in the summer
often designed to be brief to facilitate months, winter figures would have to be
broad classifications. Also called screener; seasonally adjusted to give a true in-
screening study. crease that may be caused by other fac-
tors. See SEASONAL COMPONENT.
scree plot in EXPLORATORY FACTOR
ANALYSIS, a graphic that shows (in de- seasonal component in TIME-SERIES
scending order) the EIGENVALUES of DATA, that element of the variance that
the coMMUNALiTY-adjusted correlation can be accounted for by patterns reoc-
matrix. Consider the generic example curring regularly over time (e.g., certain
below. months of the year, days of the week,
Researchers examine the plot to iden- times of day). Usually the variance asso-
tify a "break" in the curve between ciated with this component can be de-
strong, dominant factors at the top and scribed and modeled. For example,
other smaller factors at the bottom: In researchers studying substance use in
this way they can determine which fac- college sradents might need to account
tors to retain in the analysis. The plot for the days in the week (Thursday, Fri-
was named by Raymond B. Cattell, in day, and Sararday) that are associated
analogy with the sloping heaps of differ- with most use.

scree plot
 selection

secondary data information cited in a an association between two variables is

srady that was not gathered directly by reduced once other factors or variables
the current investigator but rather was that might also be related are removed
obtained from an earlier srady or source. or controlled for (i.e., are held constant
The data may be archived or may be ac- for all participants in the data set).
cessed through contact with the original For example, a researcher might want to
researcher. When consulting or analyz- examine the extent of the association
ing this information, the investigator between conservatism and political be-
should be sensitive to the original re- havior once the effects of age and educa-
search questions and the conditions tion level are removed. Compare FIRST-
under which the observations were ORDER PARTIAL CORRELATION. See also
gathered. Compare PRIMARY DATA. PARTIAL CORRELATION.
secondary s a m p l i n g u n i t in MULTI- second-order stationarity the qual-
STAGE SAMPLING, a grouping of ele- ity of a TIME SERIES such that the MEAN,
ments selected from a larger population VARIANCE, and PROBABILITY DISTRIBU-
(the PRIMARY SAMPLING UNIT). For ex- TION remain constant over time and the
ample, in a srady of coUege sradent atti- AUTOCOVARIANCE depends only on the
tudes, a researcher might first sample lag, or distance between pairs of time
small, medium, and large colleges and points. See STATIONARITY.
institutions and then specify four geo- second quartile see QUARTILE.
graphic regions that need to be repre-
sented within those larger units. Com- sectional bar g r a p h see COMPONENT
pare TERTIARY SAMPLING UNIT. BAR GRAPH.

second moment see MOMENT. SEE symbol for STANDARD ERROR OF

ESTIMATE.
second-order f a c t o r a latent con-
struct that emerges from a further FAC- segmented bar g r a p h see COMPO-
TOR ANALYSIS (i.e., a second-order factor NENT BAR GRAPH.
analysis) of the primaty dimensions de- segmented regression see PIECEWISE
rived from correlations among a set of REGRESSION.
items or variables. The higher order di-
mensions so derived are held to generate segregation analysis a statistical
the FIRST-ORDER FACTORS that in rarn method to examine the probability that
generate observed responses at the level offspring will have certain inherited at-
of the individual item or scale. tributes, traits, or phenotypes. For exam-
ple, a cUnical study of a disorder may
second-order interaction in ANALY- examine families of the affected person
SIS OF VARIANCE Or REGRESSION ANALY- seeking treatment.
SIS, an effect in which three inde-
pendent variables combine to have a selected group a SAMPLE explicitly
nonadditive influence on a dependent chosen with respect to specific criteria
variable. See HIGHER ORDER INTERAC- related to the purpose of the research.
TION. For example, in a study of attitudes of
older adults a researcher might choose a
second-order M a r k o v model see sample of citizens ages 65 and over from
FIRST-ORDER MARKOV MODEL. four geographic regions in the United
States. Also called selected sample.
second-order partial correlation
the correlation between two variables selection n. the process of choosing an
(e.g., X and y), with the effects of two ad- item (e.g., an individual, an object, a
ditional variables (e.g., a and b) re- measurement) from a larger universe of
moved. It is often useful to see whether units for a purpose, such as srady, test-

329
selection bias

ing, classifying, or working (hiring em- ple of data that could be appropriate for
ployees). this method is women's salaries: Some
aspects of the data could be modeled
selection bias see SAMPLING BIAS.
using standard regression methods, but
selection invariance in choosing the preponderance of zero salaries (re-
among applicants for, say, employment flecting a decision by some women not
or admission to college, the property of a to work) would also need to be incorpo-
selection procedure such that it is rated into the analysis. In such cases the
equally efficient (i.e., makes a similar two-stage regression method uses a model
number of errors) for all subgroups of that relies on BIVARIATE NORMALITY to
applicants (e.g., ethnic or income estimate the selection equation and an
groups). Selection invariance is the em- ordinaty LEAST SQUARES REGRESSION to
pirically testable assumption that there estimate the remainder. See also TRUN-
is equal SENSITIVITY and equal SPECIFIC- CATED DISTRIBUTION; ZERO-INFLATED
ITY across all groups. Compare MEA- POISSON REGRESSION.
SUREMENT INVARIANCE.
selection ratio the proportion of all
selection methods i n regression in those eligible to be selected for a purpose
t h e G E N E R A L LINEAR M O D E L , v a r i O U S SCtS who actually are selected. In personnel
of decision rules that help researchers selection, for example, it is the number
determine which predictors or INDE- of applicants hired to perform a job di-
PENDENT VARIABLES are related to the vided by the total number of applicants.
outcome or DEPENDENT VARIABLE. Some The lower the selection ratio, the more
methods are theoty driven and involve competitive the hiring situation will be
entering variables into the model in a and the more useful, all other factors
preconceived order (i.e., HIERARCHICAL being held constant, any given predictor
REGRESSION). Other approaches apply a will be in making selection decisions.
rule (usually based on STATISTICAL SIG- selection threat the THREAT TO VALID-
NIFICANCE) for the inclusion of a vari-
ITY arising from any form of SAMPLING
able into the model or its exclusion from BIAS (i.e., nonrandomly selecting units
the model (i.e., STEPWISE REGRESSION). for a study).
For example, FORWARD-SELECTION ap-
proaches begin with no variables in the selective dropout the nonrandom loss
model and proceed by including the of participants from a srady that occurs
variable (or variables) with the strongest when an identified fearare of the study
association with the outcome until there design (e.g., topic sradied, number of
is no increase in EXPLAINED VARIANCE tasks) interacts with respondent char-
by the addition of remaining variables. acteristics (e.g., depression, education
Conversely, BACKWARD-ELIMINATION ap- level). See ATTRITION.
proaches begin with all possible vari-
ables in the model and proceed by suc- selective observation a process in
cessive deletion of the variable (or which an individual attends to behav-
variables) that contribute least to the iors, attitudes, and interpersonal set-
prediction or explanation of the out- tings that correspond to his or her
come. current beliefs or self-interests. When
listening to a speech by a presidential
selection model a two-stage REGRES- candidate, for example, a person with
SION ANALYSIS used to estimate PARAME- strong views on national spending for
TERS and STANDARD ERRORS in an social programs may listen for aspects of
. unbiased way when scores for one aspect the speech that address those specific be-
of the sample may be affected by a liefs. Compare UNSELECTIVE OBSERVA-
nonrandom selection process. An exam- TION. See also CONFIRMATION BIAS.

330
 semantic differential

self-administered test a test in which the degree to which the descriptors

the instructions are sufficiently self- listed apply to them. Also called self-
evident not to require further clarifica- report scale.
tion (e.g., by the researcher), enabling
self-selected groups design an ex-
the respondent to complete it by him- or
perimental design in which participants
herself. Compare PROCTORED TEST.
choose their group or the condition to
s e l f - f u l f i l l i n g prophecy a belief or which they will be exposed. Because
expectation that helps to bring about its the assignment of participants to re-
own fulfillment, as, for example, when a search conditions is nonrandom, causal
person expects that nervousness will im- inference from data gleaned in such ex-
pair his or her performance during a job periments is questionable. See QUASI-
interview or when a teacher's precon- EXPERIMENTAL DESIGN.
ceptions about a sradent's ability influ-
ence the pupil's achievement for better self-selection bias a type of BIAS that
or worse. See PYGMALION EFFECT; UP- can arise when study participants choose
WARD PYGMALION EFFECT. See alSO DE- their own treatment conditions, rather
MAND CHARACTERISTICS; EXPERIMEN- than being randomly assigned. In such
TER EXPECTANCY EFFECT. cases it is impossible to state unambigu-
ously that a study result is due to the
self-monitoring observation the treatment condition and not to the pre-
study of one's own behaviors, attitudes, existing characteristics of those individ-
or emotions over time. In research stud- uals who chose to be in this condition.
ies and clinical interventions focused on Also called self-selection effect. See
weight loss, for example, participants also SAMPLING BIAS.
are encouraged to keep track of their eat-
ing patterns, their feelings related to SEM 1. abbreviation for STRUCTURAL
their eating, and triggers for not main- EQUATION MODELING. 2. Symbol for
taining their diet. This enables a clini- STANDARD ERROR OF MEASUREMENT. 3.
cian or researcher to review periods symbol for STANDARD ERROR OF THE
when the participant is outside the clini- MEAN.
cal setting or laboratoty. Compare ANA-
semantic d i f f e r e n t i a l a type of scale
LOGUE OBSERVATION; NATURALISTIC
that researchers use to assess a respon-
OBSERVATION.
dent's views on a certain topic (e.g., a
self-rating scale any questionnaire, in- stimulus such as a word or photograph,
ventoty, survey, or other instrament in the quality of some experience). Parti-
which participants are asked to assess cipants are asked to rate the topic or
their own characteristics (e.g., attitudes, stimulus on a scale that has pairs of op-
interests, abilities, performance). See also posites, such as bad-good, unpleasant-
SELF-REPORT; SELF-REPORT INVENTORY. pleasant, or competitive-cooperative, as
ANCHORS or reference points. For exam-
self-report n. a statement or series of
ple, the bipolar opposites of bad to good
answers to questions that an individual may be scaled along 7 points and the
provides about his or her state, feelings, respondent asked to position him- or
thoughts, beUefs, past behaviors, and so herself on the 7-point continuum. Gen-
forth. Self-report methods rely on the erally, the anchors are focused on three
honesty and self-awareness ofthe partic- dimensionsevaluation, activity, and
ipant and are used especially to measure potency. Responses to items are then
behaviors or traits that cannot easily be scaled in some way (theory, EXPLOR-
directly observed by others. ATORY FACTOR ANALYSIS) SO that items
self-report inventory a type of ques- can be averaged or summed to arrive at a
tionnaire on which participants indicate final index of attirades. This procedure

331
semi-interquartile range

is one of the most widely used methods probability that a test gives a positive
of assessing attitudes and may be used in diagnosis given that the individual actu-
psychometric testing or in a wide array ally has the condition for which he or
of settings, such as marketing and poli- she is being tested. Compare SPECIFIC-
tics, to gauge public reactions to a prod- ITY.
uct, issue, or personality.
sensitivity analysis an analysis that
semi-interquartile range see QUAR- measures the extent to which the overall
TILE DEVIATION. outcome of a model or system will be af-
fected by potential changes to the input.
semi-Markov process a STOCHASTIC This type of analysis is often used where
PROCESS with a finite set of states in the values of key variables are uncertain
which the "jump" (transition) from each or subject to change: In financial plan-
state to the next is memoryless but the ning, for example, it might be used to
process as a whole is not. determine whether a potential change
semiparametric model a model that in mortgage rates, cash flow, local house
combines a PARAMETRIC component that prices, or all of these might alter the
has strict assumptions relating to the decision to purchase a new home. In re-
distribution of the variables and a NON- search studies, sensitivity analysis en-
PARAMETRIC component that does not ables researchers to understand the
rely on distribution parameters. An ex- boundaries of their statistical models
ample of this flexible hybrid model is and design updated models that can ac-
the Cox proportional hazards model (see count for the data at hand.
cox REGRESSION ANALYSIS). Sen's slope estimator see THEIL-SEN
semipartial correlation see PART ESTIMATOR.
CORRELATION. separate-variances t test a variant
semistructured interview an inter- o n t h e I N D E P E N D E N T - S A M P L E S T TEST
view format that involves the inter- used where the samples being compared
viewer asking a specified set of questions have differing VARIANCES. It involves
making a special adjustment to the DE-
in a particular order, while also allowing
GREES OF FREEDOM.
for open-ended responses by the inter-
viewee and a more natural conversa- sequence effect in WITHIN-SUBJECTS DE-
fional style. In a study of health SIGNS, a difference in scores that emerges
professionals, for example, interviewers because of a particular arrangement of
might follow a guide that includes sev- treatments; that is, the presentation of
eral open-ended questions in which the one level of the independent variable
professionals are asked to reflect on their has an effect on responses to another
patients' use of their medications. By level of that variable. A researcher can
using prompts such as "Tell me more" test for a sequence effect by administer-
and "Why is that?", the interviewers ing the treatments in various different
would encourage a full and rich re- arrangements (e.g., the arrangement ABC
sponse. This approach recognizes that a VS. ACB, vs. BCA, etc.). The sequence ef-
researcher may still learn valuable infor- fect is not to be confused with the
mation even when the interviewee ORDER EFFECT.
moves away a little from the topic at
hand. See also STRUCTURED INTERVIEW. sequential analysis a class of statisti-
cal procedures in which decisions about
sensitivity n. 1. the capacity to detect sample size and the type of data to be
and discriminate differences. In SIGNAL collected are made or modified as the
DETECTION THEORY, sensitivity is mea- srady proceeds, based on the cumulative
sured by the index D PRIME (d'). 2. the findings to date. This approach contrasts

332
 Shannon index

with one in which the sample size is de- s e q u e n t i a l s u m o f s q u a r e s see TYPE I

termined in advance and data are not SUM OF SQUARES.
analyzed until the entire sample is col- s e q u e n t i a l test d e s i g n see SEQUEN-
lected. A common form of sequential
TIAL DESIGN.
analysis is one in which data are col-
lected until a desired outcome or level of s e q u e n t i a l t h r e s h o l d m e t h o d a
precision is reached; also, some clinical form of CLUSTER ANALYSIS that involves
trials require an approach that allows re- identifying the center of the variables
searchers to stop data collection if the being analyzed (the CENTROID), group-
treatment is clearly not working or if ing all variables within a certain distance
participants are being harmed in some of that point into a cluster, and then re-
way. Also called sequential hypothe- peating the process several times for pro-
sis testing. gressively larger distances. The method
emphasizes some clusters over others by
s e q u e n t i a l d e s i g n 1. a research design prioritizing those that are closest to the
that allows for termination of the study center point. See also K - M E A N S CLUSTER-
at various points of data collection if I N G . Compare P A R A L L E L T H R E S H O L D
the results do not conform to a desired METHOD.
pattern or if there is danger or cost to
s e r i a l c o r r e l a t i o n see A U T O C O R R E L A -
participants. Also called s e q u e n t i a l
TION.
test design. See SEQUENTIAL ANALYSIS.
2. see COHORT-SEQUENTIAL DESIGN. s e r i a l d e p e n d e n c e see A U T O C O R R E L A -
TION.
sequential probability ratio test set t h e o r y the branch of mathematics
(SQRT) a form of SEQUENTIAL ANALYSIS
and logic that is concerned with the
in which samples of units are drawn in properties of sets (i.e., collections of en-
sequence and tested to see if units meet tities that are themselves treated as enti-
some specified criterion. Inferential ap- ties).
proaches allow researchers to determine
the exact point at which the attribute of s e t w i s e r e g r e s s i o n see ALL-POSSIBLE-
interest reaches a required level and no SUBSETS REGRESSION.
more sampling is needed. The SQRT was s e x - s p e c i f i c r a t e the RATIO of the
developed in Q U A L I T Y C O N T R O L sradies number of instances of some variable
but has been extended to many other (e.g., birth, mortality, incidence of dis-
areas. ease) to the total population in some
specified period, as reported separately
s e q u e n t i a l r e g r e s s i o n see HIERARCHI- for males and females. Because many
CAL REGRESSION. diseases and life processes differ for men
and women, it is more useful in most
s e q u e n t i a l s a m p l i n g a method of
cases to obtain the sex-specific rate.
NONPROBABILITY SAMPLING in which
the researcher draws a group of units Shannon index a DIVERSITY INDEX
from the larger population, conducts a widely used in ecological studies, where
srady within a specified time frame, ana- it provides important information about
lyzes the data, and then determines the rarity or abundance of species in a
whether another sample is needed. The community. It is calculated by identify-
process can be repeated several times. ing the PROBABILITY of each species in
The sequential approach enables a re- the system, multiplying each probability
searcher to determine when enough value by the LOGARITHM (usually the
data have been collected and to fine- natural logarithm), and taking the nega-
tune his or her methodology over re- tive sum of those numbers. Also called
peated studies. Shannon's information measure.

333
Shapiro-Wilk test

Compare SIMPSON INDEX. [Claude Shan- shrinkage f o r m u l a any of various for-

non (1916-2001), U.S. mathematician] mulae used to estimate the degree of
S h a p i r o - W i l k test an INFERENTIAL SHRINKAGE that will occur when a statis-
TEST to determine whether a random tic or REGRESSION EQUATION is applied
sample comes from a population with a to a new data set. The degree of shrink-
NORMAL DISTRIBUTION. If the test Statis- age will generally depend on the size of
tic, W, is Significant, then the NULL HY- the initial sample used to obtain the sta-
POTHESIS that the distribution is normal
tistic and the number of predictors in
should be rejected. Because so many of the model.
the most common statistical tests are de- s h r u n k e n R^ see ADJUSTED R^.
signed for normally distributed data,
Shapiro-Wilk is a useful test to identify Sidak test see DUNN-SIDAK PROCEDURE.
those data sets that require a different Siegel-Tukey test a NONPARAMETRIC
approach. [Samuel S. Shapiro (1930- ), TEST that evaluates any difference in
U.S. statistician; Martin Wilk (1922- VARIANCE across two independent sam-
2013), Canadian statistician] ples to determine whether they repre-
short-answer test a test that uses sent two different populations. Data
such item types as multiple choice, fiU- values for the two groups are first or-
in-the-blanks, true-false, and matching dered in a single list from lowest to high-
alternatives, as opposed to one requiring est and then ranked in terms of alternate
lengthy, open-ended answers. Short- extremes, so that, for example, the high-
answer tests can be quickly scored and est and lowest values are both ranked 1,
are generally preferred when time and the next highest and next lowest 2, and
resources are limited. so on. FinaUy, the ranked values in each
group are summed. If final sums per
shotgun approach an unsystematic group do not differ, there is no evidence
way of conducting research in which the that the dispersion differs across groups.
investigator examines a large number of The test has relatively low statistical
variables, often without a theoretical POWER when data have a NORMAL DIS-
basis, in the hopes that some relation- TRIBUTION.
ships will be found. Any study effects
that emerge are unlikely to be cross- sigma n. 1. (symbol: S) the sum of a list
validated and may not be stable. of values. 2. (symbol: 0) see POPULATION
STANDARD DEVIATION.
shrinkage n. the situation in which the
strength of a CORRELATION COEFFICIENT sigmoid curve an S-shaped curve that
or REGRESSION EQUATION decreases describes many processes in psychology,
when it is applied to a new data set. Such including learning and responding to
shrinkage occurs when the initial esti- test items. The curve starts low, has a pe-
mate of the correlation reflects unique riod of accelerafion, and then approaches
characteristics of the initial sample, an ASYMPTOTE. Often the curve is char-
which are not replicated in subsequent acterized by the LOGISTIC FUNCTION.
samples. signal detection theory (SDT) a
shrinkage estimator a PENALTY, FUNC- body of concepts and techniques from
TION used to reduce the value of a statis- communication theory, electrical engi-
. tic: The estimator reflects some known neering, and DECISION THEORY that
aspect of the model, such as a high num- were applied during World War II to the
ber of independent variables (i.e., its detection of radar signals in noise. The
complexity). In MULTIPLE REGRESSION, same concepts were applied to auditory
the R^ value is often the target of this re- and visual psychophysics in the late 1950s
duction, yielding an ADJUSTED R^ value. and are now widely used in many areas

334
 Simes modified Bonferroni procedure

of psychology. An important method- statistically significant (i.e., unlikely to

ological contribution of SDT has been arise solely from chance). In its most
the refinement of psychophysical tech- common form, significance testing is
niques to permit the separation of SENSI- used to decide whether the NULL HY-
TIVITY from criterial, decision-making POTHESIS of no effect should be rejected.
factors. SDT has also provided a valuable A comparison of the probability statistic
theoretical framework for describing obtained from the test to the chosen SIG-
perceptual and other aspects of cogni- NIFICANCE LEVEL determines whether an
tion and for quantitatively relating observed effect may be due to chance
psychophysical phenomena to findings variance and hence whether the null hy-
from sensoty physiology. A key notion pothesis is or is not likely to be correct.
of SDT is that human performance in This approach may also be used to differ-
many tasks is limited by variability in entiate between two models that differ
the internal representation of stimuli in terms of the number of parameters
due to internal or external noise. See D specified in them (as in MULTIPLE RE-
PRIME; R E C E I V E R - O P E R A T I N G CHARAC- GRESSION analysis).
TERISTIC CURVE.
s i g n i f i c a n t difference the situation
signed-ranks test see WILCOXON
in which a SIGNIFICANCE TESTING proce-
SIGNED-RANKS TEST. dure indicates the statistical differences
significance n. the extent to which observed between two groups (e.g., a
something is meaningful or of conse- treatment group and a control group)
quence. In Statistics and related fields, are unlikely to reflect chance variation.
the term usually denotes STATISTICAL
sign test a NONPARAMETRIC procedure
SIGNIFICANCE. See also CLINICAL SIGNIF-
for testing the hypothesis that two re-
ICANCE; PRACTICAL SIGNIFICANCE; PSY-
lated samples have come from two dif-
CHOLOGICAL SIGNIFICANCE.
ferent populations. The procedure is
significance level (symbol: a) in SIG- appropriate to ORDINAL DATA, that is,
NIFICANCE TESTING, afixedprobability scores that have a meaningful order but
of rejecting the NULL HYPOTHESIS of no lack equal distances between points. The
effect when it is in fact true. It is set at scores within each sample are first
some value, usually .001, .01, or .05, de- ranked. For each pair of corresponding
pending on the consequences associated scores (e.g., a participant's score in the
with making a TYPE I ERROR. When a first set and its pair from the second set)
particular effect is obtained experimen- the sign of the difference (either positive
tally, the PROBABILITY LEVEL (p) associ- or negative) in the rankings is recorded.
ated with this effect is compared to the If the proportion of participants who ob-
significance level. If the p value is less tain a positive sign is significantly differ-
than the a level, the null hypothesis is ent from those obtaining a negative
rejected. Small p values suggest that ob- sign, then the data are considered to
taining a statistic as extreme as the one come from different populations.
obtained is rare and thus the null hy-
pothesis is unlikely to be trae. The Simes m o d i f i e d B o n f e r r o n i proce-
smaller the a level, the more convincing dure a method for providing strong
is the rejection of the null hypothesis. control of the FAMILY-WISE ERROR RATE
Also called alpha level. in multiple POST HOC COMPARISONS of
means. It is a less conservative variant
significance testing in HYPOTHESIS of the DUNN-BONFERRONI PROCEDURE.
TESTING, a set of procedures used to de- [R. J. Simes, Australian medical statisti-
termine whether- the differences be- cian; Carlo Emilio Bonferroni (1892-
tween two groups or models are 1960), Italian mathematician]

335
SIMEX

SIMEX abbreviation for SIMULATION TERACTION EFFECT among two inde-

AND E X T R A P O L A T I O N PROCEDURE. pendent variables, an examination of
the effect of one variable at one level of
similarities test a test in which the
the other variable. For example, if there
participant must either state the like-
were two levels of a factor, a, and 02, and
nesses between items or arrange items in
two levels of a second factor, b^ and i>2,
categories according to their similarities.
the comparison of vs. 02 at b^ would
Some neuropsychological batteries in-
represent one simple-effects analysis;
clude these assessments.
another would be a comparison of a^ vs.
s i m i l a r i t y coefficient any index that 02 at In this way a series of simple-
allows a researcher to assess the similar- effects analyses can be used to break
ity of two or more samples. In EXPLOR- down an interaction into its component
ATORY FACTOR ANALYSIS, for example, parts.
such an index is used to compare FAC-
simple event see ELEMENTARY EVENT.
TOR STRUCTURES across studies. See also
CONGRUENCE COEFFICIENT. Compare DIS- simple hypothesis a hypothesis that
SIMILARITY C O E F F I C I E N T . specifies all the parameters of a popula-
tion distribution. For example, consider
s i m i l a r i t y m a t r i x a matrix whose ele-
a researcher who wants to examine the
ments measure pairwise similarities of
hypothesis that graduates from a partic-
itemsthe greater the similarity of two
ular institution have higher than aver-
items, the greater the value entered in
age overall grade point averages if they
the matrix at this point. A variety of
have studied abroad. If the researcher
METRICS can be used to measure similar-
specified the population mean and stan-
ity, including proximities and correla-
dard deviation of grades in the full popu-
tions. Compare DISSIMILARITY MATRIX.
lation of students at that institution, this
simple analysis of variance an would be a simple hypothesis. Compare
ANALYSIS OF VARIANCE involving one COMPOSITE HYPOTHESIS.
DEPENDENT VARIABLE and One INDE- simple random sampling the most
PENDENT VARIABLE that has two or more basic approach to drawing a RANDOM
levels. SAMPLE of cases, observations, or indi-
simple comparison a contrast be- viduals from a population, in which the
tween two means, usually in the context cases are selected individually using a
of multilevel analyses of data from a fair process, such as the toss of a coin or a
FACTORIAL DESIGN. For example, con- table of RANDOM DIGITS. Also called in-
sider a researcher examining the influ- dependent random sampling.
ence of three different amounts of caf-
simple regression a type of REGRES-
feine (0 mg, 50 mg, and 100 mg) on stu-
SION ANALYSIS that has only one predic-
dent test performance. His or her evalua-
tion of the differences between 0 mg and tor or independent variable and one
50 mg would represent one possible sim- outcome or dependent variable. See also
MULTIPLE REGRESSION.
ple comparison. Compare COMPLEX COM-
PARISON. simple structure in EXPLORATORY
FACTOR ANALYSIS, a set of Criteria for de-
simple correlation the linear associa- termining the adequacy of a FACTOR RO-
tion of one variable with one other vari- TATION solution. These criteria require
able, as quantifled by a CORRELATION that each factor show a pattern of high
COEFFICIENT. FACTOR LOADINGS on Certain variables
simple-effects analysis where an and near-zero loadings on others and
ANALYSIS OF VARIANCE Or MULTIPLE RE- that each variable load on only one fac-
GRESSION analysis has identifled an IN- tor. This minimizes the complexity of

336
 simultaneous comparison method

the factor solution, allows each variable s i m u l a t i o n n. 1. an experimental

to be most strongly identified with a spe- method that is used to investigate the
ciflc factor, and increases interpretabil- behavior and psychological processes
ity. and functioning of individuals in social
and other environments, often those to
simplex model a model stating that which investigators cannot easily gain
when V variables are ordered by time, se- access, by reproducing those environ-
verity, or some other attribute, their as- ments in a realistic way. For example,
sociations in a V X V CORRELATION simulations are often used in personnel
MATRIX will have a particular stracrare. selection, where various exercises have
The variables will have stronger rela- been developed to tap job-related di-
tions the closer they are to one another mensions or behaviors; such exercises
and weaker relations the further they are need to be pretested to ensure that the
apart. Also called simplex pattern. techniques are objective and relevant for
the assessment of a candidate's potential
Simpson index a commonly used DI- for, say, a management position. 2. the
VERSITY INDEX that measures the proba-
artificial creation of experimental data
bility that two randomly selected through the use of a mathematical or
individuals from a sample will be from computer model. The purpose is usually
different groups or categories. It is calcu- to test the behavior of a statistic or
lated as the sum of the squared probabil- model under controlled conditions.
ities (p) of each group. For example, if a
sample of students contained freshmen s i m u l a t i o n a n d extrapolation
(p = .30), sophomores (p = .10), juniors procedure (SIMEX) a method used to
(p = .40), and seniors (p = .20), the estimate a CUMULATIVE FREQUENCY DIS-
Simpson index would be .30 (.09 + .01 + TRIBUTION for an entire population in
.16 + .04), meaning that two randomly the presence of MEASUREMENT ERROR. It
selected individuals have a 30% chance involves adding extra measurement
of being from different classes. Compare error in known quantities, establishing a
SHANNON INDEX. [Edward H. Simpson relationship between the bias thus in-
(1922- ), British stafisfician] duced and the variance of the error, and
extrapolating back to the case in which
Simpson's paradox a phenomenon there is no measurement error.
that can occur when data from two or
more studies are merged, giving results simulator n. a training device that sim-
that differ from those of either study in- ulates the conditions or environment of
dividually. For example, two sradies, each the actual operating siraation or that
showing a correlation of .00 between resembles the actual equipment to
two variables, x and y, may show a sfrong be used, such as a flight simulator for pi-
positive correlation between variables x lots. The training allows for a safe, lower
and y when the data are merged. This cost experience with the conditions
paradoxical effect could occur if the under study but in most cases lacks real-
mean values of each variable in one ism.
srady are both substantially different
(e.g., lower) than their mean values in simultaneous comparison method
the second study. Simpson's paradox a strategy used in sensoty discrimination
highlights the fact that important rela- research, such as color or sound discrim-
tions can be masked under certain con- ination, in which the stimuli for com-
ditions. Also called Yule-Simpson . parison are viewed simultaneously. This
effect; Yule's paradox. [Edward H. is in contrast to the successive compari-
Simpson; George Udny Yule (1871- son method, in which stimuli are succes-
1951), British statistician] sively presented.

337
simultaneous confidence intervals

simultaneous confidence intervals Also called N-of-1 design; N=l design;

joint CONFIDENCE INTERVALS that are single-participant design; single-
formed for estimating multiple parame- subject (case) design.
ters simultaneously from the same set of
single-factor analysis of variance
data. Various techniques have been de-
see ONE-WAY ANALYSIS OF VARIANCE.
veloped for obtaining confidence inter-
vals for a finite or infinite set of single-factor design see ONE-WAY DE-
parametric functions, such that the SIGN.
probability of the parametric functions
of the set being simultaneously covered single-factor m u l t i l e v e l design a
by the corresponding intervals is a preas- research design involving one INDE-
PENDENT VARIABLE (e.g., treatment mo-
signed value.
dality) with more than two conditions
simultaneous equations a set of or levels (e.g., cUnical therapy plus medi-
equations containing two or more vari- cation, medication only, clinical ther-
ables, the values for which are to be apy only). See ONE-WAY ANALYSIS OF
found simultaneously. Many MULTI- VARIANCE.
VARIATE problems in psychology that
express complex relationships among single-group v a l i d i t y the notion that
constructs require such an approach to some measures may be valid for certain
obtain parameter estimates. See STRUC- groups (e.g., Caucasians) but not others
TURAL EQUATION MODELING.
(e.g., minority ethnic groups). See also
DIFFERENTIAL VALIDITY.
simultaneous regression a type of
single-linkage clustering in HIERAR-
REGRESSION ANALYSIS in which all pre-
CHICAL CLUSTERING, a method in which
dictors or INDEPENDENT VARIABLES are
the proximity between one group of x
entered into the equation at the same items (e.g., people, objects) and another
time. Each independent variable's coef- group of y items is computed as the dis-
ficient or WEIGHT is interpreted in the tance between the two closest elements
context of all of the other independent of the groups. Also called nearest
variables in the model at that time, some neighbor. Compare AVERAGE-LINKAGE
of which may be correlated. Also called
simultaneous multiple regression. CLUSTERING; COMPLETE-LINKAGE CLUS-
Compare HIERARCHICAL REGRESSION; TERING.
STEPWISE REGRESSION. single masked see BLIND.
single b l i n d see BLIND. single-participant design see SINGLE-
C A S E DESIGN.
single-case design an approach to the
empirical study of a process that tracks a single-sample runs test a statistical
single unit (e.g., person, family, class, analysis to detect if there is a non-
school, company) in depth over time. It random pattern in a sequence of dichot-
is a WITHIN-SUBJECTS DESIGN w i t h juSt omous outcomes. For example, suppose
one UNIT OF ANALYSIS. Such studies are one flipped a coin multiple times and
useful for generating ideas for broader observed the outcomes of heads (H) and
studies and for focusing on the micro- tails (T). Defining a run as a succession of
level concerns associated with the par- identical outcomes that is followed and
ticular unit. However, data from these preceded by different outcomes, for the
studies need to be evaluated carefully series of flips yielding HTTTTHTT there
given the many potential threats to IN- are two runs oif length one for heads, a
TERNAL VALIDITY; there are also issues run of four tails, and a ran of two tails.
relating to the sampling of both the The single-sample runs test is concerned
one unit and the process it undergoes. with whether there are fewer or more

338
 skewness

rans than would be expected by chance.single-subject design see SINGLE-

Also called one-sample runs test. CASE DESIGN.
singly censored data a set of CEN-
single-sample test any of various sta-
SORED DATA in which there is only one
tistical procedures used to analyze data
censoring point (i.e., one threshold be-
from one sample and determine whether
yond which values are unknown). Most
the distribution of values differs signifi-
often, single censoring results from the
cantly from a known or theoretical dis-
nonoccurrence of the event of interest
tribufion for the larger POPULATION from
during the study observation period: It is
which the sample is believed to derive.
therefore particularly common in SUR-
Also called one-sample test. Compare
VIVAL ANALYSIS. See also DOUBLY CEN-
TWO-SAMPLE TEST.
SORED DATA.
single-sample test for the median s i n g u l a r m a t r i x a SQUARE MATRIX
any of a variety of statistical procedures whose inverse does not exist. The fol-
used to determine whether the middle lowing is an example.
value (MEDIAN) for a given SAMPLE dif-
fers significantly from a hypothesized
value of the median in the larger POPU-
LATION. Also called one-sample test
for the median. A singular matrix has a zero DETERMI-
NANT and cannot be used within FACTOR
single-sample t test a statistical proce-. ANALYSIS and other multivariate proce-
dure used to determine whether there is dures. Compare NONSINGULAR MATRIX.
a significant difference between the ob-
served mean of a SAMPLE and the known singular value decomposition
or hypothetical mean ofthe larger POPU- (SVD) a specialized form of MATRIX DE-
LATION from which the sample is ran- COMPOSITION, a process used in linear
domly drawn. In the single-sample t test, algebra to break a complex matrix down
one calculates a t value using (a) the into its simpler component elements.
mean of the sample (observed average); situational differences any distinc-
(b) the mean of the population (known tion arising from environmental charac-
or theoretically expected average); (c) teristics, as opposed to INDIVIDUAL
the STANDARD DEVIATION of the sample; DIFFERENCES. For example, in a srady of
and (d) the total number of sample ob- men and women in social settings versus
servations taken. The t value obtained is business settings, the amount of positive
then compared to a standard table of emotion displayed by the participants
values, arranged by sample size, to deter- may be explained not only by gender
mine whether it exceeds the threshold (e.g., females tending to exhibit more
of STATISTICAL SIGNIFICANCE. Also positive emotion) but also by the situa-
called one-sample t test. tion (more positive emotions in a social
setting regardless of whether the partici-
single-sample z test a statistical proce- pant is male or female).
dure used to determine how closely the
MEAN value on a variable for a given skewness n. the degree to which a set of
SAMPLE corresponds to the mean of the scores, measurements, or other numbers
same variable for a larger POPULATION are asymmetrically distributed around a
with a known mean. Also called one- central point (see example overleaf).
sample z test. See also z TEST. A normal FREQUENCY DISTRIBUTION of
data is shaped like a bell, with equal values
single-subject case design see SIN- for each of its three indices of CENTRAL
GLE-CASE DESIGN. TENDENCYthe MEAN, the MEDIAN, and

339
skew-symmetrical matrix

positive skew negative skew

skewness
the MODE. Approximately 68% of the slope analysis see REGRESSION ANALY-
scores lie within 1 STANDARD DEVIATION SIS.
of the mean and approximately 95% of
the scores lie within 2 standard devia- smallest space analysis (SSA) a sta-
tions of the mean. When a distribution tistical technique for creating a visual
has a few extreme scores toward the high representation of-data, in which more
end relative to the low end (e.g., when a closely correlated variables are grouped
test is difficult and few test takers do together. Smallest space refers to the
well), it has a positive skew (or is posi- fewest number of geometric dimensions
tively skewed), such that the mean is (e.g., one dimension may be denoted by
greater than the mode. When a distribu- a line, two by a square, three by a cube)
tion has a few extreme scores toward the by which a body of data may be ade-
low end relative to the high end (e.g., quately represented. Similar in purpose
when a test is easy and most test takers to FACTOR ANALYSIS and PRINCIPAL
do well), it has a negative skew (or is COMPONENTS ANALYSIS, Smallest space
negatively skewed). analysis may be appUed to the assess-
ment of objects, persons, attitudes, test
skew-symmetrical m a t r i x a SQUARE constracts, and other similar variables.
MATRIX whose TRANSPOSE is equal to its
negation. small expected frequencies an issue
that arises when the CHI-SQUARE TEST is
original skew-symmetrical used to study GOODNESS OF FIT for vari-
ables in a CONTINGENCY TABLE. The chi-
1 2 0 -1 -2 square test is based on the assumption
0 1 1 0 - 1 that the obtainedfrequencieswdthin any
-1 0 2 1 0 cell of the table will be distributed nor-
maUy around the theoretically EXPECTED
It is symmetrical about its MAIN DIAGO- FREQUENCY. If the expectedfrequencyis
NAL, with each element negated, as small, however, the observed frequen-
shown in the example above. See also cies cannot be normally distributed and
SYMMETRICAL MATRIX. thus the chi-square test is inappropriate.
Generally, the expected frequency for a
slope n. the steepness or slant of a line on cell should be five or more.
a graph, measured as the change of value
on the Y-AXis associated with a change small-N design a research approach in
of one unit of value on the X-AXIS. In a which only a few participants are stud-
REGRESSION EQUATION, slope is repre- ied in order to focus attention upon the
sented by the variable b, with +b indicat- individual rather than the group as a
ing an upward slope to the line and -b whole. Typically, such designs are con-
indicating a downward slope. See also ducted over a longer period of time, to
ACCELERATION. allow for the behavior of interest to sta-

340
 socially sensitive research

bilize, and employ QUALITATIVE ANALY- asked to nominate others, and so on until
SES as opposed to INFERENTIAL STAT- a sufficient number of participants is ob-
ISTICS iri analyzing results. Although tained. Researchers generally use snow-
small-N designs have several advan- ball sampUng if the population of inter-
tagesincluding reduced cost since est is hard to locate, rare (e.g., people
fewer participants are involved, better who have an infrequent condition or
control of EXTRANEOUS VARIABLES, and disease), or otherwise limited. Although
more comprehensive records of partici- this nonprobability strategy of chain re-
pant performancea significant disad- ferral is simple and cost efficient, there is
vantage is that EXTERNAL VALIDITY may the potential for SAMPLING BIAS in that
. be lower if parficipants are not representa- initial participants may tend to nomi-
tive of the larger population from which nate people they know well and thus are
they are drawn. Compare LARGE-N DESIGN. likely to share the same traits and char-
acteristics.
small-sample theory the understand-
ing and application of appropriate social desirability the bias or ten-
methods for analyzing data from groups dency of individuals to present them-
(samples) comprising 30 or fewer indi- selves in a manner that will be viewed
viduals or cases. Although there is some favorably by others. In an experiment,
SAMPLING ERROR in any group under for example, it manifests as the social
srady, the theory accounts for the fact desirability response set, which is the
that this error tends to be greatest in tendency of participants to give answers
smaller samples. that are in accordance with social norms
or the perceived desires of the researcher
Smirnov test see KOLMOGOROV-SMIRNOV rather than genuinely representative of
GOODNESS-OF-FIT TEST; KOLMOGOROV- their views. This is a CONFOUND to be
SMIRNOV TWO-SAMPLE TEST. controlled for in certain research, as it
smoothed curve a graphical represen- often reduces the validity of interviews,
tation of the relationship between two questionnaires, and other self-reports.
variables that has been adjusted to elimi- social i n d i c a t o r any fearare of a soci-
nate erratic or sudden changes in SLOPE, ety that can be measured over time and
so that its fundamental shape and direc- is presumed to reveal some underlying
tion will be evident. Also called smooth aspect of social reality and quality of life.
curve. For example, the retail price index is
smoothing n. a collection of techniques used as a measure of inflation, which in
used to reduce the irregularities (RAN- turn is taken as a key indicator of eco-
DOM VARIATION) in a data set or in a plot nomic performance. Other commonly
(curve) of that data, particularly in TIME- used indicators are derived from unem-
SERIES analyses, so as to more clearly see ployment figures, per capita income,
the underfying trends. The use of a MOV- poverty levels, labor conditions, hous-
ING AVERAGE is one example of smooth- ing costs, mental health, general health
ing. See SMOOTHED CURVE. and mortality data, nutrition informa-
tion, pollution levels, crime rates, educa-
Snedecor's F d i s t r i b u t i o n see F DIS- tion levels, opportunities for leisure and
TRIBUTION. [George W. Snedecor (1881- recreation, and the status of the elderly.
1974), U.S. statistician] Researchers and policymakers fre-
quently use social indicators to assess
s n o w b a l l s a m p l i n g a technique to the extent to which a society is "pro-
identify and recruit candidates for a gressing" as well as make predictions
study in which existing participants rec- about its furare.
ommend additional potential partici-
pants, who themselves are observed and socially sensitive research research

341
social science

on topics likely to evoke controversy in with arrows indicating the direction of

the community or strong emotional re- relationships. In most cases the diagram
sponses from participants. Such topics is organized into a meaningful pattern
would include those that have ethical by placing those individuals who are
implications affecting subgroups or cul- most frequently chosen (stars) in the
tures within society (e.g., ethnic minori- center of the diagram and the isolates
ties) or that involve potential costs and about the periphery. The method also
consequent problems for the partici- yields various indices of group structure
pants, investigators, or sponsors. For ex- and group cohesion, including choice
ample, a study that examines the status (the number of times a person is
relative merits of day care for infants ver- chosen by the other group members), re-
sus full-time care by the mother can jection status (the number of times a
have broad social implications and thus person is rejected by others), the relative
be considered socially sensitive. number of mutual pairs in a group, and
so on. sociometric adj.
social science 1. any of a number of
disciplines concerned with the common soft data subjective data that lack the
elements and coUective dimensions of rigor of HARD DATA. Soft data usually re-
human experience, sradied from a scien- sult from informal collection methods,
tific and research perspective. These disci- such as those lacking RANDOM ASSIGN-
plines traditionally have included anthro- MENT to conditions, those lacking for-
pology, economics, geography, history, mal RANDOM SAMPLING, or those based
linguistics, political science, psychiatty, only on anecdote. Soft data may be de-
psychology, and sociology, as well as asso- scriptive or qualitative and are used to
ciated areas of mathematics and biology. help interpret hard data.
Additional fields include related psycho- Solomon four-group design an ex-
logical studies in business administration, perimental design that assesses the effect
journalism, law, medicine, public health, of having been pretested on the magni-
and social work. The focus of analysis tude of the treatment effect. Participants
rangesfromthe individual to instirations are randomly divided into four groups
and entire social systems. The general goal and each group experiences a different
is to understand social interactions and to combination of experimental manipula-
propose solutions to social problems. 2. tions: The first group (A) receives the
these disciplines collectively. pretest, the treatment, and the posttest;
the second group (B) receives only the
sociometry n. a field of research in
treatment and posttest; the third group
which various techniques or sociometric
(C) receives the pretest, no treatment,
measures are used to analyze the pat-
and a posttest; and the fourth group (D)
terns of intermember relations within receives only a posttest. The major ad-
groups and to summarize these findings vantages of the Solomon four-group
in mathematical and graphic form. In design over a traditional two-group PRE-
most cases researchers ask the group mem- TEST-POSTTEST DESIGN are that it re-
bers one or more questions about their duces the influence of CONFOUNDS and
fellow members, such as "Whom do you that it can pinpoint whether changes in
most like in this group?", "Whom in the the dependent variable are due to some
group would you like to work with the INTERACTION EFFECT between the pre-
most?", or "Whom do you like the test and the treatment. A major disad-
least?". These choices can then be sum- vantage, however, is that its analysis and
marized in a sociogram, in which each statistics are complex. [Richard L. Solo-
rnember is represented by a numbered or mon (1919-1992), U.S. psychologist]
lettered symbol and the various choices
are identified by lines between them Somers's d (Somers's dg^) a NONPARA-
342
 spatial autocorrelation

METRIC TEST of whether there is an asso- SERIES ANALYSIS, where the purpose is to
ciation between two ORDINAL VARI- identity the nature of the phenomenon
ABLES when the data are presented in a represented by a sequence of measure-
C O N T I N G E N C Y T A B L E . T h e foCUS o f t h e ments, typically taken at successive
analysis is on the number of agreements points in time under the assumption
and disagreements in the set of data. For that the sequence is not random, and to
example, suppose a researcher interested forecast or predict future values. In a
in the relationship between age and atti- STARIMA model, the value of each ob-
tudes conducted a srady in which the servation in the series depends, at least
former were measured as ranges (e.g., in part, on the value of one or more of
20-25 years, 26-30 years, etc.) and the the immediately preceding observa-
latter were measured as good, neutral, or tions, and each observation in the series
poor. Somers's d, which ranges from -1 is predicted by a linear combination of
to +1, could then be used to obtain a prior observations as well as RANDOM
measure of association between the two. ERROR. When a variable, such as the
Also called Somers's delta. [Robert H. number of unemployed, is plotted over
Somers] tirhe there are likely to be considerable
seasonal or cyclical components to the
sorting test a format for assessing the variation that make it difficult to see the
ability to conceptualize, often used in underlying pattern. These components
adult neuropsychological assessments or can be eliminated by taking a suitable
in determining a child's level of cogni- moving average, which reduces random
tive development. The participant is fluctuations and makes long-term trends
asked to arrange an assortment of com- clearer.
mon objects by categoty. A common ex-
ample is the Wisconsin Card Sorting space-time autoregressive moving-
Test, in which the participant is asked to average model (STARMA model)
match a set of cards depicting shapes of an extension of the AUTOREGRESSIVE
differing color, quantity, and design; he MOVING-AVERAGE MODEL to include ad-
or she is given no matching rales but is ditional spatial and temporal dimen-
told whether a particular match is cor- sions in the exploration and forecasting
rect or incorrect. During the course of of relationships in a TIME SERIES. For
the test the matching rules are changed example, a researcher could apply a
and the time taken for the participant to STARMA model to predict traffic flow at
learn the new rules and the mistakes several adjacent intersections in a neigh-
made during this learning process are borhood using data gathered through-
analyzed to arrive at a score. out a particular year. There are likely to
soundproof r o o m a room designed to be multiple associations among multiple
balance the absorption of sound with its variables at multiple points in time that
controlled reflection. Mostly used for must be accounted for in such a situa-
the study of acoustic issues, soundproof tion, which can best be understood
rooms eliminate unwanted outside noise using a STARMA model.
and ensure sounds are sufficiently con-
spatial autocorrelation a measure-
fined to and evenly distributed through-
out the space within. , ment of a variable's association with
itself throughout space. Although statis-
space-time autoregressive inte- tical approaches often assume that mea-
grated moving-average model sured outcomes are independent of each
, (STARIMA model) an extension of the other, this may not be true for observa-
AUTOREGRESSIVE INTEGRATED MOVING- tions made at different locations. For ex-
AVERAGE MODEL to STATIONARY PRO- ample, measurements made at nearby
CESSES. The model is used in TIME- locations may be closer in value than

343
spatial data

measurements made at locations farther changes in test size. Also called Spear-
apart. Spatial autocorrelation thus pro- man-Brown prediction formula.
vides an index of the similarity of ob- [Charies Edward Spearman (1863-1945),
jects within an area, the level of inter- British psychologist and psychometri-
dependence between the variables, and cian; W. Brown, 20th-cenraty British
the nature and strength of that interde- psychologist]
pendence. Suppose one wanted to srady
Spearman correlation coefficient
the'relationships among cultures. In this
research the distance between societies (symbol: r^; p) a nonparametric measure
could be either spatial (based on phy- of statistical dependence between two
sical distance) or cultural (based on lan- variables that were measured on an OR-
DINAL SCALE; that is, the individual ob-
guage). Thus, spatial autocorrelation
models would examine the relationship servations (cases) can be ranked into two
among values of a single variable that ordered series. The Spearman correla-
arises from the geographic arrangement tion coefficient assesses how well the re-
of the areas in which these values occur. lationship between the variables can be
Spatial autocorrelation may be positive described using a MONOTONIC function.
or negative. Positive spatial autocorre- It ranges in value from +1 to -1. Also
lation occurs when similar values occur called Spearman's rank correlation
near one another, whereas negative spa- coefficient; Spearman's rho. [Charles
tial autocorrelation occurs when dissim- Spearman]
ilar values occur near one another. Spearman footrule a simpler version
When no statistically significant spatial of the SPEARMAN CORRELATION COEFFI-
autocorrelation exists, the pattem of spa- CIENT: a procedure for measuring rank
tial distribution is considered random. association that assesses the absolute dif-
See also AUTOCORRELATION. ference (D) between pairs of values. For
example, one supervisor ranks a set of
spatial data any observations or mea-
employees as 1, 2, 3, and 4, respectively,
surements with a direct or indirect refer-
on overall performance while another
ence to a specific location or geographi-
supervisor rates the same set of employ-
cal area.
ees as 2, 3,1, and 4, respectively. The ab-
spatial experiment a research design solute sum of the differences for the sets
that studies variables with respect to their of ranks is obtained by subtracting one
physical distance or separation from one rank in a pair from the other, disregard-
another. For example, a spatial experi- ing the sign of the difference, and add-
ment might examine the rates of mal- ing the values together. Thus, [1 - 2 = 1;
treatment referrals for children of a 2-3 = l;3-l=2;4-4 = 0], yielding a D
specific age across different residential = 4. The lower the absolute value of D,
areas. the closer the relationship between the
two sets of data. [Charles Spearman]
Spearman~Brown prophecy for-
m u l a the mathematical formulation of Spearman's r h o see SPEARMAN CORRE-
a basic tenet of CLASSICAL TEST THEORY LATION COEFFICIENT.
concerning the length (number of items)
speciesism n. discriminatoty, prejudi-
of a test and its influence on reliabiUty,
cial, or exploitative practices against
whereby increasing the number of items
nonhuman animals, often on the basis
with similar content results in increased of an assumption of human superiority.
reliabiUty for the test; similarly, decreas- See also ANIMAL RIGHTS. speciesist
ing the number of items leads to de- n., adj.
creased reliability. The formula allows a
researcher to estimate the gains or losses specification error in REGRESSION
in reliability that would occur with ANALYSIS, atypeof error that may occur
 spherical data

when converting a theory into a regres- specific variance in FACTOR ANALYSIS,

sion model, which involves selecting an the systematic variability that is specific
appropriate functional form for the to a particular variable and not shared
model and choosing which variables to with other variables. Specific variance
include. In specification error, an INDE- contrasts with COMMUNALITY, the vari-
PENDENT VARIABLE becomes correlated ability that a single variable shares with
with the ERROR TERM through use of an one or more of the other variables in the
incorrect functional form; through analysis, and RANDOM ERROR, the un-
omitting a variable from the model that systematic variability specific to a partic-
has a relationship with both the DE- ular variable.
PENDENT VARIABLE and one or more of
the independent variables; or through specimen record a data collection
including an irrelevant variable in the strategy in which a trained observer doc-
model. If an estimated model is uments the activities of a particular indi-
misspecified, it will be biased and incon- vidual within a particular context for a
sistent. specific period. For example, a specimen
record of a child's gross motor skills might
specification search a strategy used involve creating a detailed description of
.to find a statistical model that correctly the child's behavior while watching him
describes a population of interest. The or her play on recreational equipment
researcher first constructs artificial data for 15 minutes.
for which there is a known correct
model, then fits a misspecified model to spectral analysis in TIME-SERIES ANALY-
the data (see SPECIFICATION ERROR), and SIS, a strategy used to decompose a com-
finally adjusts this as necessary until the plex cycle of data into a small number of
"search" yields the combination of vari- underlying sine and cosine functions of
ables that best explains the phenome- different frequencies. For example, if the
non under study. The likelihood of cycle to be studied is one year12
success in a specification search is opti- monthly collections of data on the same
mal when (a) the investigator's initial variablespectral analysis examines the
model corresponds closely to the true fluctuations that occur over that 12-
model, (b) the search is allowed to con- month period and then determines which
tinue even when a statistically plausible appear to be particularly strong or im-
model is obtained, (c) the investigator portant. Also called spectrum analy-
can place valid restrictions on permissi- sis.
ble modifications, and (d) a large sample
spectral density f u n c t i o n a mathe-
is used.
matical description of how the periodic
specific f a c t o r in FACTOR ANALYSIS, a variation in a TIME SERIES may be ac-
LATENT VARIABLE that is significant only counted for by cyclic components at dif-
to a single MANIFEST VARIABLE. In con- ferent frequencies. The procedure for
trast, a common factor pertains to multi- estimating the spectral densities at vari-
ple manifest variables. ous frequencies is called SPECTRAL ANAL-
YSIS.
specificity n. 1. the quality of being
unique, of a particular kind, or limited spherical data observations or mea-
to a single phenomenon. For example, a surements that involve direction, dis-
stimulus that elicits a particular response tance, or both in three dimensions.
is said to have specificity. 2. the proba- Spherical data are most common in the
bility that a test yields a negative diagno- earth sciences but are also analyzed in
sis given that the individual does not neuroscience. For example, investigations
have the condition for which he or she is of how the nervous system processes au-
being tested. Compare SENSITIVITY. ditoty information to create spatial rep-

345
sphericity

resentations may involve quantitying INDEPENDENT VARIABLE (X) and a DE-

the direction and distance from an ob- PENDENT VARIABLE (y), in which values
server of a sound presented to him or of y vary sharply for different intervals of
her. X. Each range of x thus has a different
SLOPE, and spline functions are formed
sphericity n. an assumption, encoun-
by joining piecewise polynomials at
tered in a WITHIN-SUBJECTS ANALYSIS OF
fixed points called knots. An essential
VARIANCE of data obtained from the
feature of a spline function is that it is
same individuals on multiple occasions,
continuous; in other words, it has no
requiring the variations among each in-
breaks on the boundaries between two
dividual's set of scores to be equal or the
adjacent intervals. Spline functions (or
correlations among all time points to be
splines for short) are useful for fitting
constant. Results from analyses of vari-
data that have random components and
ance that violate sphericity require ad-
are widely used for interpolation and ap-
justments, such as the Greenhouse-
proximation of data sampled at a dis-
Geisser correction, the Huynh-Feldt cor-
crete set of points (e.g., for TIME-SERIES
rection, or the lower-bound correction,
interpolation).
to compensate for an increased propen-
sity of the researcher to draw invalid spline regression a nonparametric
conclusions by making a TYPE II ERROR. technique in which a SPLINE FUNCTION
See WITHIN-SUBJECTS DESIGN. is used to model an x variable that has
CATEGORICAL DATA.
sphericity test see BARTLETT TEST OF
SPHERICITY; MAUCHLY'S SPHERICITY split-ballot technique a procedure in
TEST. which a sample is randomly divided into
halves and each half receives a slightly
spike n. in a TIME-SERIES ANALYSIS, a different version of a questionnaire or
point in time at which there is a sharp survey designed to measure the same
increase followed by a rapid decrease in construct. The technique can be used to
measurements of the DEPENDENT VARI- determine whether different versions of
ABLE.
the survey have different outcomes (e.g.,
SpjotvoU-Stoline test in an ANALYSIS whether position of a particular item on
OF VARIANCE, a Statistical examination the questionnaire makes a difference in
of POST HOC COMPARISONS among the outcomes obtained from the two ad-
means for groups that contain an un- ministrations). Also called split-half
equal number of cases (observations). method; split-sample test.
The NULL HYPOTHESIS is that there are Split-half r e l i a b i l i t y a measure of the
no pairwise comparisons that are signifi- internal consistency of surveys, psycho-
cantly different; the ALTERNATIVE HY- logical tests, questionnaires, and other
POTHESIS is that at least one pair of
instruments or techniques that assess
means is different. The SpjotvoU-Stoline participant responses on particular con-
test is an extension of TUKEY'S HON- structs. Split-half reliabiUty is deter-
ESTLY SIGNIFICANT DIFFERENCE TEST,
mined by dividing the total set of items
which is restricted to cases in which (e.g., questions) relating to a construct of
each sample mean is based on an equal interest into halves (e.g., odd-numbered
number of observations. [Emil Spjotvoll and even-numbered questions) and
(1940-2002), Norwegian mathematician comparing the results obtained from the
and statistician; Michael R. Stoline, U.S. two subsets of items thus created. The
statistician] closer the correlation between results
spline f u n c t i o n a smoothed mathe- from the two versions, the greater the
matical representation of a disjointed or INTERNAL CONSISTENCY of the survey or
disaggregated relationship between an instrument. The RELIABILITY of the total

346

survey can be determined by applying
the SPEARMAN-BROWN PROPHECY FOR-
MULA.
Split-plot design a variation of a full
FACTORIAL DESIGN in which One of the
INDEPENDENT VARIABLES is held Constant
while all other combinations of condi-
tions are examined, often using different
sample sizes or different randomization
schemes. For example, consider a re-
searcher examining the influence on crop
yield of four different types of corn seed,
three differenttypesof fertilizer, and two
different types of planting technique. If
each planting technique requires its
own specialized set of expensive equip-
ment, such that it is not financially fea-
sible to provide all of the equipment to
all participants, the investigator instead
could have half of the participating
farmers plant all of the seed types in a
random order using one technique and
the other half plant all of the seed types
in a different random order using the
second technique. Indeed, split-plot
designs are particularly common in agri-
cultural and industrial contexts, in
which certain conditions may be dif-
ficult to manipulate or change for ex-
perimental purposes. Data from such de-
signs may be examined with a split-plot
analysis of variance.
split-sample test see SPLIT-BALLOT
TECHNIQUE.
s p r e a d n. see DISPERSION.
s p u r i o u s c o r r e l a t i o n a situation in
which variables are associated through
their common relationship with one or
more other variables but not through a
causal mechanism. For example, assume
that the data show a relationship be-
tween the total amount of loss in a fire formations generally, the goal is to ob-
and the number of firefighters at the
scene. One cannot infer from this that
calling fewer firemen would lower loss:
There is a third variablethe initial size
of the firethat irifluences both the SS symbol for S U M OF SQUARES.
amount of loss and the number of S S A abbreviation for S M A L L E S T S P A C E
firefighters present. The main problem
347
SSA
with spurious correlations is that one
typically does not know what other vari-
able is influencing the obtained rela-
tionship. If, however, one has some
insight into possible other variables,
PARTIAL CORRELATIONS Can be used, to
control for their influence. See THIRD-
VARIABLE PROBLEM.
s p u r i o u s p r e c i s i o n a value or outcome
stated with more precision than can be
justified, considering the procedures used
to obtain it. The statistics cited in adver-
tising or journalism often show spurious
precision.
S Q R T abbreviafion for SEQUENTIAL PROB-
ABILITY RATIO TEST.
s q u a r e d c o r r e l a t i o n c o e f f i c i e n t see
COEFFICIENT OF DETERMINATION.
s q u a r e d m u l t i p l e c o r r e l a t i o n co-
e f f i c i e n t see COEFFICIENT OF MULTIPLE
DETERMINATION.
s q u a r e m a t r i x a M A T R I X that has the
same number of rows as columns. The
number of rows and columns deter-
mines the order of the matrix (e.g., 3
rows by 3 columns is an order of 3), and
any two square matrices of the same
order can be added and multiplied.
Square matrices are used in REGRESSION
ANALYSIS, FACTOR ANALYSIS, and Other
multivariate analytical techniques.
s q u a r e - r o o t t r a n s f o r m a t i o n a pro-
cedure for converting a set of data in
which each value, Xj, is replaced by its
square root, another number that when
multiplied by itself yields Xj. Square-root
TRANSFORMATIONS OftCn result In HO-
MOGENEITY OF VARIANCE for the differ-
ent levels of the I N D E P E N D E N T V A R I A B L E
(x) under consideration. As with trans-
tain data that more closely meet the
assumptions of a statistical procedure
that is.to be applied.
ANALYSIS.
SSCP

SSCP symbol for SUM OF SQUARES OF sign, all samples have a common ori-
CROSS-PRODUCTS. ginthey start at the same timeand
there is a single, terminal event (typi-
SSE symbol for sum of squared errors. See
cally, death). By contrast, in a staggered
ERROR SUM OF SQUARES. entry design each participant begins the
S-shaped curve see OGIVE. study at a particular time zero and has an
ending point corresponding to the
SSregression Symbol for REGRESSION SUM
length of time that he or she was part of
OF SQUARES.
the study (i.e., until he or she experi-
Stability n. see STATISTICAL STABILITY. enced the event of interest or otherwise
ceased participation). For example, con-
stability coefficient an index of RELI- sider a study of a new treatment method
ABILITY determined via a test-retest
implemented at a specific hospital: There
method, in which the same test is ad- will be patients who survived over the
ministered to the same respondents at entire study period, others who survived
two different points in time. For exam- but entered the hospital after the study
ple, the stability coefficient of a psycho- had already begun, and still others who
logical test may be estimated by moved away and lost contact with the
determining the degree of similarity be- researcher. Although the latter two groups
tween participants' scores across time: of patients contributed only partial, in-
The more the two scores for each partici- complete information to the research
pant are alike, the higher the correlation (see CENSORED OBSERVATION), their data
between the two administrations and nonetheless should not be excluded from
the greater the stability coefficient of the the srady since those individuals are
test. A critical issue in using this strategy "survivors" and therefore reflect on the
is the interval of time between the two success of the new treatment method.
administrations. It should not be so long Staggered entry allows for increased pre-
that participants are likely to change on cision in estimating the SURVIVAL FUNC-
the construct being assessed nor so short TION.
that memory of responses on the initial
administration infiuences responses on Stahel-Donoho robust m u l t i v a r i -
the second administration (see PRAC- ate estimator in FACTOR ANALYSIS,
TICE EFFECT). Also called coefficient of DISCRIMINANT ANALYSIS, and Other sta-
stability. tistical techniques for examining multi-
stable rate 1. a constant change in the variate data, a strategy for handling
value of a DEPENDENT VARIABLE as. the OUTLIERS that weights each observation
value of an INDEPENDENT VARIABLE according to its overall "outlyingness."
changes. 2. in a REGRESSION ANALYSIS, The Stahel-Donoho estimator is neces-
the relative consistency of a BETA COEF- sary since the most common methods
FICIENT over time and across popula-
for identitying averages and variances,
tions, conditions, and so forth. the sample MEAN arid the sample CO-
VARIANCE MATRIX, are sensitive to out-
stacked bar g r a p h see COMPONENT liers and thus tenable only when the
BAR GRAPH. data foUow a NORMAL DISTRIBUTION.
[Werner Stahel, Swiss statistician; David
stage s a m p l i n g see MULTISTAGE SAM-
Donoho (1957- ), U.S. statistician]
PLING.

staggered entry design in SURVIVAL stalactite plot a graphical display of

ANALYSIS, a research design that allows the pattern of extreme values found in a
for the entry or exit of participants at set of numbers during M I N I M U M VOL-
any time throughout the course of the UME ELLIPSOID or other search proce-
study. In a typical survival analysis de- dures specifically used to detect masked
348
 standard error

Observation Number
1 2 3 4
1234567890123456789012345678901234567890
6 XX X X X XX X X X XX X X
8 XX X X X XX X X X XX X X
10 XXX X X X XX X X X XX X X
12 X X X X X X X XX X X X X XX
14 XXX X X X X X XX X X X XX X X
16 XXX X X X X X X XX X X XX X X
18 XXX X X X X X X XX X X XX X X
20 XXX X X X X X X XX X X XX X
22 XXX X X X XX X X XX X
0) 24 XXX X X X X X X XX X X XX X
N 26 XXX X X X X X X XX X X XX
28 XXX X X X X X XX X X XX
vi
4-1 30 XXX X X X X X X XX XXX
0) 32 XX X X X X X
sqn

34 X X X X X X X X XXX
36 X X X X X XX
(0 38 X X X X X X X X XX
40 X X X X X XX
42 XX X XX
44 X X X X X X
46 X X
48 X X
50 X

stalactite plot

OUTLIERS in MULTIVARIATE data. In the mean value X, the sample standard de-
hypothetical example above, xs are used viation is taken to be
to denote those observations that can be
V[i(x,- x)2/(-i)],
considered outliers within each subset
size, with observation 23 emerging as that is, the divisor is (n -1) rather than n.
the most extreme outlier. See also ROOT MEAN SQUARE.

standard Cox regression model see standard error (symbol: S) in statisti-

c o x REGRESSION ANALYSIS. cal analysis, a quantification of the in-
herent inaccuracy of a calculated POP-
standard deviation (symbol: 5D) a ULATION value that is attributable to
measure of the variability of a set of randomflucraationswithin the SAMPLE
scores or values, indicating how nar- data upon which it is based. Some degree
rowly or broadly they deviate from the of imprecision is present whenever a
MEAN. A small 5D indicates data points value for a large group (the population)
that cluster around the mean, whereas a is estimated by studying a subset of that
large standard deviation indicates data group (the sample), and the standard
points that are dispersed across many error provides a numerical description of
different values. It is expressed in the that variability. It is expressed as the
same units as the original values in the STANDARD DEVIATION of the SAMPLING
sample or population, so that the SD of a DISTRIBUTION. For example, the SAMPLE
series of measurements of weight would MEAN is the usual estimator of a POPULA-
be in pounds, for example. The 5p is TION MEAN yet different samples drawn
equal to the square root of the VARIANCE. from that same population nonetheless
If a population of n values has a mean p, wUl yield different values for the mean.
then the SD is Thus, to determine how much sample
variability exists the STANDARD ERROR
V[2:(x,.-p)%] OF THE MEAN may be obtained by taking
For a sample of the population, with a the standard deviation of all of the

349
standard error of estimate

means over all of the samples taken. any given measurement opporranity
Standard error is expressed in units represents his or her TRUE SCORE. It is
given in the same scale of measurement one of several types of STANDARD ERROR
that was used for the sample data (e.g., that may be calculated.
for a set of means that are given in
standard error o f p r e d i c t i o n see
weight the standard error unit is also a
S T A N D A R D E R R O R OF E S T I M A T E .
weight). The more samples involved in
determining the standard error, the standard error of the difference
smaller its value; the smaller the stan- an index of the degree to which statisti-
dard error, the more reliable the calcu- cal distinctions between two SAMPLES
lated population value. from a larger POPULATION reflect real
disparities between those samples. For
standard error of estimate (symbol: example, suppose a researcher analyzes
S) for a relationship between two vari- two samples, A and B, randomly drawn
ables (x and y) given by a REGRESSION from the same normally distributed
EQUATION, an index of how closely the source population and obtains a differ-
predicted value of y for a specific value of ent MEAN for Sample A than for Sample
X matches its actual value. If y' is an esti- B. The standard error of the difference
mated value from a regression line and y between the two means is used to deter-
is the actual value, then the standard mine whether those sample values truly
error of estimate is are different or instead reflect random,
V[Z(y-y')%], irrelevant dataflucraations.It is equal to

where n is the number of points. The MSD^/n,) + (SD'//7b)],

smaller the standard error of estimate, where SD^ = the population variance
the better the degree, of relationship (i.e., the square of the STANDARD DEVIA-
(CORRELATION) between x and y and the TION for the source population); = the
more confident one can be in the accu- size of Sample A; and b = the size of
racy of the estimated (predicted) y value. Sample B. It is one of several types of
It is one of several types of STANDARD STANDARD ERROR that may be calcu-
ERROR that may be calculated. Also lated.
called standard error of prediction.
standard error of the difference
standard error of measurement between t w o proportions an index
(symbol: SEM) an index of the RELIABIL- of the degree to which percentage dis-
ITY of an assessment instrament, repre- tinctions between two SAMPLES from a
senting the variation of an individual's larger POPULATION reflect real dispari-
scores across multiple administrations of ties between those samples. For exam-
the same test. A perfectly reliable instru- ple, suppose a researcher is interested in
ment will have a standard error of mea- whether the percentage of college gradu-
surement of 0, which means that an ates in Sample A (males) differs from the
individual will have the same score percentage of college graduates in Sam-
upon repeated testings with the instru- ple B (females). Both samples are ran-
ment. Otherwise, the standard error of domly drawn from the same normally
measurement will be between 0 and 1 distributed source population of all col-
and the individual will have different lege graduates from a particular univer-
scores on different occasions; the larger sity. The difference between the percent-
the error the greater the variation across ages of the two samples, A and B, is given
administrations. In essence, the stan- by
dard error of measurement provides an
indication of how confident one may be S = V([p*(l-p)*[(l/ni) + (l/')]l
that an individual's obtained score on where p is the pooled sample proportion

350
 standardized mean difference

of college graduates. See also STANDARD intended. The arrangement of scores ob-
ERROR OF THE DIFFERENCE. tained by such a standardization group
subsequently provides a point of com-
Standard error of the mean (sym- parison for the scores of other older
bol: SM; OM) a statistic that indicates adults who take the test. It is important
how much the average value (MEAN) for to note, however, that the standardiza-
a particular SAMPLE is likely to differ tion group must be representative of the
from the average value for the larger intended population of test takers in
POPULATION from which it is drawn. It is order to yield valid information. Also
equal to aHn, where a is the standard called norm group; standardization
deviation of the original distribution sample.
and n is the sample size. Less commonly
called standard error of the popula- standardized coefficient any index
tion mean. derived from an analysis of two or more
variables that have been transformed via
standard g a m m a d i s t r i b u t i o n see STANDARDIZATION, whlch ensures their
GAMMA DISTRIBUTION. value RANGES and VARIANCES are equiv-
Standardization n. 1. the process of alent and thus appropriate for compari-
establishing NORMS for a test. 2. the use son. The term is most commonly used to
of uniform procedures in test adminis- denote the standardized regression coef-
tration to ensure that all participants ficient or BETA COEFFICIENT.
take the same test under the same condi- standardized distribution a NOR-
tions and are scored by the same criteria, MAL DISTRIBUTION whose values have
which in turn ensures that results can be undergone TRANSFORMATION SO as to
compared to each other. 3. the transfor- have a MEAN of 0 and a STANDARD DEVI-
mation of data into a distribution of ATION of 1. Also called standard nor-
STANDARDIZED SCORES, often One hav- mal distribution; unit normal
ing a mean of 0 and a STANDARD DEVIA- distribution.
TION of 1, which produces derived'
measures of relative standing and allows standardized instructions direc-
comparison of raw scores from different tions for a measuring instrument that
distributions. The Z-SCORE TRANSFOR- are to be presented to all participants ex-
MATION is an example of standardiza- actly as prepared as part of the standard
tion. experimental or assessment procedure.
See STANDARDIZATION.
standardization group a sample
Standardized i n t e r v i e w see STRUC-
used to establish reliable norms for the
TURED INTERVIEW.
population that it represents. This is
done by analyzing the results of the test standardized mean difference a
administered to the sample and ascer- summaty statistic used in a META-
taining the average performance level ANALYSIS of studies that assess the same
and the relative frequency of each devia- outcome but measure it in a variety of
tion from the mean. The NORMAL DIS- ways. For example, many studies mea-
TRIBUTION thus created is then used for sure the constract of anxiety and its rela-
comparison with any specific furare test tionship to school performance, but they
score. For example, the standardization do so using different psychometric
group for a new test of computer literacy scales. In order to examine the anxiety-
in older adults might comprise a large school performance effects found across
set of test takers above age 60 whose these multiple studies, it is necessaty to
characteristics (e.g., sex, demographics, first standardize the results of each of the
ethnicity, race) reproduce those of the studies to a uniform scale and then com-
larger population for whom the test is bine the results. The standardized mean

351
standardized measure

difference therefore expresses the size of standardized test 1. an assessment in-

the effect in each srady relative to the strament whose VALIDITY and RELIABIL-
variability observed in that study; it is ITY have been established by thorough
the difference in mean outcomes be- empirical investigation and analysis. It
tween the groups divided by the STAN- has clearly defined norms, such that a
DARD DEVIATION of the outcome among person's score is an indication of how
participants in the study, COHEN'S D and weU he or she did in comparison to a
HEDGES'S G are two commonly used large group of individuals representative
types of standardized mean difference. of the population for which the test is
Also called standardized mean ef- intended. Also called standardized
fect. measure. 2. an assessment instrument
administered in a predetermined man-
standardized measure see STAN- ner, such that the questions, conditions
DARDIZED TEST. of administration, scoring, and interpre-
standardized regression coeffi- tation of responses are consistent from
cient see BETA COEFFICIENT. one occasion to another.

standardized regression equation standard n o r m a l d i s t r i b u t i o n see

in REGRESSION ANALYSIS, a formula STANDARDIZED DISTRIBUTION.
showing the average change in the DE- Standard n o r m a l variable any ran-
PENDENT VARIABLE (y) that occurs with dom variable whose probable value fol-
one unit change in an INDEPENDENT lows a NORMAL DISTRIBUTION with a
VARIABLE (X) after each has been con- MEAN of 0 and a STANDARD DEVIATION
verted into a form whose distribution of 1. Also called standard normal de-
has a MEAN of 0 and a STANDARD DEVIA- viate; standard normal variate;
TION of 1. The use of a standardized re- unit normal variable.
gression equation allows for direct
comparison of the variables despite dif- standard observer in color vision re-
ferences in their measurement scales. search, a hypothetical typical human vi-
See also REGRESSION EQUATION. sual system that is described in terms of
mathematical functions and equations
standardized residual in a REGRES- relating its quantitative visual responses
SION ANALYSIS, the error in the DEPEND- to measurable physical statistics of light
ENT VARIABLE (y) score not explained by stimuli. The descriptions are therefore
the INDEPENDENT VARIABLE (X) when psychophysical and used to achieve a
the X and y variables have undergone shared technical description of the per-
STANDARDIZATION. ceptual effects of Ught stimuli on human
observers. The equations that define the
standardized score a value derived standard observer are based on averages
from a raw score by subtracting the of laboratoty measurements of the vi-
mean value of all scores in the set and di- sual responses of human participants to
viding by the STANDARD DEVIATION of particular light stimuU under particular
the set. The advantage of standardized viewing conditions.
scores is that they are not reflective of
the units of the measuring device from standard stimulus an item used as the
which they were obtained and thus can basis of comparison in the quantitative
be compared to one another regardless investigation of physical stimuli and the
of the device's scale values. Several types sensations and perceptions they produce.
of standardized score exist, including For example, in the method of adjust-
STANINES, T SCORES, and z SCORES. Also ment, a participant may be presented
called normal score; standard score. with a sound of a particular intensity
See also STANDARDIZATION. (the standard stimulus) and asked to

352
 statistical conclusion validity

change the intensity of another sound TION strucrare remain constant over
to match. time. Most statistical forecasting meth-
ods are based on the assumption that
standard treatment c o n t r o l in a the time series of interest can be mathe-
CLINICAL TRIAL, atypeof CONTROL CON- matically transformed into approximate
DITION in which participants receive the stationarity (i.e., can be "stationarized").
treatment typically administered to a
group of participants with a given medi- Stationary d i s t r i b u t i o n a distribu-
cal problem. Another group of partici- tion describing ah invariant equilibrium
pants receives the experimental treat- state in a MARKOV CHAIN. The process
ment. After the trial, the outcomes of being modeled is associated with a sin-
the two groups are compared to see if the gle, time-independent MATRIX.
experimental treatment is better than, stationary process a STOCHASTIC
as good as, or worse than the standard PROCESS whose joint probability distri-
treatment. bution does not change when shifted in
stanine n. a method of scaling scores on time or space. In other words, the pa-
a nine-point scale that ranges from a low rameters of a stationary process, such as
of 1 to a high of 9, with a mean of 5 and a the mean and the variance, do not
STANDARD DEVIATION of 2. A Stanine is a change over time or position. See also
standard ninth, referring to the interval STATIONARITY.
used in dividing the results into nine statistic n. 1. a number measuring some
more or less equal parts. A stanine score characteristic, construct, variable, or
of 1, 2, or 3 is below average; 4, 5, or 6 is other item of interest. 2. any function of
average; and 7,8, or 9 is above average, 'A the observations in a SAMPLE that may
stanine is a type of STANDARDIZED be used to estimate the unknown but
SCORE and is mainly used with school corresponding value in the POPULA-
achievement tests. | TION. Examples include measures of
STARIMA model abbreviation for CENTRAL TENDENCY (e.g., the MEAN, ME-
SPACE-TIME AUTOREGRESSIVE INTEGRATED DIAN, MODE), measures of DISPERSION
MOVING-AVERAGE MODEL. (e.g., STANDARD DEVIATION, VARIANCE),
and distributional attributes (e.g., SKEW-
S T A R M A model abbreviation for NESS, KURTOSIS). Statistics often are as-
SPACE-TIME AUTOREGRESSIVE MOVING- signed Roman letters (e.g., M, s), whereas
AVERAGE MODEL. ] the equivalent values in the population
(called PARAMETERS) are assigned Greek
State space 1. multidimensional space,
letters (e.g., p, a).
particularly as related to the depiction of
the results of classification methods that statistical analysis any of a wide
group objects with similar characteris- range of techniques used to describe, ex-
tics and patterns of behavior. 2. in sys- plore, understand, explain, predict, and
tem behavior, an abstract representation test HYPOTHESES about data. It involves
of the potential states of the system that the examination of data collected from
emerges from the complex interactions SAMPLES within POPULATIONS as well as
of the variables that make up the system. the use of probabilistic models to make
The interactions are based on a set of inferences and draw conclusions.
nonlinear relations.
statistical conclusion v a l i d i t y the
stationarity n. in TIME-SERIES ANALY- degree to which the conclusions drawn
SIS, the property of being unchanging or from statistical analyses of data are accu-
"flat," such that the data are without rate and appropriate. In other words, sta-
trends or periodic fluctuations and the tistical conclusion validity addresses
MEAN, VARIANCE, and AUTOCORRELA- whether inferences about relationships

353
statistical control

(i.e., whether the independent variable given decision. Specific equations are
and dependent variable covary and, if used to calculate the degree of loss asso-
so, how strongly) are reasonable or not, ciated with each course of action in
given the observed data. It is related to order to determine the most advanta-
but distinct from INTERNAL VALIDITY, geous choice.
which is concerned with the causality of
statistical dependence the situation
the relationship. Statistical conclusion
in which the conditional probabUity of
validity is enhanced when there is good
one event given another event is greater
statistical POWER, RELIABILITY of mea-
than the probability of that flrst event.
sures, and use of good experimental
In other words, knowing information
methods and procedures. Conversely, it
about one variable provides information
is threatened by such factors as (a) viola-
about the other. Compare STATISTICAL
tions of the ASSUMPTIONS embedded in
INDEPENDENCE. See also DEPENDENT
a statistical test; (b) problems associated
EVENTS.
with the EXPERIMENT-WISE ERROR RATE;
(c) RESTRICTION OF RANGE; (d) use of in- statistical determinism the position
appropriate EFFECT-SIZE measures; and that the laws of probability can predict
(e) extraneous variation in the experi- the likely number of events of a given
mental setting. kind that will occur in a given popula-
tion under certain defined conditions
statistical c o n t r o l the use of statistical (e.g., the number of coin tosses per 1,000
procedures to remove the influence of a that will be heads; the number of male
particular factor that could not be elimi- Americans age 70 who will die in a 12-
nated or controlled by the experimental month period). This is analogous to, but
design in order to better analyze the rela- logically distinct from, physical deter-
tionship between two variables. For ex- minism, which is the proposition that all
ample, the relationship between age (x) eventsincluding human behaviors
and income earned (y) could be influ- are caused by prior events, conditions,
enced by a third variable, years of educa- and natural laws.
tion (z). Thus, if a researcher did not first
remove the effects of education he or she statistical difference see SIGNIFI-
might derive erroneous conclusions CANT DIFFERENCE.
about the influence of age on income statistical error see ERROR.
from his or her analysis. One type of sta-
tistical control is PARTIAL CORRELATION, statistical evaluation see STATISTI-
which shows the association between CAL TEST.
two quantitative variables after statisti- statistical hypothesis a research
cally controlling for one or more extra- question posed in a statistically testable
neous variables. A second type of form. For example, if a researcher is in-
statistical control is ANALYSIS OF CO- terested in whether one treatment leads
VARIANCE, a technique that shows the to a more positive outcome than an-
relationship between a categorical inde- other treatment, he or she could reframe
pendent variable and a quantitative the question in terms of mean differ-
dependent variable after statistically ences, such that the NULL HYPOTHESIS is
controlling for one or more extraneous 0 (no difference between treatments)
variables. and the ALTERNATIVE HYPOTHESIS is not
0 (the difference between treatments is
statistical decision theory a branch
greater or lesser than zero). See also HY-
of statistical science concerned with the
POTHESIS TESTING.
use of data to arrive at decisions. It fo-
cuses upon identifying the values, un- statistical independence the condi-
certainties, and other issues relevant to a tion in which the occurrence of one

354
 statistical table

event makes it neither more nor less participants might be required than in
probable that another event will occur. alternative strategies.
In other words, knowing information
statistical significance the degree to
about one variable provides no informa-
which a research outcome cannot rea-
tion about the other variable. Compare
sonably be attributed to the operation of
STATISTICAL DEPENDENCE. See also IN-
chance or random factors. It is deter-
DEPENDENT EVENTS.
mined during SIGNIFICANCE TESTING
statistical inference see INFERENCE. and given by a critical p value, which is
the probability of obtaining the ob-
statistical model a formal description served data if the NULL HYPOTHESIS (i.e.,
of the relationships between two or of no significant relationship between
more variables in the form of a mathe- variables) were true. Significance gener-
matical equation. It is statistical in that ally is a function of sample sizethe
the variables are related in a STOCHASTIC larger the sample, the less likely it is that
rather than a deterministic manner, one's findings will have occurred by
with each set of possible observations on chance. See also CLINICAL SIGNIFICANCE;
a variable linked to a set of probability PRACTICAL S I G N I F I C A N C E ; SIGNIFICANCE
distributions. Many STATISTICAL TESTS LEVEL.
involve comparing a particular model
with the observed data. Statistical significance testing see
SIGNIFICANCE TESTING.
statistical power see POWER.
statistical s t a b i l i t y consistency of re-
statistical p r e d i c t i o n the process of sults across samples, srady designs, and
using correlations between variables to analyses. A META-ANALYSIS may be used
hypothesize about future events and to examine the stability of means, COR-
outcomes. For example, a university ad- RELATIONS, and other PARAMETER esti-
ministrator may use a REGRESSION mates obtained from different studies of
EQUATION to predict a student's coUege the same population. See also STABILITY
grade point average with reasonable ac- COEFFICIENT.
curacy from measures of performance in
statistical surveillance continuous
h i g h school, such as scores o n tests a n d
observation and analysis of a set of data
f i n a l grades i n classes.
with the goal of detecting any important
Statistical procedure see STATISTICAL changes (e.g., in the direction of a pro-
TEST. cess) as soon as they occur. For example,
a statistical surveillance system to deter-
statistical psychology the applica- mine the end of the influenza season
tion of statistical methods and models to might rely upon NONPARAMETRIC meth-
organize, summarize, and interpret data ods to identify the decline in incidence
so as to derive descriptions and explana- that signals the season's end.
tions of cognitive, behavioral, social,
and other psychological phenomena. statistical table any table of values
used by a researcher to plan experiments
statistical sequential test any statis- or interpret results. For example, there
tical procedure in which the sample size are tables of CRITICAL VALUES in HY-
is not fixed in advance but rather the POTHESIS TESTING, tables to estimate the
process of coUecting more data (more sample size needed to obtain a desired
observations or more participants) is level of POWER, and tables to convert
continued until there is enough infor- one distribution to another (e.g., from a
mation available either to accept or to correlation to a standardized value). A
reject the NULL HYPOTHESIS. A parficular CONTINGENCY TABLE Is often used in the
benefit of this approach is that fewer CROSS-CLASSiFiCATiON of items: For ex-

355
statistical test

ample, a 2 x 2 table might show the aver- on the dependent variable. Repeated
age scores on a test according to runs are therefore undertaken and the
individuals' gender (male and female) path of steepest ascent is identified via the
and education level (high school and use of a REGRESSION EQUATION. Where
college graduates). improvement is defined as the minimiz-
ing of a particular variable, then the
statistical test any mathematical tech- steepest descent will be sought in the
nique or procedure used to evaluate the same way.
correctness of an empirical hypothesis
by determining the likelihood of the stem n. the introductory part of a FIXED-
sample results occurring by chance. Sta- ALTERNATIVE QUESTION, often an in-
tistical testing will reveal the probability complete statement that the respondent
of committing a TYPE i ERROR if the NULL is asked to complete. It is followed by a
HYPOTHESIS is rejected. See HYPOTHESIS set of response options consisting of the
TESTING; STATISTICAL SIGNIFICANCE. correct answer and several plausible but
incorrect choices (DISTRACTORS).
statistical uncertainty the degree of
inaccuracy inherent to the process of es- stem-and-leaf plot a graphical me-
timating PARAMETERS from sampled thod of presenting data that have been
data. When researchers make such esti- measured on an INTERVAL SCALE. A basic
mates they generally assume that some stem-and-leaf plot comprises two col-
exact or TRUE SCORE exists and often umns separated by a vertical line; the
specify a CONFIDENCE INTERVAL or range right column lists the last digit of each
of values that they expect this'true value data point (the "leaves") and the left col-
to fall within. For example, an estimate umn lists all of the other digits from
of 8 2 means that the researcher is con- each data point (the "stems"). Each stem
fident that the actual parameter value is listed only once and no numbers are
lies between 6 and 10. The uncertainty is skipped, even if that means some stems
the experimenter's best estimate of how have no leaves. The leaves are listed in
far an estimated quantity might be from increasing order of magnitude in a row
the true value. to the right of each stem. For example,
consider the following hypothetical val-
statistics n. the branch of mathematics ues for participants measured on a par-
that uses data descriptively or inferen- ticular variable:
tially to find or support answers for sci-
entific and other quantifiable questions. 55 57 58 59 74 75 77 79 79 83 83 86 87
It involves various techniques and proce- 92 95 99 107
dures for recording, organizing, analyz-
The stem-and-leaf plot for these values is
ing, and reporting quantitative infor-
mation. See also DESCRIPTIVE STATISTICS; 5 15 7 8 9
INFERENTIAL STATISTICS; SUMMARY STA-
TISTICS. statistical adj. statisti- 6I
cian n. 714 5 79 9
steepest ascent when conducting a 8 13 3 6 7
study with repeated runs, that region of 91259
the INDEPENDENT VARIABLE in which
there is maximum response or improve- 101 7
ment on the DEPENDENT VARIABLE of Also called stem-and-leaf diagram;
interest. If research is being carried out stem-and-leaf display; stemplot.
in a new area, it is likely that initial levels
of the independent variable will not stepdown selection see BACKWARD
provide a satisfactory level of response ELIMINATION.

356
 stochastic process

stepup selection see FORWARD SELEC- and females to elicit donations from par-
TION. ticipants instead of using a single male
and a single female.
stepwise correlation see STEPWISE
REGRESSION. stimulus value 1. the strength of a
given stimulus, measured in standard
Stepwise d i s c r i m i n a n t analysis a units (e.g., a shock of 40 volts). 2. a theo-
type of DISCRIMINANT ANALYSIS in retical characteristic of a stimulus said to
which a model that can differentiate index its effectiveness as a reinforcer.
maximally between categories is built
step by step. In forward stepwise dis- stochastic adj. 1. random or undeter-
criminant analysis, all possible predictor mined; arising from chance. 2. describ-
variables are reviewed and evaluated at ing a system or process that follows
each step to determine which one con- a probability pattern, such that events
tributes most to the discrimination be- may be analyzed according to their sta-
tween groups. That variable is then tistical probability but not accurately
included in the model, and the process predicted.
starts again. In backward stepwise stochastic matrix a SQUARE MATRIX
discriminant analysis, all variables are in which all of the rows or all of the col-
initially included in the model and then, umns sum to 1. Arightstochastic matrix
at each step, the variable that contrib- is one in which the rows contain
utes least to the prediction of group nonnegative real numbers summing
membership is eliminated. The stepwise to 1, whereas a left stochastic matrix is
procedure is structured so that the re- one in which the columns contain
spective F-TO-ENTER and F-TO-REMOVE nonnegative real numbers summing to
values of each variable can be isolated, 1. See also DOUBLY STOCHASTIC
thereby determining its unique contri-
bution to the prediction of group mem- stochastic model a model in which
bership. one or more of the inputs allow for ran-
dom variation, thus generating a range
stepwise regression a group of RE- of potential outcome values. The ran-
GRESSION ANALYSIS techniques that dom variation is usually based on fluc-
enter predictor (independent) variables tuations observed in historical data. Sto-
into (or delete them from) the REGRES- chastic models are used to estimate the
SION EQUATION one variable (or block of probabilities of various outcomes occur-
variables) at a time according to some ring under varying conditions. They are
predefined criterion. It is contrasted widely used in the social and behavioral
with SIMULTANEOUS REGRESSION, whlch sciences and also in theflnancialworld.
enters all variables at the same time. Also Compare DETERMINISTIC MODEL.
called stepwise correlation; step-
wise multiple regression. stochastic process a random process: a
sequence of events with a random prob-
stimulus s a m p l i n g a procedure for in- ability pattern such that the occurrence
creasing the generalizability of research of any event in the sequence is inde-
results by using multiple stimuli within pendent of past events. For example, the
a categoty as representative of an experi- number of people in a doctor's office
mental condition, as opposed to select- who have colds during a one-month pe-
ing a single stimulus whose unique riod could be said to follow a stochastic
characteristics may distort results. For process. In contrast to deterministic pro-
example, a study investigating the ef- cesses, stochastic processes involve
fects of gender on monetaty generosity some indeterminacy, such that their de-
would demonstrate stimulus sampling if velopment over time may only be de-
it employed a variety of different males scribed by probability distributions. A

357
stochastic variable

MARKOV CHAIN is a stochastic process. Stratified sampUng often improves the

Also called discrete-time stochastic representativeness of the selected cases
process. by reducing SAMPLING ERROR. The pro-
portion of the sample to be selected
stochastic variable see RANDOM VARI- from each subgroup is decided in ad-
ABLE.
vance.
stooge n. a colloquial name for a CON- s t r a t u m n. (pl. strata) a layer (typically
FEDERATE. one of a number of parallel layers) in
stopping rule a criterion stated at the a structure, such as any of the sub-
outset of a research project for ending populations in SAMPLING. For example,
the study early, as when one treatment if the variable of interest is gender, the
clearly has been shown to be more effec- two subgroups (strata) would be male
tive than another. and female, and members of each sub-
group would be chosen separately for re-
strata pl. n. see STRATUM. search participation.
s t r a t i f i e d log-rank test a nonpara- s t r a t u m chart see BAND CHART.
metric statistical procedure used to com-
pare the time to an event in two samples streaky hypothesis see HOT-HAND HY-
POTHESIS.
where each sample has component lay-
ers or strata. For example, if there are strength o f association in statistics,
two preventative treatments to be com- the degree of relationship between two
pared according to the time from initial or more variables, as measured by such
use until a patient's heart attack, and indices as a CORRELATION COEFFICIENT,
low-risk, medium-risk, and high-risk in- COEFFICIENT OF DETERMINATION, ETA
dividuals are in each treatment group, SQUARED, or OMEGA SQUARED.
application of the stratified log-rank test
controls for the difference between treat- strength-of-effect index see EFFECT
SIZE.
ments as a function of risk. The stratified
log-rank test is appropriate when there stress test in MULTIDIMENSIONAL SCAL-
are CENSORED DATA or when the data ING, a measure that indicates the RELI-
have positive SKEWNESS. ABILITY or GOODNESS OF FIT of the
solution. The KRUSKAL STRESS test is an
s t r a t i f i e d s a m p l i n g the process of se-
example. Stress tests also are used to de-
lecting a sample from a population com- termine the number of dimensions to be
prised of various subgroups (strata) in included in the scaling solution.
such a way that each subgroup is repre-
sented. For example, in a study of col- strong inference an inductive argu-
lege students a researcher might wish to ment based on systematic, controlled in-
examine people from the different ma- vestigation. It involves a cyclic process
jors (e.g., social sciences, physical sci- of developing several different testable
ences, humanities). The selection pro- hypotheses, devising and carrying out
cedure within each of these strata may an experiment to exclude one or more
be random, or systematic. In stratified of these, and devising and testing ad-
random sampling, a chance process ditional hypotheses to refine the pos-
(e.g., a RANDOM NUMBER GENERATOR) is sibilities that remain. Compare WEAK
used to select individuals, whereas in INFERENCE.
stratified systematic sampling an objec-
tive, orderly procedure is applied to strong l a w o f large numbers see
choose individuals (e.g., listing all of LAW OF LARGE NUMBERS.
the students within each major alpha- s t r u c t u r a l c o e f f i c i e n t in STRUC-
betically and choosing evety 10th case). TURAL EQUATION MODELING, a measure

358
 structured observation

of the amount of change expected in an ships, one might ask each of 15 people
outcome or DEPENDENT VARIABLE given to indicate who among the others is a
a one-unit change in the causal or INDE- friend. This would yield a 15 x 15 con-
PENDENT VARIABLE and no change in tingency table or matrix, with 1 repre-
any other variable.- senting "is a friend" and 0 representing
"is not a friend." Although there would
structural equation modeling be empirical zeros in some cells to repre-
(SEM) any of a broad range of multi- sent not-a-friend responses, all entries
variate analysis methods, including FAC- along the diagonal of the matrix are
TOR ANALYSIS and PATH ANALYSIS, that
structural zeros since they represent one
examine VARIANCES and COVARIANCES
person's response with regard to himself
in order to find interrelationships or herself rather than another person.
among LATENT VARIABLES. For example,
assume a researcher states that job sat- stmcture coefficient 1. in REGRESSION
isfaction leads to happiness. Both are ANALYSIS, a CORRELATION COEFFICIENT
latent variables that are not directly indicating the extent of relationship be-
observable but are defined in terms tween scores on a particular MANIFEST
of other measurable variables, such as VARIABLE and scores on a latent or pre-
judgments of job performance from su- dicted variable derived from the entire
pervisors and peers; self-reports about set of variables. It is used to quantity the
attitudes toward pay, supervision, type importance of an observed variable. 2.
ofwork, and other job characteristics; re- in DISCRIMINANT ANALYSIS, a value rep-
sults from PSYCHOLOGICAL TESTS; and resenting the association between a par-
so forth. The researcher could use the ticular variable being considered as a
measurable data to generate an equation differentiator of group membership and
representing the strength and nature of the DISCRIMINANT FUNCTION scores de-
the links among the latent variables. rived from the full set of predictor vari-
Also called structural modeling. ables.

structural time-series model a RE- structured i n t e r v i e w a method for

GRESSION EQUATION in which the ex- gathering information, used particularly
planatory variables are functions of time in surveys and personnel selection, in
and the REGRESSION COEFFICIENTS asso- which questions, their wordings, and
ciated with the interval of time are al- their order of administration are deter-
lowed to vaty. For example, a researcher mined in advance. The choice of an-
could srady changes in emotional be- swers tends to be fixed and determined
havior as a person grows older. Struc- in advance as well. With stractured in-
tural time-series models are used not terviews, answers can be aggregated and
only to describe trends and patterns but comparisons can be made across differ-
also to FORECAST outcomes. ent samples or interview periods; inter-
viewees can be assessed consistently
structural v a l i d i t y see FACTORIAL VA- (e.g., using a common rating scale); and
LIDITY. ORDER EFFECTS, whereby the answer to a
particular question can depend on the
structural zero an entry in a table that
preceding questions, are minimized.
is certain to be zero because it corre-
Also called standardized interview.
sponds to an impossible outcome, as op- Compare UNSTRUCTURED INTERVIEW.
posed to an entry that has an empirical See also SEMISTRUCTURED INTERVIEW.
value of zero. The occurrence of zeros
may be inherent to the issue of interest structured observation a systematic
or, in some applications, introduced in- method of collecting behavioral data
tentionally through the experimental within a controlled environment, often
design. For example, in studying friend- used in research with infants and young

359
structured Q sort

children, in which observers measure DARDIZATION) between the observed

overt actions and interpersonal processes. value and the value predicted by the RE-
In stractured observation, researchers (a) GRESSION EQUATION when the potential
select which behaviors are of interest outlier is included, minus the same dif-
and which are not, (b) clearly define the ference when the potential outlier is
characteristics of each behavior so that excluded. A large Studentized deletion
observers all agree on the classification, residual (e.g., greater than 2) provides
and (c) note the occurrence and fre- evidence that the observation in ques-
quency of these targeted behaviors in tion is indeed an outlier biasing the
the situation under analysis. Observers results of the analysis. Also called Stu-
may monitor situations unobtrusively dentized deleted residual. [Student,
or from within as active participants (see pseudonym of WiUiam S. Gosset]
PARTICIPANT OBSERVATION). Structured
observation differs from NATURALISTIC Studentized m a x i m u m modulus
OBSERVATION, which involves observing
d i s t r i b u t i o n a distribution of p values
individuals outside of the laboratoty in (see PROBABILITY LEVEL) used when an
I N D E P E N D E N T V A R I A B L E haS m u l t i p l e
their own environments.
conditions (e.g., very high, high, average,
structured Q sort see Q SORT. low, and very low) that a researcher
wishes to compare two at a time (e.g.,
structured stimulus a well-defined, very high to high, very high to average,
weU-organized stimulus, the perception etc.). The Studentized maximum modu-
of which is influenced more by the char- lus distribution adjusts the p values
acteristics of the stimulus than by those needed to indicate significance in order
of the perceiver. Compare UNSTRUC- to minimize the possibiUty that a
TURED STIMULUS. PAIRWISE COMPARISON Will be found
Stuart test see COX-STUART TEST FOR significant simply because multiple
TREND. comparisons are being made. [Student,
pseudonym of William S. Gosset]
stub f a c t o r in FACTOR ANALYSIS, that
part of a FIRST-ORDER FACTOR that re- Studentized range d i s t r i b u t i o n a
mains after the variance due to a second- theoretical PROBABILITY DISTRIBUTION
order or other HIGHER ORDER FACTOR that is used in MULTIPLE COMPARISON
has been removed. TESTS when assessing differences be-
tween pairs of means for significance. It
Studentization n. a procedure to elimi- is similar to the T DISTRIBUTION but dif-
nate a NUISANCE PARAMETER in particu- fers in accounting for the number of
lar calculations. It transforms a statistic means under consideration: The more
whose distribution of probable values re- means being considered, the larger the
lies upon the unknown parameter into CRITICAL VALUE Of the STUDENTIZED
one whose distribution relies on quanti- RANGE STATISTIC must be in order to re-
ties that can be derived from the sample ject the NULL HYPOTHESIS. [Student,
data. Such transformed statistics are pseudonym of William S. Gosset]
described as "Studentized." [Student,
Studentized range statistic (sym-
pseudonym of William S. Gosset (1876-
bol: q) a statistic used in TUKEY'S HON-
1937), British statistician]
ESTLY SIGNIFICANT DIFFERENCE TEST and
Studentized deletion residual in Other MULTIPLE COMPARISON TESTS to
REGRESSION ANALYSIS, a Statistic com- establish CRITICAL VALUES for rejecting
puted to determine if any observations the NULL HYPOTHESIS. It is the RANGE of
or cases are significantly different from a sample expressed in units of the stan-
the others and thus can be considered dard distribution for that sample. [Stu-
OUTLIERS. It is the difference (after STAN- dent, pseudonym of WUliam S. Gosset]
 subjective method

Studentized residual in REGRESSION responses or performance are reported

ANALYSIS, a Standardized statistic de- or evaluated; less frequently, the subject
scribing the variation between obtained may also be an institution, group, or
and predicted values. It is calculated by other entity, PARTICIPANT is now often
obtaining the difference between the em- the preferred term for human subjects,
pirical value of an observation and the because the word subject is held to be de-
value of that observation predicted by a personalizing and to imply passivity and
model, and then dividing that difference submissiveness on the part of the
(called a RESIDUAL) by an estimated experimentee.
STANDARD DEVIATION. Studentlzed re-
siduals are used to identify REGRESSION subject a t t r i t i o n see ATTRITION.
EQUATIONS that are a poor fit for the ob- subject bias the influence of partici-
served data. [Student, pseudonym of pants' knowledge of the purpose of the
William S. Gosset] research upon their responses to experi-
Student's t d i s t r i b u t i o n see T DISTRI- mental conditions and manipulations.
BUTION. [Student, pseudonym of Wil- For example, a participant who knows
liam S. Gosset] he or she is in the treatment group as op-
posed to the control group might be-
Student's t test see T TEST. [Student, have differently than he or she would
pseudonym of William S. Gosset] otherwise. See DEMAND CHARACTERIS-
study n. a research investigation con- TICS.
ducted for the purpose of understand-
ing, explaining, describing, or predict- subject history 1. background infor-
ing some phenomenon of interest. It mation about a participant in a study,
may be conducted in the laboratoty or such as sociological, occupational, and
natural environment, and it may yield educational data. 2. events that a re-
quantitative or qualitative data. search participant experiences outside of
the controlled environment of the
study a r t i f a c t see ARTIFACT. study, which may threaten its INTERNAL
sturdy statistic see ROBUST ESTIMA- VALIDITY. See HISTORY EFFECT.
TOR. s u b j e c t i v e adj 1. taking place or existing
Sturges's rule a rule for determining only within the mind. 2. particular to a
the width of CLASS INTERVALS or bars on specific person and thus intrinsically in-
a HISTOGRAM, that is, for deciding the accessible to the experience or observa-
number of groups into which a distribu- tion of others. 3. based on or influenced
tion of observations should be classified. by personal feelings, interpretations, or
It is given as 1 + log2, where n is the prejudices. Compare OBJECTIVE.
number of observations. [Herbert A.
Sturges (1882-1958), mathematician] subjective error any systematic devia-
tion of an individual's OBSERVED SCORE
subgroup analysis the division of a from the TRUE SCORE that can be attrib-
sample into smaller groups for the pur- uted to individual variations in percep-
pose of evaluating differences among tion or particular interpretive biases
those subgroups. For example, if the held by that person.
variable of interest is gender, subgroups
of males and females could be created in subjective e x a m i n a t i o n see SUBJEC-
order to examine differences between TIVE TEST.
them on a dependent variable.
subjective method any of various ap-
subject (S) n. generally, the individual proaches to the collection, evaluation,
human or animal who takes part in an and interpretation of data that depend
experiment or research study and whose upon the researcher's or analyst's per-

361
subjective prior

sonal judgments, feelings, attitudes, or participant in a srady. It is an issue of par-

inraitions. ticular concern in DECEPTION RESEARCH,
as participants may significantly modity
subjective p r i o r see PRIOR DISTRIBU-
their behavior (e.g., specifically search-
TION.
ing for any evidence of deceit in the ex-
subjective probability a person- perimental procedure) as they become
specific estimate, derived from individ- increasingly knowledgable about the na-
ual experience, of the likelihood of a ture of deception in experiments. Sources
given event or outcome. Also called per- of sophistication include participating
sonal probability. in research previously, talking to partici-
pants from other studies, and reading
subjective scoring see OBJECTIVE about studies in the mass media.
SCORING.
subject-specific m o d e l in the analysis
subjective test an assessment tool that
of LONGITUDINAL DATAfor example,
is scored according to personal judg-
ment or to standards that are less sys- via REGRESSION ANALYSIS Or TIME-SERIES
tematic than those used in OBJECTIVE ANALYSISa mathematical representa-
TESTS, as in some essay examinations. tion in which the PARAMETERS are fit to
Although there are no necessarily right a given person's data. Compare POPULA-
TION-AVERAGED MODEL.
or wrong answers, responses are scored
based on appraisals of their appropriate- subjects' rights see PARTICIPANTS'
ness or quality. Also called subjective RIGHTS.
examination.
subjects-to-variables r a t i o the num-
subjectivism n. in general, any position ber of research participants compared to
holding that judgments of fact or value the number of research variables (e.g., 5
reflect individual states of mind rather to 1) in a srady. The subjects-to-variables
than states of affairs that can be said to ratio provides a guideline for determin-
be trae or false independently of indi- ing what size of sample to use to ensure
viduals. Compare OBJECTIVISM. sub- the greatest POWER when conducting a
jectivist n., adj. statistical analysis, such as MULTIPLE RE-
GRESSION or FACTOR ANALYSIS.
subjectivity n. the tendency to inter-
pret data or make judgments in the light subject variable an experience or
of personal feelings, beliefs, or experi- characteristic of a research participant
ences. Compare OBJECTIVITY. that is not of primaty interest but none-
theless may influence srady results and
subject m a t c h i n g see MATCHING.
thus must be accounted for during ex-
subject m a t u r a t i o n see MATURATION. perimentation or data analysis. Exam-
ples include age, marital status, religious
subject role a coherent set of behaviors affiliation, and intelligence. A variable
assumed by a research participant in re- of this type is neither manipulated by
sponse to his or her perceptions of what the experimenter, as an INDEPENDENT
is required or expected in the siraation. VARIABLE might be, nor is it usually
See APPREHENSIVE-SUBJECT ROLE; FAITH-
changed in the course of the experi-
FUL-SUBJECT ROLE; GOOD-SUBJECT ROLE;
ment, as a DEPENDENT VARIABLE might
NEGATIVISTIC-SUBJECT ROLE. be. Also called background variable;
subject selection effect see SAM- individual-difference variable; par-
PLING BIAS. ticipant variable.
subject sophistication a familiarity subpopulation n. a subgroup of a larger
with general research procedures that POPULATION of individuals or cases. For
may influence a person's behavior as a example, if the population of interest

362
 summated rating scale

is all U.S. college undergraduates, one substitution m a t r i x see TRANSITION

could divide the students into subpop- MATRIX.
ulations by state (e.g., individuals at-
substitution test any examination in
tending college in California, New York,
which the test taker exchanges one set of
etc.).
symbols for another. For example, a per-
subsample n. a subset of a SAMPLE of in- son might be required to substitute
dividuals or cases selected for study. For numbers with letters according to a spe-
example, if one chooses a random group cific code or to substitute a word in a
of college students for research pur- sentence with a grammatically equiva-
poses, one could divide the sradents into lent alternative. Substiration tests often
subsamples by major (e.g., individuals are used in neuropsychology to evaluate
from the group who are majoring in psy- cognitive function and identify individ-
chology, mathematics, etc). uals with brain damage, dementia, and
other conditions.
subscale n. a SCALE that taps some spe- subtest n. a separate division of a test or
cific constituent or otherwise differenti- instrament, usually with an identifiable
ated categoty of information as part of content (e.g., the multiplication subtest
a larger, overall scheme. For example, of a mathematics test).
a test of intelligence might consist of
several subscales (or subtests) assessing successive comparison method see
verbal and performance aspects or di- SIMULTANEOUS COMPARISON METHOD.
mensions of intelligence, which in com- sufficient estimator an ESTIMATOR
bination yield a verbal intelligence score, that utilizes all the relevant or useful in-
a performance inteUigence score, and an formation from a sample in order to ap-
overall intelligence score. proximate a population parameter. For
example, the SAMPLE MEAN is a suffi-
subset n. in mathematics and logic, a set cient estimator for the POPULATION
(collection of entities) that is part of a MEAN. Also called sufficient statistic.
larger set. See also VENN DIAGRAM.
summary statistics a set of statistics
substantive hypothesis a statement, used to communicate the most impor-
derived from a theoty being tested, that tant descriptive information about a col-
indicates expectations about the type, lection of raw data in succinct form. For
strength, or direction of relationship example, the VARIANCE, CENTRAL TEN-
among variables. The substantive hy- DENCY, SKEWNESS, and KURTOSIS may
pothesis is not empiricaUy testable until be used to summarize a sample data set.
each of its terms is given an OPERA- summary table any tabular presenta-
TIONAL DEFINITION. For example, if the tion of cracial descriptive information
substantive hypothesis is that those in- about a data set. An ANOVA SUMMARY
terested in biology will do better in med- TABLE is an example.
ical school, the concept of "interest in
biology" could be operationally defined summated r a t i n g scale a series of
as the score on a questionnaire assessing statements measuring the same con-
one's curiosity about and experiences in stract or variable to which respondents
this area, and "doing better" could be indicate their degree of agreement or
operationally defined as one's grade disagreement. The number of response
point average in medical school. See also options for each item varies, often from
RESEARCH HYPOTHESIS. 5 to 7 points (e.g., from strongly agree to
strongly disagree). The response values for
substantive significance see PRACTI- individual items may be summed to ob-
CAL SIGNIFICANCE. tain a total or average score that reflects

363
summative evaluation

a person's general attirade toward the (7 - 5)^ = 2^ = 4: This would be added to

construct of interest. A LIKERT SCALE is the squared deviations of all other indi-
the most commonly used summated rat- viduals in the sample. Various types of
ing scale. Also called summated rat- sums of squares are calculated in ANALY-
ing method; summated scale. SIS OF VARIANCE, REGRESSION ANALYSIS,
and other statistical procedures. Also
summative evaluation in evaluation called sum of the squared devia-
research, an attempt to assess the overall tions. See BETWEEN-GROUPS SUM OF
effectiveness of a program in meeting its SQUARES; REGRESSION SUM OF SQUARES;
objectives and goals after it is in opera- ERROR SUM OF SQUARES; WITHIN-
tion. This is in contrast to FORMATIVE GROUPS SUM OF SQUARES.
EVALUATION, which is used to help in
the development of the program. sum of squares of cross-products
(symbol: SSCP) the value obtained by
sum o f cross-products a statistical
first multiplying the squared differences
value obtained for two sets of variables Xj of individual x and y scores in a sample
and yj defined by the sumrnation from their respective means and then
2(^, - A ) (y,-y), adding all of the results. It is used in cer-
tain statistical analyses to determine the
where x is the mean value of Xj and y the COVARIATION between two sets of scores
mean value of y,-. It is used in MULTI- for a group. For example, if a participant
VARIATE ANALYSIS OF VARIANCE. in a sample has x and y scores of 4 and 7,
sum o f products the value obtained by respectively, and the means of the x and
multiplying each pair of numbers in a y variables are 2 and 4, respectively, the
set and then adding the individual to- squared cross-product for the partici-
tals. For example, for the set of number pant would be (4 - 2)^ X (7 - 4)^ = 4 X 9 =
pairs 36; this deviation would then be added
to the deviations for all other partici-
2,4 pants in the sample to yield the sum of
3, 5 squares of cross-products.
6, 6
1,4 . sum o f the squared deviations see
SUM OF SQUARES.
the siim of cross products is (2 x 4) + (3 x
5) + (6 X 6) + (1 X 4) = 8 + 15 + 36 + 4 =supersaturated
63. design a type of FAC-
TORIAL DESIGN in which dozens of INDE-
sum-of-ranks test see RANK-SUM TEST.
PENDENT VARIABLES that influence a
sum o f squared errors (symbol: SS) single DEPENDENT VARIABLE of interest
see ERROR SUM OF SQUARES. are assessed simultaneously to identify
those vital few that are most influential.
sum of squares (symbol: SS) the num- In a supersararated design, the number
ber obtained by determining the devia- of variables being investigated exceeds
tion of each point in a data set from the number of runs or trials conducted;
some value (such as a mean or predicted algorithms and other criteria exist to de-
value), multiplying each deviation by it- termine the appropriate number of runs
self, and adding the resulting products. to conduct. Such designs commonly are
Thus, for a set of variables x^, used for screening purposes, wherein the
goal is to identify a small number of
dominant active factors with minimal
where x is the mean value of Xj. For ex- cost.
ample, if an analysis yields a mean score
of 5 but a person's acraal score is 7, the suppressor effect a reduction in the
squared deviation for that individual is correlation between two variables due to

364
 survival probability

the influence of a third variable. Com- methods used to model a variety of time-
pare REINFORCER EFFECT. related outcomes. The simplest applica-
tion of survival analysis involves esti-
suppressor variable 1, a condition or mating the amount of time until the
characteristic that is associated with an occurrence of an event (e.g., death, ill-
INDEPENDENT VARIABLE, SUCh that the ness, graduation, marriage) for a group
correlation between the independent of individuals, but the technique also
variable and the DEPENDENT VARIABLE is may be applied to compare durations for
less than what it would be otherwise. 2. two or more groups and to build
in REGRESSION ANALYSIS, a piedlctor multivariate models that explain varia-
variable that is highly correlated with tion in duration. Survival analysis can be
Other predictors but that has a small cor- more informative than other techniques
relation with the outcome variable. The because it gives insight into the relation-
suppressor variable thus serves to re- ship between time and the outcome of
move the other predictors' variance that interest. That is, it enables the researcher
is irrelevant to the outcome, resulting in to determine not just whether an out-
stronger relationships with the depend- come is likely to occur but whether it
ent variable of interest. See also PAR- will occur early or late and whether the
TIALING.
chances of event occurrence increase
surrogate end p o i n t in CLINICAL TRI- gradually or sharply over time. Also
ALS, a measure that is intended to substi- called duration analysis; event his-
tute for or predict a particular outcome. tory analysis.
Surrogate end points often are used survival curve in SURVIVAL ANALYSIS,
when the primary end point is undesir- a graph showing the probabUity of an
able or rare, thus making it unethical or event's occurrence at specific points in
impractical to conduct a clinical trial to time, as in the illustration overleaf using
gather a statistically significant number hypothetical data.
of end points. For example, a researcher In this example, the survival curve de-
testing a drag to decrease the likelihood picts the number of individuals who
of death from heart disease may use died from a disease of interest at various
blood pressure as a surrogate end point. points in time relevant to the study of
surrogate v a r i a b l e see PROXY VARI- the specific disease. Also called survival
ABLE. plot.
survey n. a study in which a group of s u r v i v a l f u n c t i o n in SURVIVAL ANALY-
participants is selected from a popula- SIS, a mathematical formula that de-
tion and some selected data about or scribes the relationship between .the
opinions of those participants are col- likelihood of a particular event occur-
lected, measured, and analyzed. Infor- ring and a given time frame.
mation typically is gathered by inter- s u r v i v a l model in SURVIVAL ANALYSIS,
view or self-report questionnaire, and a model that shows the probability of
the results thus obtained may then be the occurrence of a well-defined event
extrapolated to the whole population. at different time points while taking
survey error the degree to which the into account the influence of certain
results of a SURVEY are inaccurate, due to predictors or explanatoty variables. See
such factors as SAMPLING BIAS, content CONTINUOUS-TIME SURVIVAL MODEL;
or measurement flaws, or the RESPONSE DISCRETE-TIME SURVIVAL MODEL.
BIASES of participants. See MEASURE- survival plot see SURVIVAL CURVE.
MENT ERROR; SAMPLING ERROR.
survival p r o b a b i l i t y in SURVIVAL
survival analysis a family of statistical ANALYSIS, the likelihood of experiencing

365
survival rate

survival curve

the event of interest as a function of conditions in a specific pattern that

time. It often is computed using the controls for personal- and time-related
KAPLAN-MEIER ESTIMATOR. variations that may influence individu-
als' responses. The pattern can be ex-
survival rate in SURVIVAL ANALYSIS, a pressed as a combination of rows and
value used to indicate the frequency of a columns. The number of columns usu-
particular event's occurrence at a spe- ally equals the number of unique pairs
cific point in time. For example, in a of experimental conditions and the
health study, the survival rate (i.e., oc- number of rows tends to be three or four,
currence of death) may be lower within regardless of the total number of condi-
the first year of the diagnosis of a moder- tions being studied. For example, sup-
ately serious disease but become progres- pose a researcher is investigating the
sively higher over time as the disease effect of threetypesof instraction (A, B,
becomes more serious. and C) on the performance of six partici-
survival ratio in SURVIVAL ANALYSIS, a pants, each of whom has a different
value used to indicate the elapsed time learning curve. A switchback arrange-
prior to an event of interest (e.g., disease ment for such a study might be
onset, death) in relation to the total
ACBACB
amount of time studied. Larger survival
ratios indicate a longer time until the CBABAC
event's occurrence.
ACBACB
survival time in SURVIVAL ANALYSIS, C BABAC
the amount of time that elapses untU the
occurrence of a negative event (e.g., Each row represents a distinct time period
death, disease). Also called failure time. and each column represents a specific par-
ticipant, with only two instraction types
suspended rootogram see HANGING switching back and forth for any given
ROOTOGRAM. participant.
SVD abbreviation for SINGULAR VALUE
symmetrical confidence interval an
DECOMPOSITION.
estimated range of values (CONFIDENCE
switchback design a type of WITHIN- INTERVAL) for an unknown population
SUBJECTS DESIGN in which participants PARAMETER that foUows a NORMAL DIS-
are assigned to multiple experimental TRIBUTION. Each side of the interval mir-

366
 systematically biased sampling

rors the other, such that half of the synchronic adj. describing research
values are in the left tail of the distribu- that focuses on events or other phenom-
tion and the other half are in the right ena at a given point in time. For exam-
tail. Compare ASYMMETRICAL CONFI- ple, a synchronic approach to linguistics
DENCE INTERVAL. would seek to characterize linguistic fea-
symmetrical d i s t r i b u t i o n a distribu- tures at a particular time, without refer-
tion in which the frequency of values ence to their historical development.
above the MEAN are a mirror image of Synchronic research contrasts with
those below the mean. Compare ASYM- diachronic research, which focuses on
METRICAL DISTRIBUTION.
processes of change over time.
symmetrical m a t r i x a SQUARE MA- synergism n. the joint action of differ-
TRIX in which the values above the MAIN ent elements such that their combined
DIAGONAL are a mirror image of the val- effect is greater than the sum of their in-
ues below the diagonal. It has the prop- dividual effects. For example, synergism
erty A = (fly), orfly= ajj, with / denoting occurs in organizational behavior when
rows and / denoting columns. a work group's overaU performance ex-
ceeds the aggregate performance of its
individual members. See also INTERAC-
TION EFFECT. synergistic adj.
synthetic approach the combining
Thus, in the example matrix above the (synthesizing) of various processes, sys-
third value in column 2 is the same as tems, skills, or other components into a
the second value in column 3 (i.e., 1). more complex whole as a means of
learning or better understanding the
symmetry n. equality relative to some whole. For example, a synthetic ap-
axis. More specifically, it is a condition proach to learning to read is one in
in which values are arranged identically which the child first learns to recognize
above and below the middle of a data set written letters and understand their as-
(see NORMAL DISTRIBUTION) or abovesociated sounds before learning to com-
and below the diagonal of a matrix (see bine letters into syllables and words.
SYMMETRICAL MATRIX). Many Standard Compare ANALYTIC APPROACH.
statistical techniques are appropriate
only for symmetrical data, such that synthetic risk m a p a graphical dis-
nonsymmetrical data often are trans- play of the results of a HAZARD ANALYSIS,
formed into a roughly symmetrical form which uses multivariate statistical proce-
prior to analysis. Compare ASYMMETRY. dures to identity and understand haz-
symmetrical adj. ardous or risky situations. Synthetic risk
maps provide an estimate of what the
symmetry test a method of determin- prevalence rate of the hazard (e.g., can-
ing whether a graphical representation cer mortality from airborne toxins) in
of a data set demonstrates SYMMETRY a given area is expected to be given the
about its X-axis, y-axis, or origin, such demographics of the area.
that knowing the arrangement of values Consider the example overleaf from
for one portion of the graph (e.g., values the 2002 National-Scale Air Toxics As-
above the mean) enables one to deter- sessment of the U.S. Environmental Pro-
mine the values for the opposite, mirror- tection Agency.
image portion of the graph (i.e., values
below the mean). If the data do possess synthetic variable see LATENT VARI-
ABLE.
such symmetty, they can be analyzed
using statistical techniques that assume systematically biased sampling
a NORMAL DISTRIBUTION. any sampUng method that consistently

367
systematic error

Average Risk Level

<1 in a Million
1 ^ 1 - 25 in a Million
^ 25 - 50 in a Million
I 50 - 75 in a Million .
I 75 -100 in a Million
H >100 in a Million

synthetic risk map

favors some outcomes or characteristics measurement imprecision). Systematic

over others, such that the resulting sam- error variance is generally positive or
ple does not accurately represent the generally negative and makes it more
larger population from which it was difficult to identify the unique effects of
drawn. For example, suppose a re- the INDEPENDENT VARIABLE in experi-
searcher conducts a phone survey using mental manipulations. Compare RAN-
telephone directories to identify partici- D O M ERROR VARIANCE.
pants. This sampling approach would be
systematically biased in not including in systematic naturalistic observa-
the sample those who do not have t i o n an objective, consistent method
phones, those who have opted out of re- for recording behavior as it occurs in the
ceiving calls of certain types, and those natural environment, without any inter-
who have unlisted numbers. vention or manipulation of variables by
the researcher. It is a more focused form
systematic error error in which the of NATURALISTIC OBSERVATION in that it
data values obtained from a sample devi- employs a structured system to examine
ate by a fixed amount from the trae val- a single behavior of interest (or one spe-
ues within the population. For example, cific aspect thereof), measuring its fre-
a scale that repeatedly provides readings quency, duration, latency, or other
0.5 g lower than the true weight would characteristics as applicable.
be demonstrating systematic error. Sys-
tematic errors tend to be consistently systematic observation an objective,
positive or negative and may occur as a well-ordered method for close examina-
result of SAMPLING BIAS or MEASURE- tion of some phenomenon or aspect of
MENT ERROR. Also called systematic behavior so as to obtain reliable data un-
bias. Compare RANDOM ERROR. biased by observer interpretation. Sys-
tematic observation typically involves
systematic error variance a consis- specification of the exact actions, attrib-
tent directional discrepancy in scores that utes, or other variables that are to be re-
is produced by extraneous factors (e.g.. corded and precisely how they are to be

368
 systems theory

recorded. The intent is to ensure that, strategy, the coding and analysis of the
under the same or similar circumstances, included studies, and other procedures
all observers will obtain the same results. are explicitly defined in advance in
order to ensure that the process is trans-
systematic r a n d o m s a m p l i n g see parent and can be replicated, with the
SYSTEMATIC SAMPLING. aim of minimizing bias in terms of what
is reviewed and the conclusions that are
systematic r e p l i c a t i o n the process of drawn.
conducting a study again but with cer-
tain consistent differences, often in an systematic sampling a type of sam-
attempt to extend the original research pling process in which all the members
to different settings or participants. For of a population are listed and then some
example, a systematic replication could objective, orderly procedure is used to
refine the design (e.g., by using more randomly choose specific cases. For ex-
participants) or the methodology (e.g., ample, the population might be listed al-
by using more standardized procedures phabetically and evety seventh case
or objective measures). Compare DIRECT selected. Also called quasi-random
REPLICATION. sampling; systematic random sam-
pling.
systematic review an organized
method of locating, assembling, and systematic variance see UNSYSTEM-
evaluating a body of literature on a par- ATIC VARIANCE.
ticular topic using a specific set of crite- systems theory see GENERAL SYSTEMS
ria. Standards for inclusion, the search THEORY.

369
 Tt
T symbol for the statistic obtained from simple printed form, used for checking,
the WILCOXON SIGNED-RANKS TEST. counting, or scoring a variable. For ex-
ample, a researcher might use a tally
see HOTELLING'S T^ TEST.
sheet to record the frequency of occur-
table n. a presentation of data in the rence of various behaviors or other
form of an ordered arrangement of over- events. Also called tally chart.
laid vertical columns and horizontal T A R abbreviation for THRESHOLD AUTO-
rows. As with a GRAPH, the purpose of a REGRESSION.
table is to communicate information (ei-
ther in words or numerical values) in a target n. 1. an area or object that is the
concise, space-efficient manner that can focus of a process, inquity, or activity. 2.
be assessed at a glance and interpreted the goal object in a task. For example,
easily. The columns have headings (the the target in a visual search might
leftmost column, which usually lists the be to find a letter S in a randomly ar-
independent variable, is referred to as ranged array of letters. In some concept-
the stub column). The intersection of a discovety tasks, the target is the rule that
column and row is called a CELL. Tables classifies objects as belonging or not be-
are often accompanied by explanatory longing to a category. Where a search
notes. tabular adj. has more than one item as its goal, these
are known as the target set.
tachistoscope n. a device that displays
target group the specific group of peo-
(usually by projecting) visual material
ple within a TARGET POPULATION that is
on a screen for a specific amount of time,
the focus of research. The target group
usually at very brief intervals. Words, may be defined by age, gender, marital
numbers, pictures, and symbols can be status, or other similar background vari-
rapidly presented in the right or left vi- ables. Often, a particular combination of
sual field. The device is used in experi- variables, such as men ages 20 to 50,
ments that are concerned with visual may define the target group.
perception, recognition speed, and
memory. It is also widely used in market target p o p u l a t i o n the population
research concerned with advertising, that a study is intended to research and
logos, branding, and so forth. Also called to which generalizations from samples
T-scope. are to be made. Also called reference
population. See also TARGET GROUP.
tactile test any test designed to mea-
sure how people perceive something target stimulus a specific stimulus to
through the sense of touch. which participants in a test or experi-
mental procedure must attend or re-
t a i l of a d i s t r i b u t i o n in a PROBABIL- spond. For example, in tests of hearing
ITY DISTRIBUTION, especially one that is the target stimulus may be a specific
graphically displayed, the region (or re- tone that must be identified.
gions) of least frequently occurring val-
ues. This is often the CRITICAL REGION in Tarone-Ware test in SURVIVAL ANAL-
YSIS, a test that can be used to determine
tests of STATISTICAL SIGNIFICANCE.
whether two survival curves are equiva-
tally sheet an instrument, usually a lent when there are reasons to think that
 temporal validity

the traditional survival test may not give is the sampling distribution of the statis-
useful results. This may be because test tic (M-[IQ)/S, where [LQ is the mean of the
assumptions have been violated, such as population from which the sample is
the assumption of independence of CEN- drawn, M is the estimate of the mean of
SORING, or because there are factors un- the population as obtained from sample
accounted for in the analysis that affect data, and s is the standard deviation of
survival or censoring times or both. The the data set. Also called Student's t
Tarone-Ware test is designed to be used distribution.
with a large percentage of censored data
and with nonnormal distributions. telephone i n t e r v i e w an interview
[Robert E. Tarone, U.S. epidemiologist; that is conducted by telephone rather
James H. Ware, U.S. statistician] than face to face, QUESTIONNAIRES and
surveys involving large numbers of geo-
task analysis 1. the breakdown of a graphically dispersed participants are
complex task into component tasks to often carried out in this way. Some-
identify the different skills needed to times, telephone interviews may be used
correctly complete the task. In educa- to screen participants in order to narrow
tion, for example, it entails the break- the pool of those who will be invited for
down of a subject or field of study to in-person interviews.
identify the specific skills the student
must possess in order to master it. 2. a t e m p o r a l consistency the CORRELA-
method of evaluating a product or sys- TION between measurements obtained
tem in which researchers interview ac- when the same test or instrument is ad-
tual or target users in order to find out ministered to the same sample on two
information such as (a) what tasks are different occasions. Temporal consis-
performed, (b) which of these are most tency is an index of the RETEST RELIABIL-
frequently performed and which are ITY of an instrament. This approach
most important, (c) how and in what se- assumes that there is no substantial
quence the tasks are performed, (d) what change in the CONSTRUCT being mea-
standards of performance apply, and (e) sured between the two occasions. The
how different categories of user vaty in longer the time gap, the greater the like-
their answers to the above. Although lihood of a lower correlation. Also called
some scripted questions are asked, the temporal stability. See RELIABILITY.
interviews are otherwise unstrucrared, temporal frequency the number of
the better to reflect users' actual experi- occurrences of a repeating event per unit
ence. time. For example, if 80 repeating events
t a u test see KENDALL'S TAU. occur within 20 seconds, the frequency,
f, is 80/20 = 4.
Taylor series an infinite sum of terms
that are calculated from the values of t e m p o r a l precedence in establishing
a function's DERIVATIVES at a single cause-effect relationships between two
point a. The process provides for a Tay- variables, the principle that the cause
lor expansion of the function f(x) about must be shown to have occurred before
x = a. [Brook Taylor (1685-1731), British the effect. Two other requirements are
mathematician] those of COVARIATION and nonspu-
riousness (i.e., there are no plausible al-
Tchebechev's inequality see CHE- ternative explanations for the observed
BYSHEV'S INEQUALITY.
relationship).
t d i s t r i b u t i o n a theoretical PROBABIL-
temporal stability see TEMPORAL
ITY DISTRIBUTION that plays a central
CONSISTENCY.
role in testing hypotheses about popula-
tion means, among other parameters. It temporal v a l i d i t y a type of EXTERNAL

371
terminal event

VALIDITY that refers to the generaliza- testable hypothesis a HYPOTHESIS

bility of a study's results across time. that generates predictions of what is
Also called temporal external valid- likely to occur (or not occur) that are ca-
ity. pable of being evaluated empirically. To
be truly testable, a prediction must be
t e r m i n a l event in a series of related capable of being operationalized such
events, an event that can occur only that reliable measures to collect unbi-
once and after which no other event of ased data are available.
interest can occur. In many clinical and
observational studies the terminal event test a f l m i n i s t r a t i o n the giving of a
is death, whereas the nonterminal or re- test for the purpose of obtaining infor-
current events include hospitalizations, mation, especially in a standardized
relapses, repeat of behaviors and symp- manner that can be rephcated.
toms, and the like. For the analysis of
such situations and data, one must take test age see AGE EQUIVALENT.
into account the dependence among dif- test analysis a detailed statistical analy-
ferent types of recurrent events and that sis bf a test's PSYCHOMETRIC properties,
between the recurrent events and the including an evaluation of the quaUty of
terminal event. the test items and of the test as a whole.
It usually includes information such as
tertiary s a m p l i n g u n i t a unit se-
the MEAN and STANDARD DEVIATION for
lected in the third stage of sampUng. For
the test scores in the population used to
example, suppose the job satisfaction of
develop the test as well as data on the
employees in a supermarket chain is
test's RELIABILITY; it may also include
being studied. The first sampling could
data on such factors as item DIFFICULTY
be drawn by identitying stores geo-
VALUE, ITEM DISCRIMINABILITY, and the
graphically across the United States; this
impact of ITEM DISTRACTORS.
would be the PRIMARY SAMPLING UNIT.
Then, within each geographical loca- test battery a group, series, or set of sev-
tion, stores from large urban areas would eral tests designed to be administered as
be selected for possible inclusion in the a unit in order to obtain a comprehen-
research; this would be the SECONDARY sive assessment of a particular factor or
SAMPLING UNIT. Finally, cashiers from phenomenon. For example, a health re-
within the urban stores would be se- searcher may administer a battery of
lected to complete the job satisfaction health surveys to a group of individuals
survey; this last sample would be the ter- diagnosed with a particular disease to as-
tiaty sampUng unit. sess multiple facets of the disease. De-
pending on the purpose of testing,
test 1. n. any procedure or method to ex- individual tests may measure the same
amine or determine the presence of or different areas (or both) and may be
some factor or phenomenon. 2. n. a scored separately or combined into a
standardized set of questions or other single score.
items designed to assess knowledge,
skills, interests, or other characteristics test bias the tendency of a test to sys-
of an examinee. See PSYCHOLOGICAL ternaticaUy over- or underestimate the
TEST. 3. n. a set of operations, usually true scores of individuals to whom that
statistical in nature, designed to deter- test is administered, for example because
mine the VALIDITY of a hypothesis. 4. they are members of particular groups
vb. to administer a test. (e.g., ethnic minorities, genders). See
also CULTURAL TEST BIAS.
testability n. the degree to which a hy-
pothesis or theory is capable of being test construction the creation of a
evaluated empirically. test, usually with a clear intent to meet

372
 test f o r two independent proportions
the usual criteria of VALIDITY, RELIABIL-
ITY, NORMS, and other elements of test
standardization.
test d a t a 1. any data gathered from a
TEST. 2. data that have been specifically
identified for use in testing models, pro-
grams, or assumptions. For example,
data may be used in a confirmatoty way,
typically to verify that a given set of
input to a given function produces some
expected result. Other data may be used
in order to challenge the abilify of the
program to respond to unusual, extreme, test f o r i n d e p e n d e n c e a procedure
exceptional, or unexpected input.
test f o r c o n t r a s t s any procedure used
to determine which of the specific groups
examined i n an A N A L Y S I S O F V A R I A N C E
are significantly different from each
other, whether this takes the form of
a POST HOC COMPARISON Carried out
after the data have been observed or an A
PRIORI COMPARISON formulated before
observation. For example, suppose one test f o r n o r m a l i t y any procedure used
wished to analyze the outcomes from
three different modes of training: (a)
video-based instraction, (b) live lecture
by an instructor, and (c) text reading
only. After conducting an analysis of
variance and obtaining results indicat-
ing there is a significant difference be- the statistical tests may not be warranted.
tween the three modes, one could use
a test for contrasts to evaluate various
combinations of modes to identify
where exactly the differences lie. More
specifically, one comparison could be
between video-based testing versus lec-
ture, which assesses two modes that in-
volve an instractor explaining content
to the sradents; another comparison
could be the video and live lecture
groups versus the text reading, which as-
sesses an instructor presenting informa-
tion against learning via reading; and so
on until all possible combinations have
been evaluated. Examples of specific
tests for contrasts include the SCHEFFE
test f o r t w o i n d e p e n d e n t p r o p o r -
TEST, T U K E Y ' S . H O N E S T L Y SIGNIFICANT
DIFFERENCE TEST, and the FISHER LEAST
SIGNIFICANT DIFFERENCE TEST.
test f o r e q u a l i t y o f v a r i a n c e any
373
procedure used to test for H O M O G E N E I T Y
OF VARIANCE. Examples include the
BARTLETT TEST FOR EQUALITY OF VARI-
ANCE and the LEVENE TEST FOR EQUALITY
OF VARIANCE. The assumption that the
VARIANCES in the populations sampled
are substantially equal is basic to many
statistical procedures. If this assumption
is violated, it may be necessaty to trans-
form the data or use NONPARAMETRIC
TESTS. Also called test of homogene-
ity.
used to test the hypothesis of association
or relationship between two variables.
The test compares the observed frequen-
cies of a variable with the frequencies
that would be expected if the N U L L H Y -
POTHESIS of no association (i.e., statisti-
cal independence) were true. The CHI-
SQUARE TEST is often used for this pur-
pose.
to test whether a data set follows a NOR-
M A L DISTRIBUTION. Many statistical pro-
cedures are based on the assumption
that the R A N D O M VARIABLE is normally
distributed. When this assumption is vio-
lated, interpretation and inference from
Often normality is most conveniently as-
sessed using graphical methods, such as
a S T E M - A N D - L E A F PLOT, to visualize the
differences between an empirical distribu-
tion and the standard normal distribu-
tion. Altematively, numerical methods
can be used to present summary statis-
tics, such as SKEWNESS and KURTOSIS,
which indicate the degree of nonnor-
mality. In addition, there are statistical
tests of normality, such as the K O L M O -
G O R O V - S M I R N O V GOODNESS-OF-FIT TEST,
the SHAPIRO-WILK TEST, the DARLING
TEST, and the C R A M E R - V O N MISES G O O D -
NESS-OF-FIT TEST.
t i o n s a statistical test used to determine
whether a PROPORTION created by a ran-
dom sample represents the proportion
for the entire population or whether
testing effect

the difference between two proportions, which some treated participants are
measured in two samples, is statistically likely to experience extreme reactions in
significant. An example is the z TEST for either direction is one that generally is
a population proportion. subject to intentional distortion of re-
sponses. Tests of extreme reactions are
testing effect the research finding that typically tests of ranks and are used
the long-term retention of information when tfiere is an indication beforehand
is significantly improved by testing learn- that the experimental condition may
ers on the information. Exams or tests cause such reactions.
seem to activate retrieval processes that
facilitate the learning of study material test of homogeneity see TEST FOR
and cause knowledge to be stored more EQUALITY OF VARIANCE.
effectively in long-term memory.
test o f significance any statistical test
test interpretation the clinical, edu- or procedure, such as a T TEST, Z TEST, F
cational, vocational, or other practical TEST, or CHI-SQUARE TEST, used in SIG-
implications and inferences given to a NIFICANCE TESTING.
particular test result. Such conclusions test o f simple effects in an ANALYSIS
are typically drawn by an expert in test- OF VARIANCE, a test to determine the ef-
ing or by suitable computer software. fect of one INDEPENDENT VARIABLE On
test i t e m a constituent part, or the the DEPENDENT VARIABLE at a Single level
smallest scoreable unit, of a test. It is the of a second independent variable; the
stimulus (question or task) to which a test examines the effects of one of the in-
test taker responds. dependent variables with the other in-
dependent variable held constant. When
test n o r m the standard of performance a statistically significant interaction is
typically attained in a test, as established found, in which, for example, there
by testing a large group of people (the are two independent variables, a and
STANDARDIZATION GROUP) and analyz- b, with each independent variable hav-
ing their scores. In NORM-REFERENCED ing two levels, a^ and 02 ^nd b^ and
TESTING, subsequent test takers' scores ^2, the question of interest turns to a
on the test are compared with the test
systematic examination of the narare of
norm to provide an estimate of the posi-
the interaction. In this case, a test of sim-
tion of the tested individual in a prede-
fined population, with respect to the ple effects involves two statistical tests:
trait being measured. one of the difference between and fl2
at i>i, and a second of the difference be-
test of association any of a category of tween and 3t ^2- The comparisons
STATISTICAL TESTS that examines the de- can also be undertaken by examining
gree of relationship or dependence be- the difference between and ^2 3t
tween variables. An example is the CHI- and the difference between bj and b2 at
SQUARE TEST. (li-
test power see POWER.-
test o f extreme reactions a proce-
dure used to test for differences in range test p r o f i l e an overall description that
between the responses (scores) of a treat- summarizes an individual's relative
ment group and those of the control standing or characteristics by collating
group. In some experiments, the treat- the findings from a series of tests or
ment is likely to increase the scores of subtests. For example, a personality pro-
some participants and at the same time file may present the data gathered on
decrease the scores of other participants; personality and other tests of interest
in contrast, the control group does not and be used to evaluate the individual in
experience such reactions. A situation in areas related to his or her personal, edu-
 tetrachoric correlation coefficient

cational, and professional lives, including measurement. Theoretical frameworks

temperament, decision-making methods include CLASSICAL TEST THEORY, GEN-
and communication style, and general ERALIZABILITY THEORY, and ITEM RE-
attirade toward work and life. SPONSE THEORY.
test r e l i a b i l i t y see RELIABILITY.
testwise adj. describing a test taker who
test-retest correlation the degree of has developed skills and strategies that
association between measurements of are not related to the constract being
the same variable when the same test is measured in the test but facilitate an in-
applied on separate occasions. It is a sim- creased test score. Experience with simi-
ple estimator of the RELIABILITY of a test lar tests, coaching, or the ability to
or instrument and is indexed by the respond advantageously to items that
test-retest coefficient. contain extraneous clues and sugges-
tions may yield a score that is higher
test-retest r e l i a b i l i t y see RETEST RE-
than the "trae" ability of the test taker.
LIABILITY.
tests a h d measurements in psychol- testwise a l p h a level in HYPOTHESIS
ogy, a field of interest that focuses on TESTING, the SIGNIFICANCE LEVEL (I.e.,
psychological testing and assessment. It the level of risk of a TYPE I ERROR) se-
encompasses the knowledge and skills lected for each individual test within
required in understanding, selecting, a larger experiment. This is in contrast to
scoring, and interpreting individual and the EXPERIMENT-WISE ALPHA LEVEL, which
group-administered psychological, edu- sets the total risk of Type I error for the
cational, and employment tests. The experiment. As more significance tests
field also includes the principles neces- are conducted, the experiment-wise alpha
saty to develop tests and to evaluate level goes up, unless there is an adjust-
their usefulness. ment to lower the testwise alpha level.
test score a numerical value assigned as See TESTWISE ERROR RATE.
a measure of performance on a test. testwise error rate in a test involving
test sensitization the design of a clas- MULTIPLE COMPARISONS, the probabil-
sification test so that it achieves opti- ity of making a TYPE I ERROR on any spe-
mum accuracy in identitying those cific test or comparison, FACTORIAL
participants who fall into particular cat- DESIGNS allow for the possibility of per-
egories. For example, a test may be sensi- forming many such individual con-
tized by including certain items that are trasts, and the related FAMILY-WISE
designed specifically to identity the ERROR RATE reflects the possibility of
highest performing students or employ- Type I error across the entire set of com-
ees. parisons. Also called comparison-wise
test statistic 1. the numerical result of error rate; per-comparison error
a STATISTICAL TEST, which is used to de- rate. See TESTWISE ALPHA LEVEL. See
termine STATISTICAL SIGNIFICANCE and
also EXPERIMENT-WISE ERROR RATE.
evaluate the viability of a hypothesis. 2. tetrachoric correlation coeffi-
any of the statistics relating to a test or cient (symbol: r^^^) an index reflecting
its components, such as indices of item the degree of relationship between two
difficulty, item RELIABILITY, DISCRIMI- continuous variables both of which
NABILITY, and so on. See TEST ANALYSIS.
have been dichotomized. For example,
test theory the body of theoty underly- a researcher may need to correlate pass-
ing the interpretation and use of test fail on a test and graduate-nongraduate
scores. Of central concern is the concept from school, where pass-fail is the
of RELIABILITYits definition and dichotomization of continuous scores

375
tetrad difference criterion

on a test and graduate-nongraduate is a estimator. [Henri Theil; Pranab K. Sen

dichotomization of grade point average. (1937- ), Indian-born U.S. statistician]
tetrad difference criterion in FAC- thematic analysis a QUALITATIVE re-
TOR ANALYSIS, a procedure used to study search strategy that identifies, analyzes,
the intercorrelations among tests, par- and reports recurrent identifiable pat-
ticularly intelligence tests. Suppose terns or themes within data. There are
there are five tests presumed to be mea- multiple phases to this process: The re-
suring five specific abilities (e.g., numer- searcher (a) familiarizes him- or herself
ical ability, verbal ability, perceptual with the data;,(b) generates initial codes
ability, spatial ability, memory), thus or categories for possible placement of
yielding 10 possible intercorrelations. themes; (c) collates these codes into po-
With the tetrad difference criterion, one tential themes, gathering all data rele-
examines four elements of the CORRELA- vant to each potential theme; (d) reviews
TION MATRIX at a time for each of the the chosen themes and checks that these
possible tetrads, cross-multiplying them work in relation to the coded extracts
and determining whether the differ- and the entire data set, effectively gener-
ences between the products approxi- ating a thematic "map" of the analysis;
mate to zero. Withfivetests, there are a (e) defines and names the themes, using
total of 15 tetrads. If all of the differences continuous ongoing analysis to refine
between tetrads approximate to zero, the specifics of each theme and to gener-
the assumption is that there is one un- ate clear definitions and names for
derlying factor andfivespecific factors. each one; and (f) produces the report,
Also called tetrad difference .which should aim to be a vivid and com-
method. peUing account of the data. See also
GROUNDED THEORY; THEORETICAL SAM-
Theil's method a NONPARAMETRIC ap- P L I N G .
proach to fitting a straight REGRESSION
LINE to a set of data; it is typically used theorem n. in mathematics and logic, a
to reduce the effect of OUTLIERS on statement or formula that can be de-
the SLOPE and INTERCEPT esfimates and duced from previously established or ac-
when the assumption of NORMALITY is cepted statements.
not required. There are two versions of
theoretical construct an explanatory
the method, Theil's incomplete method
concept that is not itself directly ob-
and Theil's complete method, which dif-
servable but that can be inferred from
fer in the amount of data that enter into
observed or measurable data. In psychol-
the calculations. [Henri Theil (1924-
ogy, many hypothesized internal pro-
2000), Dutch econometrician]
cesses are of this kind, being presumed
Theil's test f o r l i n e a r i t y a test to de- to underlie specified overt behaviors. For
termine whether a linear or nonlinear fit example, a personaUty dimension, such
is best for a REGRESSION LINE. [Henri as neuroticism, might be described as a
Theil] theoretical construct that is measured by
means of a questionnaire.
Theil-Sen estimator a popular rank-
based NONPARAMETRIC procedure for theoretical d i s t r i b u t i o n a DISTRIBU-
fitting a straight line to data. It is effi- TION that is derived from certain princi-
cient in conditions of HETEROGENEITY ples or assumptions by logical and
OF VARIANCE and resistant to OUTLIERS. mathematical reasoning, as opposed to
The procedure is based on choosing the one derived from real-world data ob-
median slope among all lines through tained by empirical research. Examples
pairs of data points. Also called Kendall of such distributions include the NOR-
robust line-fit method; Sen's slope MAL DISTRIBUTION, the BINOMIAL DIS-

376
 therapy outcome research

TRiBUTiON, and the POISSON DIS- tion. In order to gain a deeper un-
TRIBUTION. In general, the procedures of derstanding of the constructs involved,
INFERENTIAL STATISTICS involve taking the researcher samples new research
one or more EMPIRICAL DISTRIBUTIONS sites, cases, incidents, time periods, or
and referring these to an appropriate data sources to compare with those that
theoretical distribution. When there is have already been studied. In this way
correspondence between an empirical he or she seeks to build a theory from the
and a theoretical distribution, the latter emerging data while continuing to se-
may be used to make inferences (predic- lect new samples to examine and elabo-
tions) about the probability of future rate on the theory. See GROUNDED
empirical events. Also called reference THEORY.
distribution. theoretical statistics the study of sta-
theoretical frequency distribu- tistics from a mathematical and theoret-
t i o n the FREQUENCY DISTRIBUTION that ical perspective involving PROBABILITY
would result if data conformed to the THEORY, DESCRIPTIVE STATISTICS, IN-
values predicted by a theoty or law as op- FERENCES, and MODEL BUILDING. For ex-
posed to their acraal observed values. ample, a researcher could use theoretical
statistics to describe a set of achievement
theoretical p r o b a b i l i t y the mathe- data, conduct HYPOTHESIS TESTING, and
matical PROBABILITY of a particular create models assessing possible predic-
event occurring, as determined by divid- tors of achievement. Also called mathe-
ing the number of positive outcomes by matical statistics. Compare APPLIED
the total number of possible outcomes. STATISTICS.
For example, when flipping a coin, the
theoretical probabihty that a tail will theory n. 1. a principle or body of inter-
occur is 1 divided by 2 (the number of related principles that purports to ex-
possible outcomes), giving a probability plain or predict a number of interrelated
of .5 or 50%. Compare EMPIRICAL PROB- phenomena. See CONSTRUCT; MODEL. 2.
ABILITY. in the philosophy of science, a set of log-
ically related explanatoty hypotheses
theoretical relative frequency dis- that are consistent with a body of empir-
t r i b u t i o n for a discrete variable x, a ical facts and that may suggest more em-
hypothesized or expected distribution of pirical relationships. See SCIENTIFIC
observations or scores that will be ob- EXPLANATION, theoretical adj.
tained at each of the possible values of x
in relation to the total number of obser- theory-led thematic analysis a
vations to be made (i.e., their theoretical form of THEMATIC ANALYSIS in which
relative frequencies). In contrast to a a specific theoretical or epistemological
RELATIVE FREQUENCY DISTRIBUTION, approach is used to provide a direction
which indicates the relative frequencies for the analysis.
with which the values of x actually occur theory t r i m m i n g in PATH ANALYSIS or
in a chosen sample, a theoretical relative STRUCTURAL EQUATION MODELING, de-
frequency distribution specifies the leting nonsignificant paths from the
likely number of occurrences of each of model to improve model fit.
the possible values of x with reference to
the anticipated total. theory v e r i f i c a t i o n the process of de-
veloping and citing empirical evidence
theoretical s a m p l i n g a sampling to increase or bolster the tenability of
strategy, often adopted in QUALITATIVE theories.
RESEARCH, that Involves the PURPOSIVE
therapeutic t r i a l see CLINICAL TRIAL.
SAMPLING of further data while a theo-
retical framework is still under construc- therapy outcome research research

377
thick description

that investigates the end results of treat- incidental relationship between the two
ment or other interventions to which variables, or changes independently in
patients are exposed. The focus is pri- variable a and variable b. For example, as
marily on the "cure" (or not) of patients, the sales of air conditioners increase, the
but the research also evaluates their ex- number of drownings also increases: The
periences, preferences, and values, as unintentional third variable in this case
well as the wider impact on society. The would be the increase in heat. See HID-
aim is to identify shortfalls in practice DEN VARIABLE.
and to develop strategies to prevent or
mitigate problems and improve care. See thought experiment a mental exer-
also TREATMENT OUTCOME RESEARCH; cise in which a hypothesis, theoty, or
TREATMENT PROCESS RESEARCH. idea is put to the test without actually
conducting an experiment or research
t h i c k description in QUALITATIVE RE- project. The purpose is to explore the
SEARCH, a strategy that aims to describe logical consequences of a hypothesis
and interpret observed behavior within or principle. Thought experiments often
its particular context so that the behav- involve arguments about events or
ior becomes meaningful to an outsider. states of affairs of a hypothetical or
The context may be a small unit (a counterfactual nature, which neverthe-
family or work environment) or a larger less have implications for the actual
unit (a community or general culture). world. They can be used to challenge the
The researcher not only accurately de- intellectual status quo, correct misinfor-
scribes observed behavior or social mation, identityflawsin an argument,
actions but also assigns purpose, motiva- or generate ideas as part of a problem-
tions, and intentionaUty to these actions solving exercise. Thought experiments
by explaining the context within which are most famUiar in philosophy but are
they took place; thick description con- also used in the physical sciences, gener-
veys the thoughts and feelings of partici- ally as a step toward designing a physical
pants as well as the complex web of experiment. Also called idealized ex-
relationships among them. In contrast, periment.
thin description is a superficial account
that does not explore underlying mean- threats to v a l i d i t y factors that may
ings. threaten the VALIDITY of inferences
drawn from the results of an experiment
t h i r d moment see MOMENT. or research program. For example, com-
t h i r d quartile see QUARTILE. mon threats to the INTERNAL VALIDITY
of an experiment include (a) lack of clear
t h i r d variable see HIDDEN VARIABLE. TEMPORAL PRECEDENCE among vari-
third-variable p r o b l e m the fact that ables, leading to confusion of cause
an observed correlation between two and effect; (b) SAMPLING BIAS or other
variables may be due to the common nonrandom factors in the assignment of
correlation between each of the vari- participants to the different conditions
ables and a third variable rather than be- of the experiment; (c) the possibility
cause the two variables have any that events extraneous to the experi-
underlying relationship (in a causal ment, including REGRESSION TOWARD
sense) with each other. In other words, THE MEAN and naturally occurring
when two variables, a and b, are found changes over time, could cause the ob-
to be positively or negatively correlated, served effect; and (d) ATTRITION of par-
it does not necessarily mean that one ticipants, especially where differential
causes the other: It may be that changes attrition rates between groups produce
in an unmeasured or unintentional third artificial effects.
variable, c, are causing a random and co- Common threats to the converse EX-

378
 three-way analysis of variance

TERNAL VALIDITY of an experiment (i.e., mathematical function of these parame-

the extent to which its results can be ters.
generalized) include (a) the possibility three-period crossover design a
that an effect found with certain units or specific CROSSOVER DESIGN that is used
participants may not hold if other units in treatment research. As in other such
had been sradied; (b) the possibility that designs, part way through the experi-
the results obtained in one siraation, ment, all participants "cross over" or are
such as studying the effects of a treat- switched to another experimental con-
ment program over one year, will not be dition. A three-period design usually in-
the same in another situation (e.g., if the volves two different treatments and a
program was extended to two years); (c) control. For example, a researcher may
the possibility that the results obtained be interested in studying the effects of a
with one measure or observation may low dose (Dl) and high dose (D2) drug
not hold if other measures or observa- on the behavior of participants; the ex-
tions were used; and (d) the possibility periment will also involve administra-
that an effect found in one setting (e.g., tion of a placebo (DO). A three-period
an inner city) may not hold in another crossover design would involve the fol-
setting (e.g., a rural area). See also CON- lowing sequences of drug intervention:
STRUCT VALIDITY; STATISTICAL CONCLU- sequence #1 as DO, D l , D2; sequence #2
SION VALIDITY. as D l , DO, D2; and sequence #3 as D l ,
three-mode f a c t o r analysis an ex- D2, DO. Observations are collected be-
tension of classical two-mode FACTOR fore treatment and after treatment in
ANALYSIS to three-mode data, that is, each of the dosage conditions.
data classified by three characteristics at three-quarters h i g h rule a general
once. It enables the researcher to per- rale of thumb stating that the height (y-
form a simultaneous factor analysis of axis) of a graph should be roughly three
all three modes. For example, the scores fourths the length (j;-axis) of the graph.
of a sample of individuals on a battery of This is to avoid the distortion of the data
tests could be classified by the individu- that can result from choosing arbitraty
als in the sample as well as by the tests in (or deliberately misleading) scales of
the battery: The sample of individuals measurement along the x- and y- axes.
would constitute one mode of the data The representation of numbers on the
and the battery of tests a second mode. If graph should be directly proportional to
the battety of tests were administered the numerical quantities that are being
to the sample of individuals on several represented.
occasions, the set of occasions would
constitute a third mode. A three-mode three-sigma m l e a rule of thumb stat-
factor analysis would identity factors ap- ing that in a NORMAL DISTRIBUTION
pearing from correlations in each of the nearly all (approximately 99%) of the
different modes and those extending values will Ue within three STANDARD
across modes. DEVIATIONS (or sigmas) of the mean. In
certain problems in probability theoty
three-parameter model in ITEM RE- and mathematical statistics, an event is
SPONSE THEORY, a model that specifies considered to be practically impossible if
three parameters affecting an individ- it lies outside this region of values.
ual's response to a particular test item:
three-stage s a m p l i n g see MULTI-
(a) the difficulty level of the item; (b) the
STAGE SAMPLING.
DISCRIMINATING, POWER of the item;
and (c) in multiple-choice items, the ef- three-way analysis o f variance an
fect of guessing. The probability of a cor- ANALYSIS OF VARIANCE that isolates the
rect response to the item is held to be a MAIN EFFECTS of three independent vari-

379
three-way classification

Television violence
Stress level None Moderate High
Sugar
High 20 30 51
Low 17 15 16
No sugar
High 10 20 32
Low 10 12 8

three-way table

ables, fl, b, and c, on a dependent vari- way interaction is a type of HIGHER

able and their INTERACTION EFFECTS ORDER INTERACTION.
one THREE-WAY INTERACTION, axbxc,
and three two-way interactions, axb, ax three-way mixed design an experi-
c, and bxc. mental design in which there are three
variables of interest and where at least
three-way classification 1. a classifi- one variable is a BETWEEN-SUBJECTS FAC-
cation of observed data into groups or TOR (e.g., gender) and at least one vari-
classes based on a consideration of three able is a WITHIN-SUBJECTS FACTOR (e.g.,
characteristics at the same time. For ex- pretest and posttest scores for the same
ample, a college population could be participants); the third factor can be ei-
classified by gender, year in college, ther within subjects or between subjects.
and major. The correlations in such data Data from such designs often are evalu-
can be analyzed using a THREE-MODE ated with a three-way mixed design
FACTOR ANALYSIS. 2. a three-way CHI- analysis of variance.
SQUARE TEST used to determine the sig-
three-way repeated measures de-
nificance of the difference between the sign see THREE-WAY WITHIN-SUBJECTS
frequencies of occurrence in three cate- D E S I G N .
gories.
three-way table in THREE-WAY CLASSI-
three-way design an experimental de- FICATION, a tabular presentation of the
sign in which three INDEPENDENT VARI- values for the three classes or variables.
ABLES are examined simultaneously to This typically takes the form of a set of
observe their separate MAIN EFFECTS and two-way tables, as in the above hypo-
their joint INTERACTION EFFECTS on a thetical example of aggression ratings in
DEPENDENT VARIABLE of interest. Data a sample of 120 children as a function of
from such designs often are evaluated stress levels, level of violence in a televi-
with a THREE-WAY ANALYSIS OF VARI- sion program, and sugar intake.
ANCE.
three-way within-subjects design
three-way interaction in a THREE- an experimental design in which there
WAY ANALYSIS OF VARIANCE, the joint ef- are three variables or factors of interest
fect of all three independent variables, a, and the same participants are sradied
b, and c, on a dependent variable. A sta- in each of the levels of each of the fac-
tistically significant three-way interac- tors. For example, suppose a, b, and c
tion indicates that one or more of the represent type of training (lecture vs.
three possible two-way interactions (a x reading), type of content (physics vs.
b, axe, and bxc) differ across the levels psychology), and multiple tests over a
of a third variable. For example, the axb nine-month training course (tests at 3,6,
interaction may differ for one level of c and 9 months). All participants receive
compared to another level of c. A three- all treatment conditions of all three fac-
 tied ranks

tors; that is, all participants receive both the behavior or action of a participant)
types of training, for both contents, and changes from one state (e.g., unwilling
are tested three times. Data from such to buy a product) to another (e.g., will-
designs often are evaluated with a three- ing to buy the product). Also called
way within-subjects analysis of variance. threshold-crossing model.
Also called three-way repeated mea-
sures design. Thurstone scaling a method for de-
veloping a scale to assess attitudes to-
threshold autoregression (TAR) an ward a single subject of interest. There
AUTOREGRESSIVE MODEL used in the are six basic steps in the general ap-
analysis of nonlinear TIME SERIES. Non- proach: (a) identify the CONSTRUCT of
linear time-series models have the ad- interest (e.g., attitudes toward providing
vantage of being able to capture public funds for charter schools); (b)
asymmetries, jumps, and time irrevers- generate statements that describe spe-
ibility in complex empirical phenom- cific attirades that people might have to-
ena, such as financial and economic ward such a construct; (c) have judges or
data (e.g., changes in interest rates). The subject matter experts rate each state-
TAR model allows for changes in the ment on a numerical scale in terms of
model parameters when the value of an how far each statement indicates a fa-
exogenous variable (i.e., one from out- vorable attitude toward public funding
side the time series of interest) crosses a for charter schools (e.g., the rating 1 =
particular threshold. extremely unfavorable attitude toward pub-
lic funding for charter schools and the rat-
threshold-crossing data in a THRESH-
ing 11 = extremely favorable attitude
OLD MODEL, data used to determine the
toward public funding for charter schools);
point in time at which a THRESHOLD EF-
(d) compute scale score values for each
FECT occurs in the dependent variable.
item, together with the mean or median
This may have to be imputed (see IM-
and some measure of VARIANCE; (e) se-
PUTATION) if observation has not been
lect the final scale items, choosing state-
continuous.
ments that are at equal intervals across
threshold-crossing model see THRESH- the range of means or mediansfor ex-
OLD MODEL. ample, select one statement for each of
the 11 mean values, with the constraint
threshold effect an effect in a DEPEND-
that there be little variance in ratings for
ENT VARIABLE that does not occur until
the statements; and (f) administer the
a certain level, or threshold, is reached
scale by asking participants to agree or
in an INDEPENDENT VARIABLE. For exam-
disagree with each statement. To get an
ple, a drug may have no effect at all until
individual's total scale score, the scale
a certain dosage level (the threshold
scores of all the items that the person
value) is reached. The change in the de-
agreed with are averaged. [Louis Leon
pendent variable can be either an irre-
Thurstone (1887-1955), U.S. psycholo-
versible dichotomous change, for
gist]
example, from "functional" to "not
functional," or it can be continuous,
such that the dependent variable t i e d ranks in a NONPARAMETRIC TEST
changes each time the independent vari- that involves ranking data, the ranks as-
able crosses the threshold in either direc- signed to two or more data points that
tion. have the same values. Such data are usu-
ally given a rank that is the mean of their
threshold model any model specify- positions in the ascending order of
ing that there is a value in a range of val- ranks. For example, suppose the data re-
ues on an INDEPENDENT VARIABLE at veal that Case 1 has the lowest value, fol-
which the DEPENDENT VARIABLE (e.g., lowed by Cases 2, 3, and 4, each of

381
time-constant covariate

which yields the same value; Cases 5 and ple, a time-lag study of intelligence
6 have the next-to-highest and highest might compare a group of people who
values, respectively. Here Cases 2, 3, and were 20 years old in 2005 with groups
4 can be awarded the tied rank of 3, as who were 20 years old in 2006, 2007,
this is the mean value of 2 + 3 + 4. and 2008. Time-lag designs have the
benefit of controlling for time-of-testing
time-constant covariate see TIME-
effects. Thus, although no differences
INDEPENDENT COVARIATE.
between ages are examined, the re-
time-dependent covariate a PREDIC- searcher can determine differences due
TOR VARIABLE whose values change over to changes in the environment over
time or the course of the study for an in- time. However, there are drawbacks of
dividual (e.g., blood pressure, dose of low INTERNAL VALIDITY and the diffi-
drug). Time-dependent variables are rel- culty in separating COHORT EFFECTS
evant in SURVIVAL ANALYSIS. Also Called f r o m AGE EFFECTS.
time-varying covariate. Compare
time-lagged correlation the correla-
TIME-INDEPENDENT COVARIATE. tion of a measure at one point in time
time-dependent Cox regression with the value of that same measure at a
model see cox REGRESSION ANALYSIS. different point in time. An example is
the correlation of IQ scores of individu-
time-homogeneous Markov chain als at 5 years of age with their IQ scores
a TIME-SERIES model in which an event's when they are 10 years of age. See STA-
probability is dependent only upon the BILITY COEFFICIENT.
immediately preceding event in the se-
ries and the transition probabilities do time-lag study a LONGITUDINAL DE-
not change over time (i.e., the probabil- SIGN in which participants are measured
ity of going from State A to State B today on two variables at two or more different
is the same as it will be at any time in points of time, with a view to determin-
the future). Also called time-homoge- ing whether one is more likely to
neous stationary chain. See MARKOV "cause" the other. Ideally, the measure-
CHAIN.
ments of the variables are separated by
the length of time it takes for the cause
time-independent covariate a PRE- to influence the effect. There are limi-
DICTOR VARIABLE that is measured at the tations to the procedure for inferring
outset of a study and whose values do causality, particularly the OMITTED-
not change over time or the course of VARIABLE BIAS or the THIRD-VARIABLE
the srady for an individual (e.g., gen- PROBLEM.
der, year of birth). Time-independent
covariates are relevant in SURVIVAL time-location s a m p l i n g a method of
ANALYSIS. Also Called time-constant finding research participants in which
covariate. Compare TIME-DEPENDENT members of a hard-to-reach target popu-
lation (e.g., homeless persons, migrant
COVARIATE.
workers) are recraited from specific loca-
time-lag design a type of QUASI- tions at which they may be found dur-
EXPERIMENTAL RESEARCH in which par- ing specific time periods when they
ticipants of the same age are compared are likely to be present. The sample is se-
at different time periods. It is typically lected in stages: (a) The SAMPLING
used in developmental, educational, FRAME is determined, comprising all of
and social psychological research to the locations at which there is sufficient
srady whether there are differences in a attendance by persons in the population
given characteristic for samples of equal of interest to make sampling worth-
age but drawn from different cohorts while, (b) A random sample of locations
measured at different times. For exam- is chosen from this frame. If attendance

382
 Tobit analysis

depends on day of the week and time of obtained are age-related only). In a time-
day, a sampling period is then chosen sequential design a second age group is
for each location in the sample, (c) A added to a TIME-LAG STUDY, and two or
sample of participants is chosen, usually more cross-sectional comparisons are
randomly, during each sampling event. made at different times of testing.
Also called time-location cluster
sampling; time-space sampling; time series a set of measures on a single
venue-based sampling; venue sam- attribute, variable, or constract obtained
pling; venue time-based sampling. repeatedly over time.
time-series analysis a branch of statis-
time-reversible M a r k o v c h a i n a tics that involves the analysis of changes
MARKOV CHAIN in whlch it is not possi- in a single variable recorded repeatedly
ble to determine, given the states at a over time. The data may have an inter-
number of points in time after ranning nal stracture (such as AUTOCORRELA-
the STOCHASTIC PROCESS, which state TION, trend, or seasonal variation) that
came first and which state arrived later. should be accounted for and that pro-
time s a m p l i n g in DIRECT OBSERVA- vides input allowing for the prediction
TION, a data collection strategy that in- of furare values of the variable. Compare
volves noting and recording the CROSS-SECTIONAL ANALYSIS.
occurrence of a target behavior when- time-series data a sequence of mea-
ever it is seen during a stated time inter- surements taken at successive times sep-
val. The process may involve fixed time arated by specified tim^ intervals. The
periods (e.g., evety 5 minutes) or ran- data have a natural temporal ordering
dom time intervals. For example, a re- (e.g., chronological order).
searcher may observe a group of
children for 10 seconds evety 5 minutes time-series design an experimental
for a specific 30-minute period each day, design that involves the observation of
noting the occurrence or nonoccurrence units (e.g., people, countries) over a de-
of particular behaviors (overt actions). fined time period. Data collected from
Observations taken during these periods such designs may be evaluated with
are known as time samples. An individ- TIME-SERIES ANALYSIS.
ual score is made on the basis of one or time-series plot a graph on which
more of the following: (a) number of TIME-SERIES DATA are connected by
time units in which the defined behav- straight lines, used to show the changes
ior occurs, (b) total frequency of occur- of the measured variable (y-axis) relative
rence of the defined behavior in the to the different measurement times (x-
total observational time, and (c) average axis).
frequency of the defined behavior per
unit of time. See PARTIAL-INTERVAL RE- time-varying covariate see TIME-
CORDING; WHOLE-INTERVAL RECORD- DEPENDENT COVARIATE.
ING. Tobit analysis a type of REGRESSION
time score a score based on the amount ANALYSIS used when a DEPENDENT VARI-
ABLE with values above or below a cer-
of time used to , complete a particular
task. An example is the number of min- tain threshold takes on the value of that
utes a 3-year-old child requires to solve a threshold; that is, the exact value of the
simple puzzle. variable is unknown or ignored (see CEN-
SORED DATA). For exarnple, in an analy-
time-sequential design an experi- sis of aptitude in coUege students, the
mental design to separate AGE EFFECTS dependent variable could be a score on a
from time of measurement and COHORT standardized aptitude test that has an
EFFECTS (i.e., to determine if the results upper limit of 500 (the threshold value).

383
tolerance

Sradents who answer all questions on will be computed to describe the sample,
the test correctly receive a score of 500, such as the MEAN. Relatedly, each indi-
even though these sradents may not vidual condition has its own DEGREES OF
have equal aptitude. In such a case the FREEDOM as well: 20 - 1 = 19.
influence of INDEPENDENT VARIABLES,
t o t a l effect in the study of causal ef-
such as reading and math scores, on the
fects, the total extent to which the de-
dependent variable of academic apti-
pendent (or outcome) variable is
rade is more appropriately sradied with
changed by the independent (or predic-
Tobit regression analysis than with ordi-
tor) variable, including any indirect ef-
nary LEAST SQUARES REGRESSION be-
fect through a MEDIATOR. In a simple
cause the threshold values would bias
example, if the independent variable, x,
the slope of the obtained REGRESSION
is presumed to cause the outcome vari-
LINE. Also called censored regression;
able, y, the PATH COEFFICIENT of this di-
Tobit model. [James Tobin (1918-
rect effect. A, is the total effect. If there is
2002), U.S. economist]
an intervening variable, linked by two
tolerance n. permissible or allowable path coefficients, B and C, this indirect
deviation from a specified value or stan- effect is BC, and the total effect is A + BC.
dard. tolerant adj. See also PATH ANALYSIS.
tolerance i n t e r v a l a range of values total p r o b a b i l i t y l a w a law relating
within which, with some probability, a to BAYES THEOREM and conditional
specified proportion of a population probabilities, stating that, for two
falls. For instance, one may be 95% con- events, A and B, the probability of A oc-
fident that 90% of the population will curring is equal to the probability that
fall within the range specified by the tol- both A and B occur plus the probability
erance interval. It differs from a CONFI- that A occurs and B does not.
DENCE INTERVAL, which bounds a single total sum of squares (symbol: T55) in
population parameter (e.g., the mean or ANALYSIS OF VARIANCE and REGRESSION
variance). ANALYSIS, the SUM OF SQUARES due to
tolerance l i m i t the upper or lower the effects of treatment plus the sum of
value of a TOLERANCE INTERVAL. squares due to error.
topastic error o f measurement the t o t a l variance in ANALYSIS OF VARI-
partly random, partly systematic error or ANCE and REGRESSION ANALYSIS, the
variance that results when the taker of a variability that is due to the effects of
multiple-choice test has the opportunity treatment (TRUE VARIANCE) plus the
to get some of the answers correct by variability that is due to error (ERROR
VARIANCE).
guessing.
top-down clustering (top-down hi- trace n. in multivariate statistics and lin-
ear algebra, the sum of the elements on
erarchical clustering) see DIVISIVE
the MAIN DIAGONAL of a SQUARE MA-
CLUSTERING.
TRIX.
t o t a l degrees of freedom the total
t r a i n i n g sample a representative sam-
number of observations in an analysis
ple of a larger population that is typi-
minus one. For example, in an experi-
cally used to develop a model or to
ment in which there are four conditions,
obtain conditional probabilities in
with 20 participants randomly assigned
to each condition, there are 80 inde- BAYESIAN analysis.
pendent observations: The total degrees t r a i n i n g set a portion of a TIME SERIES
of freedom is 79 (80 -1) observations, as used to predict furare values, that is,
there is one constraint on any value that to train (fit) a model for prediction.

384
 transformed score

Training sets are used in artificial intel- different traits. Also called aptitude-
ligence (e.g., for neural networks), sta- treatment interaction (ATI).
tistical modeling, and other areas of in-
formation science. t r a i t v a l i d i t y the degree to which a test
is capable of measuring a trait. A test has
t r a i t n. 1. an enduring personality char- trait validity if it demonstrates an associ-
acteristic that describes or determines an ation between the test scores and the
individual's behavior across a range of prediction of an underlying trait. See
situations. 2. in ITEM RESPONSE THEORY, also MULTITRAIT-MULTIMETHOD MATRIX.
an individual's level of competence on a
certain task or aptitude measurement. transfer f u n c t i o n model in TIME-
SERIES ANALYSIS, atypeof model used to
t r a i t p r o f i l e a graphic display of test forecast a time series that is influenced
scores in which each score represents an by present and past values of other time
individual TRAIT. These scores or ratings series.
are often arranged on a common scale to
enable them to be interpreted quickly. t r a n s f o r m a t i o n n. the conversion of
The hypothetical example below shows data to a different form through a rule-
a respondent's results (given in T SCORES) based, usually mathematical process, for
on a personality quesfionnaire assess- example, changing Fahrenheit to Cel-
ing neuroticism, extraversion, openness, sius. In statistics, a RAW SCORE is of-
agreeableness, and conscientiousness. ten transformed into a STANDARDIZED
Also called psychogram. See TEST PRO- SCORE for purposes of comparison. See
FILE. also LINEAR TRANSFORMATION; NONLIN-
EAR TRANSFORMATION. transform
t r a i t r a t i n g a technique in which a vb. transformational adj.
given behavioral feature or trait (e.g., a
character trait or attribute) is observed, t r a n s f o r m e d score a score that has
rated, and recorded. been changed into another scale to
allow direct comparison and meaning-
trait-treatment interaction (TTI) ful interpretation with other scores. For
the interaction between traits (e.g., gen- example, a RAW SCORE of 44 on a first
der, aptitude) and conditions of treat- test might not mean the same thing as a
ments (e.g., methods of instruction) as it raw score of 44 on the second test. The
affects a dependent variable (e.g., scores transformation could be to percentages:
on an academic test). Using ANALYSIS 44 on the first test could be out of 100
OF VARIANCE Or MULTIPLE REGRESSION items, and thus represent 44%, while on
analysis, it is possible to determine the the second test it could be out of 50
best type of treatment for people with items, and thus have a transformed

llll
E O
Trait
A C

trait profile

385
transient state

Subsequent behavior
Initial behavior 1 2 3 4 5 6 7 8 9 10
I. Change talk .24 .06 .16 .11 .49 .36 .04 .04 .01 .30
2. Question positive .15 .15 .17 .30 .32 .25 .02 .04 .02 .25
3. Question negative .06 .02 .08 .10 .22 .09 .02 .01 .01 .32
4. Question neutral .12 .02 .36 .32 .15 .05 .01 .01 .01 .02
5. Simple reflection .09 .05 .28 .02 .10 .28 .00 .02 .00 .02
6. Complex reflection .26 .36 .32 .02 .12 .31 .00 .00 .00 .02
7. Raise concem .04 .06 .40 .10 .04 .02 .00 .03 .02 .10
8. Give affirmation .10 .06 .26 .09 .03 .02 .00 .01 .03 .09
9. Give advice .07 .03 .01 .01 ,02 .01 .01 .01 .00 .11
10. Other .06 .01 .03 .02 .02 .04 .01 .01 .00 .50

transition matrix

score of 88%. Similarly, STANDARDIZED t r a n s i t i o n p r o b a b i l i t y the probabil-

SCORES also can be transformed. For ex- ity of moving from one state of a system
ample, a z SCORE can be transformed to into another state. If a MARKOV CHAIN is
a T SCORE. in state /, the transition probability, is
the probability of going into state / at the
transient state in MARKOV CHAIN next time step.
analysis, a state (i) that can be left for an-
other state (j) from which there is no translation and back-translation
possible return to /. Compare ABSORB- a method of ensuring that the transla-
ING STATE. tion of an assessment instrument into
another language is adequate, used pri-
t r a n s i t i o n m a t r i x a SQUARE MATRIX marily in cross-cultural research. A biUn-
used to describe the transitions of a gual person translates items from the
MARKOV CHAIN and the probabilities of source language to the target language,
moving from one state to another. Each and a different bUingual person then in-
row contains the probabilities of moving dependently translates the items back
from the state represented by that row to into the source language. The researcher
the other states. can then compare the original with
For example, a researcher might con- the back-translated version to see if any-
struct a transition matrix similar to the thing important was changed in the
hypothetical example above to repre- translation.
sent conversational exchange during a
therapy session. transpose n. in matrix algebra, a matrix
The numbers give the probabilities formed by interchanging the rows and
that if the speech act iri the row category columns of the original matrix.
occursfrom either the therapist or the
original transpose
clientthe response by the other person
will be the speech act in the column cat- (X) (X^
egory. Also called Markov matrix; 12 3 14 7
4 5 6 2 5 8
probability matrix; substitution 7 8 9 3 6 9
matrix.
t r a n s i t i o n model a model represent- t ratio the formula used for the TTEST, in
ing the TRANSITION PROBABILITIES frOm which the numerator is the difference
one state to another, thereby defining between the two means of the groups
the possible changes in state for the and the denominator is a measure of the
events being studied. Also called state DISPERSION of the scores. The t value is
transition model. See also MARKOV compared to a T DISTRIBUTION table to
CHAIN. determine if the difference between the

386
 treatment variability

means is significant, or likely to have one treatment better than another? If a

been a chance finding. treatment is effective, do some levels of
the treatment produce better outcomes
t r e a t m e n t n. the intervention to which
than others? Are the benefits of treat-
some participants in an experimental
ment worth the cost? See also THERAPY
design (the EXPERIMENTAL GROUP or
OUTCOME RESEARCH.
treatment group) are exposed, i n con-
trast to a CONTROL GROUP, who do not t r e a t m e n t p o p u l a t i o n the larger
receive the intervention. Also called group to which the results obtained
t r e a t m e n t c o n d i t i o n . See TREAT- from the E X P E R I M E N T A L G R O U P (or
M E N T LEVEL. treatment group) in a study will be gen-
eralized.
treatment-by-blocks d e s i g n see
RANDOMIZED BLOCK DESIGN. t r e a t m e n t process r e s e a r c h re-
search, usually of a cUnical narare, that
t r e a t m e n t - b y - s u b j e c t s d e s i g n see
investigates how interventions activate
WITHIN-SUBJECTS DESIGN.
mechanisms of behavior change. See
t r e a t m e n t c o m b i n a t i o n 1. the par- also T R E A T M E N T O U T C O M E R E S E A R C H .
ticular combination of treatments ad-
treatment-received analysis assess-
ministered to a participant in a study. 2.
ment of the outcome of a RANDOMIZED
the combination of levels of different
CLINICAL TRIAL that focuses only on par-
FACTORS in a F A C T O R I A L D E S I G N .
ticipants who acraally received the treat-
t r e a t m e n t c o n d i t i o n see TREATMENT. ment, rather than those to whom it
was prescribed (see INTENTION-TO-TREAT
t r e a t m e n t e f f e c t the magnirade of the
ANALYSIS). Also caUed efficacy subset
effect of a treatment (i.e., the, I N -
analysis.
DEPENDENT V A R I A B L E ) upon the re-
sponse variable (i.e., the DEPENDENT t r e a t m e n t s - b y - s u b j e c t s a n a l y s i s o f
VARIABLE) in a srady. It is usually mea- v a r i a n c e see WITHIN-SUBJECTS A N A L Y -
sured as the difference between the level SIS OF VARIANCE.
of response under a control condition
t r e a t m e n t t r i a l a research study de-
and the level of response under the treat-
signed to evaluate the effectiveness of an
ment condition in standardized units.
experimental intervention or procedure,
See EFFECT SIZE.
its possible adverse effects, and other in-
t r e a t m e n t g r o u p see EXPERIMENTAL formation that would contribute to the
GROUP; TREATMENT. decision to use the procedure in the fu-
rare. During the trial the experimental
t r e a t m e n t l e v e l the specific condition
procedure is compared with an existing
to which a group or participant is ex-
one, which acts as a control. When the
posed in a study or experiment. For ex-
intervention being evaluated is a new
ample, in a design employing four
drug the term CLINICAL TRIAL is used in-
groups, each of which is exposed to a dif-
stead.
ferent dosage of a particular drag, each
dosage amount represents a level of the t r e a t m e n t v a l i d i t y the extent to
treatment factor. which an instrament is of value in iden-
titying those individuals who are likely
t r e a t m e n t o u t c o m e r e s e a r c h re-
to benefit from a particular treatment or
search designed to evaluate the efficacy
intervention. The term is used particu-
of interventions and to investigate the
larly in the field of special needs educa-
mechanism by which effective interven- tion.
tions produce change. It is designed to
answer such questions as the following: t r e a t m e n t v a r i a b i l i t y the degree to
Is treatment better than no treatment? Is which the response of participants to

387
treatment variable

Successful Correct -
retrieval recall

Sufficient
storage
1-9,

Unsuccessful. Incorrect
retrieval recall
1-e.

Insufficient Incorrect
storage recall

tree diagram

a treatment differs according to the do- dratic increases in response to increases

main being assessed, the setting in in the level of an INDEPENDENT VARI-
which the treatment occurred, and the ABLE (e.g., increased dosage levels). Such
intensity of the treatment. See GEN- analysis is often used to predict future
ERALIZABILITY. events. Also called trend test. See CHI-
SQUARE TEST FOR TREND; COX-STUART
treatment variable the INDEPENDENT TEST FOR TREND; CUZICK'S TREND TEST.
VARIABLE, whose impact on a DEPEN-
DENT VARIABLE is Studied in a research t r e n d correlation coefficient a
project. CORRELATION COEFFICIENT that de-
scribes a trend in TIME-SERIES DATA. If
tree d i a g r a m 1. a diagram for generat- there is a positive correlation, then the
ing and depicting a probability distribu- trend is for increases in the variable of
tion. It shows all the possible outcomes interest over time; a negative correlation
of an event and is used to determine the represents a decrease in the variable over
probability of getting specific results time.
where the possibilities are nested. For ex-
ample, consider an experimental mem- t r e n d line a line on a graph that depicts
oty task in which participants view the overall pattern of relationship be-
multiple target stimuli one at a time and tween variables. See also TREND ANALY-
then determine whether subsequently SIS.
presented cues are targets or distractors. t r e n d study a LONGITUDINAL DESIGN in
The researcher may use a simple tree dia- which data are collected at periodic in-
gram similar to the one above to illus- tervals on samples drawn from a particu-
trate the recall process. lar population and used to reveal trends
Here, 9, represents the proportion of (systematic tendencies or patterns) over
recall trials where there is sufficient stor- time.
age and, across all such trials in which
there is sufficient storage, 9^ represents t r e n d test see TREND ANALYSIS.
the proportion of times with successful t r i a l n. 1. in testing, conditioning,
retrieval. 2. see DENDROGRAM. 3. more or other experimentation, one perfor-
generally, any branching depiction of a mance of a given task (e.g., one run
process or condition. A DECISION TREE is through a maze) or one presentation of a
an example. stimulus (e.g., an ordered list of three-
letter words). 2. see CLINICAL TRIAL;
t r e n d analysis any of several analytic
TREATMENT TRIAL.
techniques designed to uncover system-
atic patterns (trends) in a set of variables, t r i a l design the strategy or design used
such as linear growth over time or qua- to conduct a CLINICAL TRIAL for the pur-
 true variance

pose of evaluating the efficacy of a new t r i m m e d mean a mean calculated by

treatment. averaging the scores in a distribution
after removing equal numbers of the
t r i a n g u l a r contingency table a spe-
highest and lowest values. For example,
cial class of incomplete CONTINGENCY
a researcher may decide to exclude the
TABLE that contains stracraral zeros in
top and bottom 10% of the distribution;
one or more cells above or below the
the mean is then calculated on the 90%
MAIN DIAGONAL.
remaining scores.
t r i a n g u l a r m a t r i x a SQUARE MATRIX
t r i m m i n g n. 1. the exclusion of a fixed
in which the values of all numbers either
percentage of cases at each end of a dis-
above or below the MAIN DIAGONAL are
tribution before calculating a statistic on
zero (a lower triangular matrix or upper
the set of data. This is done to eliminate
triangular matrix, respectively).
the influence of extreme scores on the
t r i a n g u l a r test a test involving SE- estimate. See TRIMMED MEAN. 2. in PATH
QUENTIAL ANALYSIS that can be stopped ANALYSIS, the removal of nonsignificant
when a particular POWER is reached and paths after the first ran of analyses. The
results can be evaluated. It is typically analysis is then reran unfil the best fitting
used in clinical trials when it is difficult model is obtained. The ultimate model
to recruit participants. should then be CROSS-VALIDATED on
another data set.
t r i a n g u l a t i o n n. the process of con-
firming a hypothesis by collecting evi- t r i p l e b l i n d see BLIND.
dence from multiple sources. There are
true experiment a srady in which par-
several different types of triangulation.
ticipants are assigned at random to two
In data triangulation various sampling
or more experimentally manipulated
methods are used: The data are collected
treatment conditions or to a treatment
at different times, from different groups
group and a control group. Thistypeof
of people, and so forth. Investigator tri-
experiment is in contrast to QUASI-
angulation involves multiple researchers
E X P E R I M E N T A L D E S I G N S , such as NATURAL
in an investigation; theory triangulation
involves using more than one theoreti- EXPERIMENTS and FIELD EXPERIMENTS.
cal scheme to interpret the phenome- true-false test a test in which the par-
non; and METHODOLOGICAL TRIANGU- ticipant must respond to statements,
LATION involves using interviews, obser- words, and the like with either "true" or
vations, questionnaires, documents and "false."
other data-collection methods.
true score in CLASSICAL TEST THEORY,
triggered causal variable an INDE- that part of a measurement or score that
PENDENT VARIABLE, causally related to reflects the actual amount of the attrib-
the DEPENDENT VARIABLE, that does not ute possessed by the individual being
produce any change in the dependent measured.
variable unless preceded by another
causal variable. true variance naturally occurring vari-
ability within or among research partici-
t r i g r a m n. any three-letter combina- pants. This VARIANCE is inherent in the
tion, particularly a nonsense syllable nature of the participant and is not due
used in studies of learning and memoty. to measurement error, imprecision of
2. in sradies of language processing, a se- the model used to describe the variable
quence of three words, syllables, or of interest, or other extrinsic factors.
other items in which the identity of the It represents the variance of the TRUE
first two items is used as a basis for pre- SCORES among the participants taking
dicting the third. the measure.

389
true zero

true zero see ABSOLUTE ZERO. TTI abbreviation for TRAIT-TREATMENT

truncated data a set of data in which INTERACTION.
some values are excluded as a matter of T u k e y - K r a m e r procedure an exten-
deliberate selection. For example, if the sion of TUKEY'S HONESTLY SIGNIFICANT
distribution of age being studied focuses DIFFERENCE TEST (HSD test) SO that it can
on 21 to 65 years of age, those under 21 be used with samples of unequal size.
and over 65 are excluded from the analy- Whereas the HSD test computes a single
ses. Trancated data can be contrasted to CRITICAL DIFFERENCE (CD) for each pair
CENSORED DATA, in which certain values of means, the Tukey-Kramer test uses a
in the data sample are unknown owing different CD as riequired to evaluate the
to some random cause. significance of the difference between
truncated f l i s t r i b u t i o n a set of scores each pair of means. Qohn Wilder Tukey
lacking values beyond a specific maxi- (1915-2000 and Clyde Y. Kramer, U.S.
mum point, below a specific minimum statisticians]
point, or both. See TRUNCATED DATA. Tukey l a m b d a d i s t r i b u t i o n a CON-
truncated Poisson d i s t r i b u t i o n a TINUOUS DISTRIBUTION that Is defined
POISSON DISTRIBUTION in which a value by a particular shape parameter known
is excluded. For example, this may occur as lambda (k): It is used mainly to gener-
when the count variable cannot take on ate a CORRELATION COEFFICIENT for a
the value zero (zero-truncated Poisson PROBABILITY PLOT of a data set. On the
distribution). basis of the coefficient, a model can be
suggested for the set; in particular, it
T-scope see TACHISTOSCOPE. can be used to identity whether a distri-
T score a STANDARDIZED SCORE based on bution has a long or a short tail. This
a score distribution that has a mean of technique is not appropriate for a distri-
50 and a STANDARD DEVIATION of 10. bution (such as the normal distribution)
For example, a RAW SCORE that is 1 stan- that has only location and scale parame-
dard deviation above its mean would be ters and no shape parameter. [John Wilder
converted to a T score of 60. See TRANS- Tukey]
FORMED SCORE. T u k e y line a procedure for fitting a
T-squared test see HOTELLING'S T"" straight line to data that reduces the im-
TEST. pact of OUTLIERS. The data are divided
into three groups and the fitted line is
TSS symbol for TOTAL SUM OF SQUARES. determined from the group medians.
t test any of a class of statistical tests [John Wilder Tukey]
based on the fact that the test statistic
Tukey q u i c k test a NONPARAMETRIC
follows the T DISTRIBUTION when the
test used to compare two independent
NULL HYPOTHESIS Is true. Most t tests
samples and to test the NULL HYPOTHE-
deal with hypotheses about the mean of
SIS of no differences in population medi-
a population or about differences be-
ans or means. The test does not require
tween means of different populations,
that any special assumptions are met.
where the populations show NORMAL
Also called Tukey pocket test. [John
DISTRIBUTIONS and the variances are
Wilder Tukey]
unknown and need to be estimated. The
test can be used with independent Tukey's honestly s i g n i f i c a n t dif-
groups (e.g., test scores of those who ference test (Tukey's HSD test) a
were given training vs. a control group MULTIPLE COMPARISON procedure that
without training) or dependent groups is used to test for significant differences
(e.g., test scores before vs. after training). between all possible pairs of mean values
Also called Student's t test. on a variable for groups of research par-

390
 two-by-two factorial design

ticipants; it is generally applied after an cal but as there is no shared environ-

ANALYSIS OF VARIANCE has determined ment any disparity in intelligence must
that there is a significant difference result from the different environments,
among three or more means. The proce- (b) Comparisons between identical
dure simultaneously compares all possi- twins reared together and fraternal twins
ble pairs of means based on a single reared together. Here one can assume
quantity, called the honestly significant that each pair of twins shares the same
difference (HSD), such that if the differ- environment, but while the identical
ence between any two group means twins have 100% of their genes in com-
exceeds the HSD the corresponding pop- mon, the fraternal twins share only 50%
ulation means are said to be signifi- of their genes. The assumptions made in
cantly different from each other as well. these studies are, however, never com-
Tukey's HSD test preserves the FAMILY- pletely fulfilled. For example, the identi-
WISE ERROR RATE by adjusting the over- cal twins reared apart have had some
all SIGNIFICANCE LEVEL to take into ac- common environment, if only their
count the fact that multiple T TESTS are intrauterine experiences. Moreover,
being conducted. Also called Tukey's identical twins reared together usually
range test. [John Wilder Tukey] have more similar environments than
fraternal twins raised together. These
Tukey test o f a d d i t i v i t y an approach differences can make the estimations of
used in TWO-WAY ANALYSIS OF VARI- heritability of intelUgence open to some
ANCE to assess whether the independent doubts.
variables are additively related to the EX-
PECTED VALUE of the dependent vari- two-bend t r a n s f o r m a t i o n the use of
able, or whether there is an INTER- a TRANSFORMATION to stretch out the
ACTION EFFECT. Also called Tukey's tails of a distribution, on a graph, when
one-degree-of-freedom test; Tukey the relationship between two variables
test of nonadditivity. [John Wilder produces an S-shaped line, thus elimi-
Tukey] nating the two bends in the line.
twin control in a TWIN STUDY, a The ARC SINE TRANSFORMATION, LOGIT
method in which the target twinthat transformation, and PROBIT TRANSFOR-
is, the one who has had certain experi- MATION can be used for this purpose. See
ences or training or has been exposed to also ONE-BEND TRANSFORMATION.
the experimental conditionsis com-
two-by-two chi-square test see
pared against the twin who has not had
the experiences, training, or treatment TWO-WAY CHI-SQUARE TEST.
and therefore serves'as a CONTROL. Also two-by-two contingency table see
called cotwin control. TWO-BY-TWO TABLE.
t w i n study research utilizing twins. The two-by-two crossover design see
purpose of such research is usually to as- CROSSOVER DESIGN.
sess the relative contributions of hered-
ity and environment to some attribute. two-by-two f a c t o r i a l design an ex-
Specifically, twin studies often involve perimental design in vvhich there are
comparing the characteristics of identi- two INDEPENDENT VARIABLES each hav-
cal and fraternal twins and comparing ing two levels. When this design is de-
twins of both types who have been picted as a matrix, two rows represent
reared together or reared apart. For ex- one of the independent variables and
ample, two types of srady have been two columns represent the other inde-
used to investigate intelligence in twins: pendent variable. Also called two-by-
(a) Identical twins reared apart. Here the two design; two-way factorial de-
genotypes (genetic makeups) are identi- sign. See FACTORIAL DESIGN.

391
two-by-two table

two-by-two table a type of TWO-WAY ITEM DISCRIMINABILITY. See alSO RASCH

TABLE used to display and analyze data MODEL; THREE-PARAMETER MODEL.
for two DICHOTOMOUS VARIABLES. For
example, suppose a survey of a group of two-period crossover design see
CROSSOVER DESIGN.
100 participants reported information
on two variables: (a) gender (male or fe- two-sample runs test see W A L D -
male) and (b) major in college (social sci- W O L F O W I T Z TEST.
ences or humanities).
two-sample test any procedure in
Major Male Female Total
which data are collected on two samples
and then subjected to a test for SIGNIFI-
Humanities 20 40 60 CANT DIFFERENCES between the two
samples. Compare SINGLE-SAMPLE TEST.
Social sciences 15 25 40
two-sample t test a type of T TEST in
The results could be shown in a two-which the mean value on a variable ob-
tained by one group is compared to the
by-two table similar to the one above.
Also called fourfold table; two-by- mean value obtained by another distinct
two contingency table. group. In other words, two discrete ex-
perimental groups are evaluated against
two-factor analysis o f variance see one another. By contrast, in the SINGLE-
TWO-WAY ANALYSIS OF VARIANCE. SAMPLE T TEST, the results of only one
two-factor design see TWO-WAY DE- experimental group are compared to
SIGN. some standard of reference.
two-mode data information that re- two-sigma rule a rule of thumb stating
cords patterns of relationship among that in a NORMAL DISTRIBUTION approx-
two classes of entities (typically actor imately 95% of the values lie within two
and event) in such a way that it is possi- STANDARD DEVIATIONS (or sigmas) of
ble to describe ties between two sets of the mean. Thus, if the mean is 50 and
nodes at two different levels of analysis. the standard deviation is 5, then approx-
For example, data can be coUected on imately 95% of the scores will lie be-
which of 50 women were present at tween the values of 40 and 60 (i.e., 50
events sponsored by the local rotary club +/- 2 standard deviations or 10 points).
(e.g., 15 events) during the course of the two-stage least squares regression
social season in a community. By exam- an extension of ordinary LEAST SQUARES
ining patterns of which women were REGRESSION (OLS) to cover models that
present (or absent) at which events, it violate some of the assumptions of
would be possible to infer an underlying OLS, such as correlation among ERROR
pattern of social ties, factions, and TERMS, error in the dependent variable,
groupings among the womeri. This reciprocal causality, or the correlation of
would be one mode of analysis. At the one or more independent variables with
same time, by exarnining which events UNMEASURED VARIABLES affecting the
were attended by which women, it dependent variable. In the first stage,
would be possible to infer underlying the dependent variable is regressed on
patterns in the similarity of the events. all of the independent variables in the
This would be a second mode of analysis. model and the predicted values from the
two-parameter model in ITEM RE- regression are obtained. The purpose of
SPONSE THEORY, a model that specifies the first stage is to create new dependent
two parameters affecting an individual's variables that do not violate the assump-
response to a particular test item: (a) the tions of OLS regression. In the second
difficulty level of the item and (b) the stage, a standard OLS regression is car-
 two-way mixed design

tied out using the predicted values from DIRECTIONAL HYPOTHESIS). For a signifi-
the first-stage regression analysis in cance value of .05, .025 marks the CRITI-
place of the original dependent variable. CAL REGION at each end of the
The new dependent variable will be un- distribution.
correlated with the error term of the two-tailed test see NONDIRECTIONAL
original dependent variable because it is TEST.
a function of the independent variables.
two-way analysis o f covariance an
two-stage s a m p l i n g see MULTISTAGE ANALYSIS OF COVARIANCE in which
SAMPLING. there are two INDEPENDENT VARIABLES
two-Stage stopping rule in a CLINI- and a COVARIATE whose effects the re-
CAL TRIAL comparing two treatments, a searcher wishes to bring under statistical
strategy in which results are examined control.
after only a fraction of the planned two-way analysis o f variance an
number of participants in each group ANALYSIS OF VARIANCE that Isolates the
have completed the trial (usually either MAIN EFFECTS of two Independent vari-
half or two thirds of the patients). At this ables, a and b, and their INTERACTION
point, the test statistic is computed and EFFECT, a X b, on a dependent variable.
the trial stopped if the difference be- Also called two-factor analysis of
tween treatment means is significant at variance.
the SIGNIFICANCE LEVEL Set for this first
stage. If not, the remaining participants two-way chi-square test a CHI-
in each group are studied, the test statis- SQUARE TEST used to determine whether
tic is recomputed, and the means com- there is a significant relationship be-
pared at a significance level set for the tween the variables summarized in a
second stage. The significance levels for TWO-WAY TABLE. Also called two-by-
the two stages should be such that they two chi-square test.
equal the overall significance level for two-way design atypeof FACTORIAL
the trial, usually .05 or .01. DESIGN in which two INDEPENDENT
VARIABLES are manipulated. Also called
two-tailed alternative hypothesis two-factor design.
see N O N D I R E C T I O N A L H Y P O T H E S I S .
two-way f a c t o r i a l design see TWO-
two-tailed confidence i n t e r v a l in BY-TWO FACTORIAL DESIGN.
statistical testing, a CONFIDENCE INTER-
VAL that specifies both upper and lower two-way interaction in a TWO-WAY
limits to the population parameter (see ANALYSIS OF VARIANCE, the jolnt effect
CONFIDENCE LIMIT). of both independent variables, a and b,
on a dependent variable. A statistically
two-tailed hypothesis see NON- significant two-way interaction indi-
DIRECTIONAL HYPOTHESIS. cates that there are differences in the in-
two-tailed p value in a NONDIREC- fluence of each independent variable at
TIONAL TEST of significance, a probability
their different levels (e.g., the effect of]
value for making a TYPE TERROR that and at b^ is different from the effect of
falls below the SIGNIFICANCE LEVEL for fll and fl23t b2). See also HIGHER ORDER
the test, leading to the rejection of the INTERACTION.
NULL HYPOTHESIS. In Other words, as- two-way m i x e d design an experi-
suming the null hypothesis to be true, mental design in which there are two in-
the p value is the probability that ran- dependent variables of interest, where
domly selected samples would have one variable is a BETWEEN-SUBJECTS FAC-
means that are different, with either sam- TOR (e.g., gender) and the other variable
ple having the larger mean (see NON- is a WITHIN-SUBJECTS FACTOR (e.g., pre-

393
two-way repeated measures design

test and posttest scores for the same par- called the POWER of the test, where
ticipants). Data from such designs may power =1-3. Also called beta error.
be evaluated with a two-way mixed-
design analysis of variance. Type III error 1. the error that occurs
when there is a discrepancy between the
two-way repeated measures design research focus and the hypothesis acra-
see TWO-WAY WITHIN-SUBJECTS DESIGN. ally tested. For example, a Type 111 error
two-way table a table in which the would be committed if a researcher col-
joint FREQUENCY DISTRIBUTION Of tWO lected data on INDIVIDUAL DIFFERENCES
within a sample and determined the
INDEPENDENT VARIABLES is arrayed. See
causes of variation when the question of
also TWO-BY-TWO TABLE.
interest concerned differences between
two-way within-subjects design a populations. In other words, a Type III
type of WITHIN-SUBJECTS DESIGN in error involves providing the right an-
which there are two variables or factors swer for the wrong question. Also called
of interest, a and b, and the same partici- Type 0 error. 2. the error that occurs
pants are studied in each of the levels of when a researcher correctly rejects the
each of the factors. For example, sup- NULL HYPOTHESIS of no difference be-
pose a and b representtypeof training tween samples but then makes an incor-
(lecture vs. reading) and type of content rect inference about the direction of the
(physics vs. psychology), respectively. difference. Researchers investigating the
All participants receive all treatment direction of a relationship (e.g., "Which
conditions of both factors; that is, all is more?" or "Which is better?") will
participants receive bothtypesof train- make a Type III error if they use a NON-
ing and are exposed to both contents. DIRECTIONAL TEST to make a directional
Data from such designs may be evalu- decision. 3. in clinical tests, attributing a
ated with a two-way within-subjects lack of results to the weakness of a treat-
analysis of variance. Also called two- ment when, in fact, the problem was
way repeated measures design. that the treatment was not administered
properly.
Type 0 error see TYPE III ERROR.
Type I error the error of rejecting the Type I sum of squares the reduction
in the ERROR SUM OF SQUARES obtained
NULL HYPOTHESIS when it is in fact true.
Investigators make this error when they by adding a particular factor (e.g., Factor
believe they have detected an effect or a A) to a flt that already includes the other
relationship that does not actually exist. factors in the model (e.g., Factors B, C,
The projected probability of committing and D): In other words, it is the amount
a Type I error is called the SIGNIFICANCE of the sum of squares attributable only
LEVEL or alpha (a) level. Also called to Factor A. A Type I sum of squares de-
alpha error; rejection error. pends on the order in which factors are
listed in the model statement. For some
Type II error the error of failing to re- data sets, there would be different results
ject the NULL HYPOTHESIS when it is in for a model that states B then A than for
fact not true. Investigators make this a model that states A then B; this is be-
error if they conclude that a particular cause each factor is adjusted only for the
effect or relationship does not exist factors that were entered before it. If one
when in fact it does. The probability of flts two models, one with A then B, the
committing a Type II error is called the other with B then A, not only can the
beta (P) level of a test. Conversely, the Type 1 sum of squares for Factor A be dif-
probability of not committing a Type II ferent under the two models, but there
error (i.e., of detecting a genuinely sig- is no way to predict whether the sum
nificant difference between samples) is of squares will go up or down when A
 Type III sum of squares

comes second instead offirst.Also called quential sum of squares. Compare

sequential sum of squares. Compare T Y P E I S U M O F S Q U A R E S ; T Y P E III S U M O F
T Y P E II S U M O F S Q U A R E S ; T Y P E III S U M O F SQUARES.
SQUARES. Type III sum o f squares the SUM OF
Type II sum of squares the reduction SQUARES for a particular factor (e.g.. Fac-
in the ERROR SUM OF SQUARES obtained tor A) after correcting, controlling for, or
by adding a particular factor (e.g.. Factor taking into account all of the other fac-
A) to a model consisting of all other fac- tors in a model (e.g., Factors B, C, and
tors (e.g.. Factors B, C, and D) that do D). The Type III sum of squares does not
not contain the factor in question. The depend upon the order in which effects
Type II sum of squares is appropriate for are specified in the model. It is also unaf-
model-building procedures in which fected by the frequency of observations,
one would adjust, for example, B for A, A making it appropriate for designs in
for B, and then AB (the INTERACTION EF- which there are unequal numbers of ob-
FECT) for A and B. Examination of the servations in each group. Also called
different reductions provides insight marginal sum of squares. Compare
into the best model. Also called hierar- T Y P E I S U M OF S Q U A R E S ; T Y P E II S U M OF
chical sum of squares; partially se- S Q U A R E S .

395
ultimate cause the underlying reason RANDOM ERROR and therefore self-
for some observed result or event, as canceling in the long run.
compared to the PROXIMATE CAUSE di-
unbiased error see RANDOM ERROR.
rectly preceding it. For example, in a
study of the link between a social factor, unbiased estimator a quantity calcu-
such as socioeconomic status, and an lated from sample data whose value is
outcome behavior, such as health be- representative of the true quantity in the
havior, the data may show that lower so- larger population. In other words, when
cioeconomic groups practice poorer data from samples are used to make in-
health behaviors. However, the ultimate ferences about unknown quantities (pa-
cause may be that lower socioeconomic rameters) in populations, an unbiased
groups are subject to greater hazard or estimator is one that over repeated sam-
environmental harm, which inramdis- pling has an average equal to the true
courages healthy behavior. Also called value of the parameter in the popula-
distal cause; remote cause. tion. For example, an unbiased estima-
tor of variance provides an accurate
U M P test abbreviation for UNIFORMLY index of the variability of measurements
MOST POWERFUL TEST. for a given phenomenon in a given pop-
unbalanced design an experimental ulation of interest based on calculations
design having multiple INDEPENDENT made from the SAMPLE VARIANCE. Also
VARIABLES in which the number of mea- called unbiased statistic. Compare BI-
surements or observations obtained ASED ESTIMATOR.
is different for each condition under
unbiased s a m p l i n g selecting individ-
study. Although BALANCED DESIGNS
uals for a study using a process that
generally are preferred for their greater
yields a group exemplitying the larger
POWER in statistical analyses, unbal-
population from which it derives. In
anced designs nonetheless may arise due
practice, no strategy produces a com-
to participant ATTRITION or other un-
pletely unbiased sample but RANDOM
avoidable factors. For example, if a re-
SAMPLING yields a good approximation,
searcher is investigating how sleep and
as it introduces the minimum possible
diet influence academic performance
amount of error in representing the pop-
and only 35 of the 40 college undergrad-
ulation. Compare BIASED SAMPLING.
uates recraited to participate were able
to do so through the full term of the pro- unbiased statistic see UNBIASED ESTI-
ject, certain data will be missing for the MATOR.
five people who left the study prema-
rarely and the design will become unbal- uncertainty analysis in an experi-
anced. ment or study, an analysis used to assess
the accuracy of measurements and
unbalanced l o n g i t u f l i n a l data see model predictions, taking into account
BALANCED LONGITUDINAL DATA. such possible sources of error as instm-
mentation, methodology, and the pres-
unbiased adj. impartial or without net ence of CONFOUNDS.
error. In unbiased procedures, studies,
and the like any errors that do occur are u n c o n d i t i o n a l model a type of HIER-

396
 underidentified model

ARCHICAL LINEAR MODEL that amalgam- degrees then education would be identi-
ates the effect of explanatory or PRE- cal for all individuals and thus serve as a
DICTOR VARIABLES across the different CONTROL VARIABLE.
levels of analysis. For example, imagine
uncorrected moment a MOMENT cal-
a study having three levels of analysis:
culated on the basis of the raw data in a
students (Level 1), nested within class-
distribution, rather than from the same
rooms (Level 2), further nested within
data after CORRECTION for error or some
schools (Level 3). An unconditional
unwanted effect. The term is often ap-
model would represent variation in the
plied to moments calculated from a
outcorne measure or DEPENDENT VARI-
GROUPED FREQUENCY DISTRIBUTION (as
ABLE across the three levels (sradent,
in a corrected moment).
classroom, and school), whereas a CON-
DITIONAL MODEL would explain the in- underdetermination n. a situation in
dividual influence of each predictor which the available evidence is insuffi-
variable on the dependent variable. cient to know what beliefs to hold in re-
sponse. For example, if one knows that a
u n c o n d i t i o n a l variance the long- person spent $10 on apples and oranges
term trend in a sequence of observations and that apples cost $1 while oranges
made over time. Unconditional variance cost $2, then one knows that the person
implies that knowledge of individual did not buy six oranges but not whether
fluctuations between previous observa- the person bought one orange and eight
tions is irrelevant to identifying and un- apples, two oranges and six apples, and
derstanding the overaU variability of the so forth. underdetermined adj.
sequence as a whole! Compare CONDI-
TIONAL VARIANCE. underestimation n. a situation in
which systematic discrepancies between
u n c o n f o u n d e d comparison a com-
measurements and model outputs lead
parison of two or more groups of re-
researchers to calculate a PARAMETER of
search participants that yields an
interest as being significantly lower than
unbiased estimate of the effect of the
its true value. For example, applying a
treatment or other condition under in-
REGRESSION ANALYSIS In whlch there are
vestigation. For comparisons to be un-
BIASED ESTIMATORS could lead a medical
confounded, studies must be designed
researcher to obtain a lower value for the
to ensure identical handUng of all partic-
presence of a particular illness in the
ipants, in addition to their RANDOM AS-
general population than is acraally the
SIGNMENT to different experimental
case.
groups, such that any differences between
the groups may be attributed solely to u n d e r i d e n t i f i c a t i o n n. a situation,
the experimental manipulation rather such as may occur during STRUCTURAL
than the influence of other related factors. EQUATION MODELING, in whlch it Is not
Compare CONFOUNDED COMPARISON. possible to estimate all of the model's
parameters based on the sample data
uncontrolled variable a characteris-
being analyzed. Compare OVERIDENTI-
tic factor that is not regulated or mea-
FICATION.
sured by the investigator during an
experiment or study, such that it is not underidentified model a theoreti-
the same for all participants in the re- cally identified model in which one or
search. For example, if the investigator more PARAMETER esfimates cannot be
collects data on participants having made because of an insufficient number
varying levels of education, then educa- of available data points. This may occur,
tion is an uncontrolled variable. If the for example, when there is high
investigator,' however, were to collect MULTICOLLINEARITY among the vari-
data only on participants with college ables studied. Empirically underidenfi-

397
underlying dimension

fied parameters are vety unstable. from a larger population in which differ-
Compare JUST-IDENTIFIED MODEL; OVER- ent units have different likelihoods of
IDENTIFIED MODEL. being chosen for inclusion. In SAMPLING
u n d e r l y i n g dimension an explana- WITHOUT REPLACEMENT, for example,
tory or descriptive concept, inferred each time a unit is chosen to be part of
from empirical evidence or theory, that the sample it is removed from further
is used to interpret observed outcomes. consideration, thus altering the likeli-
For example, researchers sradying infant hood of inclusion for the remaining
attachment may conclude that the dif- units. This contrasts with equal proba-
ferent attachment styles demonstrated bility sampling, in which all population
can be explained by two underlying di- units have the same likelihood of being
mensions of avoidance and anxiety. In selected for the sample.
FACTOR ANALYSIS and Other contexts, unexplained variance see ERROR
such underlying dimensions are called V A R I A N C E .
LATENT VARIABLES.
u n f o l d i n g n. a unidimensional SCALING
u n d e r l y i n g d i s t r i b u t i o n the THEO- procedure in which respondents evalu-
RETICAL DISTRIBUTION fot a given popu- ate a set of items and their choices are
lation of interest. used to construct a continuum along
underspecified model in CAUSAL which their relative preferences are
MODELING, a predictive formula from placed. For example, consider a scale
which relevant explanatory variables measuring attitudes toward marijuana
have been excluded, such that the effect sales. A person who wholeheartedly fa-
estimates it provides are both biased vors the item "The city should legalize
and inefficient. See OMITTED VARIABLE the sale of marijuana" would be located
BIAS. at a different point along the continuum
from a person who completely opposes
undirected graph a display in which the sale of marijuana, and a person who
various points (nodes) of information endorses the item to some extent would
are connected by lines (edges) having be located at yet another point along
no direction. The value between two the continuum. See also MULTIDIMEN-
nodes often is called a WEIGHT. Thus, SIONAL UNFOLDING.
the following example has several nodes
(e.g.. A, B, F, E) and several weights (e.g., unidimensionality n. the quality of
5, 9, 12, 18). measuring a single construct, trait, or
other attribute. For example, a uni-
dimensional personality scale, attitude
scale, or other scale would contain items
related only to the respective concept of
interest. Compare MULTIDIMENSIONAL-
ITY. unidimensional adj.
u n i f o r m association model a model
for a CONTINGENCY TABLE in which a
constant ODDS RATIO is assumed across
all of the component cells.
In contrast, a DIRECTED GRAPH uses ar- u n i f o r m d i s t r i b u t i o n a theoretical
rows or numbered nodes to show CONTINUOUS DISTRIBUTION in which
directionality, such as A leading to B, the probability of occurrence is the same
which in turn leads to E, and so forth. for all values of x, represented by f(x) =
unequal p r o b a b i l i t y s a m p l i n g a l/(b -fl),where a is the lower limit of the
strategy for selecting a sample of units distribution and b is its upper limit. For

398
 unit-specific model

example, if a fair die is thrown, the prob- the system. It is given as 1 - hj^, where hj^
ability of obtaining any one of the six is the COMMONALITY of the jth variable.
possible outcomes is 1/6. Since all out- Each of the observed variables in the
comes are equally probable, the distribu- data set being analyzed can be expressed
tion is uniform. If a uniform distribution as a combination of a common factor
is divided into equally spaced intervals, shared among all variables plus a unique
there will be an equal number of mem- factor associated with a measurement
bers of the population in each interval. enor or another specific, individual source
Also called rectangular distribu- of variation.
tion.
u n i t m a t r i x see IDENTITY MATRIX.
uniformly most powerful test
u n i t n o r m a l d i s t r i b u t i o n see STAN-
(UMP test) a statistical test of one hy-
DARDIZED DISTRIBUTION.
pothesis against another that has the
greatest POWER among all tests available u n i t n o r m a l variable see STANDARD
at a given ALPHA value. Suppose the two NORMAL VARIABLE.
hypotheses are the NULL HYPOTHESIS of
no difference between two groups (Sam- unit of analysis in research, the group
ple 1 mean = Sample 2 mean) and the of people, things, or entities that are
ALTERNATIVE HYPOTHESIS of some dif- being investigated or studied. For exam-
ference between the groups (Sample 1 ple, in organizational contexts, data can
mean * Sample 2 mean). Since the latter be collected from employees, who in
actually is a composite of several possi- turn are part of departments, which in
bilities, a UMP test is one defined by a turn are part of the larger organization,
CRITICAL REGION that is best suited to which may have multiple sites in several
test the null hypothesis against each of countries. The unit of analysis chosen
the simple hypotheses comprising the influences the methodological and ana-
larger alternative hypothesis. lytical procedures used (e.g., studying
groups within organizations may re-
u n i m o d a l d i s t r i b u t i o n a set of scores quire a HIERARCHICALLY NESTED DE-
with a single peak, or MODE, around SIGN). Also called analysis unit; level
which values tend tofluctuate,such that of analysis.
the frequencies at first increase and then
decrease. See also BIMODAL DISTRIBU- unit root test a statistical procedure
TION; MULTIMODAL DISTRIBUTION. that uses an AUTOREGRESSIVE MODEL to
determine whether TIME-SERIES DATA
u n i p o l a r r a t i n g scale a type of instru- exhibit systematic trends, which must
ment that prompts a respondent to eval- be mathematically removed before any
uate the degree to which a single quality analyses may be conducted. Two com-
or attribute is present. For example, con- monly used unit root tests are the
sider a five-point scale with the foUow- Dickey-Fuller test and the Phillips-
ing anchors or benchmarks: (1) not at all Perron test.
satisfied, (2) slightly satisfied, (3) moder-
ately satisfied, (4) very satisfied, and (5) unit-specific model a HIERARCHICAL
completely satisfied. Since there is no an- LINEAR MODEL that aims to identify those
chor that represents the opposing qual- outcomes specifically associated with a
ity of dissatisfaction, the scale has one parficular level, magnitude, or categoty
pole. Also called unipolar scale. Com- of the independent variable being ana-
pare BIPOLAR RATING SCALE. lyzed. For example, a researcher may be
investigating sradents, who are nested
uniqueness n. in FACTOR ANALYSIS, that within classrooms, who are further
part of the variance of a variable that it nested within schools. A unit-specific
does not share with any other variable in model would address questions about

399
unity

one of these three levels or units of analy- universe n. see POPULATION.

sis: sradents (Level 1), classrooms (Level
2), or schools (Level 3). Compare POPU- universe of admissible generaliza-
LATION-AVERAGED MODEL.
tion in GENERALIZABILITY THEORY, the
entire set of conditions to which a re-
u n i t y n. a synonym for the value of 1. searcher wishes to generalize from a sub-
Statisticians often state that the statistic set of actual measurements obtained
or value they are working with "ap- under specific conditions. It is impor-
proaches unity," which means it ap- tant when specifying the universe of ad-
proaches the value of 1. missible generalization to isolate and
univariate adj. characterized by a single estimate as many potential sources of
variable. For example, a researcher may MEASUREMENT ERROR as is reasonably
collect univariate data by recording how and economically feasible. Otherwise,
many hours a day students in a particular unreliable findings may be used to make
course spend outside of class on complet- improper generaUzations; Also called
ing their homework. Compare BI- universe of generalization.
VARIATE; MULTIVARIATE. universe of admissible observa-
univariate analysis a statistical exam- tions in GENERALIZABILITY THEORY, the
ination of data for only one variable of set of all possible measurements or
interest. For example, a univariate analy- scores that could reasonably be substi-
sis of srady habits for a sample of college rated for one that was acraally obtained
students would examine habits across all by a researcher. In other words, any ac-
individuals without taking into account tual behavioral measurement (e.g., a test
whether a particular student was a fresh- score) is considered to be a sample from
man, sophomore, junior, or senior. Also a larger universe of potentially inter-
called univariate statistics. Compare changeable measurements. Although re-
MULTIVARIATE ANALYSIS. search participants cannot be exposed to
all possible levels or conditions of a mea-
univariate d i s t r i b u t i o n a distribu- sure (e.g., an infinite pool of test items
tion of values on a single RANDOM VARI- about a topic), the goal is to characterize
ABLE according to their observed or the variability of a participant's score
expected frequency. If this is a NORMAL over all possible combinations of such
DISTRIBUTION it is known as a univariate levels or conditions.
normal distribution. Compare MULTI-
VARIATE DISTRIBUTION. universe score in GENERALIZABILITY
THEORY, an average score for a given per-
univariate outlier an extreme or un- son derived from an actual sample of ob-
usual value occurring for a single vari- servations and a given UNIVERSE OF
able. For example, if a student in a ADMISSIBLE OBSERVATIONS.
college course obtains a score of 100
points on an exam when all other stu- unmeasured variable 1. in PATH
dents in the course score below 80 ANALYSIS, any hypothetical variable, LA-
points his or her perfect score would be a TENT VARIABLE, or unobservable vari-
univariate outlier. Compare MULTI- able. 2. in REGRESSION ANALYSIS, any
VARIATE OUTLIER. variable that is not specifically assessed
or evaluated but that is presumed to in-
univariate research research that em- fluence the relationship observed be-
ploys only one DEPENDENT VARIABLE. tween two other variables, x and y. The
Compare MULTIVARIATE RESEARCH. influence of such an unmeasured vari-
universal set in SET THEORY, the entire able is subsumed in the ERROR TERM for
collection of elements currently under y. In many statistical models of relation-
consideration. ships, causal estimates may be biased be-
 unweighted means analysis

cause relevant variables have not been color. Unselective observation is a more
included in the causal systems investi- objective strategy than SELECTIVE OB-
gated, a problem generally referred to SERVATION.
as OMITTED VARIABLE BIAS. See THIRD-
unstandardized score see RAW
VARIABLE PROBLEM.
SCORE.
unnumbered graphic rating scale
see VISUAL ANALOGUE SCALE. unstructured i n t e r v i e w an inter-
view that is highly flexible in terms of
unobtrusive measure a measure ob- the questions asked, the kinds of re-
tained without disturbing the partici- sponses sought, and the ways in which
pant or alerting him or her that a the answers are evaluated across inter-
measurement is being made. For exam- viewers or across interviewees. For ex-
ple, a researcher may observe passersby ample, a human resource staff member
in a public park from a nearby cafe and conducting an unstructured interview
document their activities. The behavior with a candidate for employment may
or responses of such participants are ask open-ended questions so as to allow
thus assumed to be unaffected by the in- the spontaneity of the discussion to re-
vestigative process or the surrounding veal more of the applicant's traits, inter-
environment. Also called concealed ests, priorities, and interpersonal and
measure; nonreactive measure. verbal skills than a standard predeter-
Compare REACTIVE MEASURE. mined question set would. Also called
unobtrusive observation the process nondirective interview. Compare
STRUCTURED INTERVIEW.
of coUecting UNOBTRUSIVE MEASURES,
such as by the use of hidden cameras.
unstructured stimulus a vague,
u n p l a n n e d comparison see POSTpoorly organized, and not clearly identi-
HOC COMPARISON. fiable stimulus, such as an inkblot in the
Rorschach Inkblot Test. The perception
unpredicted variance see ERROR VAR-
of unstmctured stimuli is often thought
IANCE.
to be influenced more by the character-
unrelated t test see INDEPENDENT- istics of the perceiver than by those
SAMPLES T TEST. of the stimulus. Compare STRUCTURED
unrestricted model see FULL MODEL. STIMULUS.
unsaturated model see SATURATED unsystematic error see RANDOM
MODEL. ERROR.

unselected sample an informal name unsystematic variance the haphaz-

for a random sample. See RANDOM SAM- ard or random fluctuation of data for in-
PLING. dividuals over time. It is one of two types
of variance identified in research, the
unselective observation a process in other being systematic variance arising
which an individual watches and re- from the effects of the INDEPENDENT
cords all occurrences of a prespecified VARIABLES Studied.
event within a given period of time. For
example, if a researcher wishes to deter- unweighted means analysis in
mine whether drivers who have cars of a ANALYSIS OF VARIANCE, a Strategy for
certain color tend to exceed the speed handling missing data in which the ac-
limit more often, he or she could note tual sample size in each of the different
the speed and color of every car that conditions or levels of the INDEPENDENT
passes a particular point within a 5- VARIABLE is replaced by the average sam-
minute period and then compare the ple size. When participants become ill,
different percentages of speeding cars by withdraw, or otherwise fail to complete

401
up-and-down method

a research study, there are no longer the extreme right on the x-axis in the
equal numbers of cases in each experi- plot of a FREQUENCY DISTRIBUTION, de-
mental condition. An unweighted means noting the probabUity of obtaining the
analysis takes this inequality into ac- highest value in the distribution. Upper-
count by applying an adjusted sample tail probabilities are useful when con-
size value across all conditions, thus en- ducting statistical tests of experimental
suring that data from all experimental hypotheses. Compare LOWER-TAIL PROB-
groups contribute equally to the deter- ABILITY.
mination of treatment effects. Also
called unweighted means proce- upper w h i s k e r in a BOX-AND-WHISKER
dure. PLOT, the line extending from the value
at the 75th PERCENTILE to the largest
up-and-down method a strategy in value within one INTERQUARTILE RANGE
which the administration of a stimulus of that percentile score. Compare
event or item depends on a participant's LOWER WHISKER.
response to the previous item. The up-
and-down method employs fixed, dis- u p w a r d P y g m a l i o n effect an effect
crete levels and sequential transition in which the expectations of followers
rales of moving one level up, moving or subordinates lead to behavior on the
one level down, or remaining at the cur- part of the leader or superior that is con-
rent level. For example, in PSYCHO- sistent with these expectations. The be-
PHYSICAL RESEARCH, a participant who havior of the leader does not reflect his
correctly detects a stimulus during one or her true abilities or personality traits,
trial would be presented with a reduced but rather the perception of the leader
value of the same stimulus in the next by subordinates. Compare PYGMALION
trial; conversely, if he or she did not de- EFFECT. See SELF-FULFILLING PROPHECY.
tect the original stimulus its value would
u r n model a THOUGHT EXPERIMENT in
be increased in the next trial. Thus, the
which objects of interest (e.g., people,
stimulus value that is presented varies
events) are represented as colored balls
sequentially until a stable estimate is ob-
placed in an urn or other container. In
tained of the individual's sensitivity to
imagination, the experimenter ran-
the stimulus. Similarly, in computer
domly removes one "baU" from the urn,
adaptive testing there is a finite set of
notes its color, and places it back before
possible items with known difficulty lev-
repeating the process; the goal is to de-
els, and depending on the individual's
termine the probability of drawing one
response to the initial item the next item
color or another. The urn model is a con-
administered wiU have the same diffi-
venient way to calculate certain basic
culty or be slightly more or less difficult.
probabilities using CONDITIONAL PROB-
The process continues until an estimate
ABILITIES.
of the individual's ability on the trait
being measured can be obtained. V-shaped d i s t r i b u t i o n a graphical
representation of a FREQUENCY DISTRI-
upper hinge the point in a distribution BUTION that is shaped more or less like
of values above which lie one fourth of the letter U, with the maximum fre-
the data and below which lie the other quencies at both ends of the range of the
three fourths of the data. It is equivalent variable. For example, the number of
to the third QUARTILE. Compare LOWER people infected by the flu each year may
HINGE.
have a U-shaped distribution by age,
upper quartile see QUARTILE. with those who are very young or very
old having the highest frequency of oc-
upper real l i m i t see REAL LIMIT.
currence, as shown in the hypothetical
upper-tail p r o b a b i l i t y the value at illustration opposite.

402
 U-shaped distribution

10,000

7,500

w
ra
O
5,000
o

2,500

"T" I
20 40 60 80 100

Age (years)
U-shaped distribution

In an inverted U-shaped distribution, die of the distribution and the least fre-
the most frequent values are in the mid- . quent values are at the extremes.

403
 Vv
vague p r i o r in BAYESIAN statistics, an and the legitimacy of included values
imprecise probability specification ap- (e.g., impossible values are not present),
plied in the estimation of a population confirming that information is consis-
parameter when more conclusive infor- tent across records, and confirming that
mation about its true value is lacking. no records are missing.
For example, a researcher may use a
mean obtained previously for another v a l i d i t y coefficient an index, typi-
population to constract a CONFIDENCE cally a CORRELATION COEFFICIENT, that
INTERVAL within which the mean for reflects how well an assessment instru-
the current population of interest may ment predicts a well-accepted indicator
be considered to lie. of a given concept or criterion. For ex-
ample, if a measure of criminal behavior
validation n. the process of establishing is valid, then it should be possible to use
the trath or logical cogency of some- it to predict whether a person (a) will be
thing. An example is determining the arrested in the future for a criminal vio-
accuracy of a research instrument in lation, (b) is currently breaking the law,
measuring what it is designed to mea- and (c) has a previous criminal record. A
sure. validate vb. validity coefficient could be used to re-
late scores on the measure to each of
validity n. the degree to which empiri-
these criteria and thus determine how
cal evidence and theoretical rationales
useful the measure actually is for behav-
support the adequacy and appropriate-
ioral forecasting.
ness of conclusions drawn from some
form of assessment. Validity has multi- valiflity criterion an external concept
ple forms, depending on the research or standard of comparison that is used to
question and on the particular type of define the attribute an instrament is pur-
inference being made. For example, the ported to measure and that is applied in
three major types of test validity are CRI- estimating how well the measurement
TERION VALIDITY, based on correlation instrament acraally fulfills its intended
with an accepted standard; CONSTRUCT purpose. See CRITERION VALIDITY.
VALIDITY, based on the concepraal vari-
able underlying a test; and CONTENT VA- v a l i d i t y d i a g o n a l the MAIN DIAGONAL
LIDITY, based on the subject matter of a in a MULTITRAIT-MULTIMETHOD MA-
test. Other forms of validity prominent TRIX, which represents correlations be-
in the social sciences include ECOLOGI- tween the same construct or trait
CAL VALIDITY, EXTERNAL VALIDITY, IN- measured using different methods. See
TERNAL VALIDITY, and STATISTICAL CON- CONVERGENT VALIDITY.
CLUSION VALIDITY. See also THREATS TO
VALIDITY, valid adj. v a l i d i t y generalization the use of
META-ANALYSIS and other statistical pro-
v a l i d i t y check the process of verifying cedures to assess the evidence of a test's
that a data set is free of errors and ad- adequacy and appropriateness in multi-
heres to standard or intended rales. Per- ple siraations and settings. Validity gen-
formed manually or using software, a eralization typically involves correcting
validity check may involve such things all of the correlations being examined
as verifying the accuracy of calculations for methodological and statistical limi-
 variance explained

tations and flaws and providing esti- PARISONS can be made with equal preci-
mates of correlations or results that sion. In other words, the amount of vari-
would have been obtained in the ab- ation or inaccuracy involved when
sence of such limitations. estimating the average effect for one
treatment will be the same as that in-
value analysis a type of CONTENT
volved when estimating the average ef-
ANALYSIS of written material consisting
fect for all other treatments.
of a table, or other systematic notation,
documenting the frequency of appear- variance components analysis any
ance in the material of all expressions re- statistical procedure for examining
ferring to specified values. MIXED-EFFECTS MODELS and RANDOM-
EFFECTS MODELS that decomposes the
v a r i a b i l i t y n. the degree to which
members of a group or population differ total variance on a DEPENDENT VARIABLE
from each other, as measured by such into that stemming from the effects of
statistics as the RANGE, STANDARD DEVI- the INDEPENDENT VARIABLE (the TRUE
ATION, and VARIANCE. VARIANCE) and that produced by extra-
neous factors (the ERROR VARIANCE). An
variable n. a condition in an experi- ANALYSIS OF VARIANCE provides an ex-
ment or characteristic of an entity, per- ample. Assume a researcher is studying
son, or object that can take on different whether caffeine and gender influence
categories, levels, or values and that can students' test performance. The former
be quantified (measured). For example, is a random variable, as it has numerous
test scores and ratings assigned by judges possible levels from which to choose a
are variables. Numerous types of vari- subset to study (e.g., 0 mg, 50 mg, 100
ables exist, including CATEGORICAL mg, 150 mg), and the latter is a flxed
VARIABLES, DEPENDENT VARIABLES, IN- variable, as it may assume only one of
DEPENDENT VARIABLES, MEDIATORS, two values (male or female). The re-
MODERATORS, and RANDOM VARIABLES. searcher might use an analysis of vari-
Compare CONSTANT. ance to apportion the variance in the
variable error see RANDOM ERROR. response measuretest performance
among the two different factorslevel
variable stimulus in PSYCHO- of caffeine and being male or female.
PHYSICAL RESEARCH, any one of a set of The central output is a table that shows
experimental stimuli that are to be sys- the proportion of variance attributable
tematically compared to a constant to the main effects of the factors, the
stimulus. proportion attributable to interactions
variance (symbol: a^) n. a measure of between the factors, and the proportion
the spread, or DISPERSION, of scores attributable to error and other external
within a sample or population, whereby factors.
a small variance indicates highly similar variance components model see
scores, aU close to the sample MEAN, and RANDOM-EFFECTS MODEL.
a large variance indicates more scores at
a greater distance from the mean and varianee-covariance m a t r i x see
possibly spread over a larger range. See COVARIANCE MATRIX.
also STANDARD DEVIATION. variance estimate an index of varia-
variance analysis see ANALYSIS OF VAR- tion in a population that has been calcu-
IANCE. lated using a sample of that population.
For example, a sample STANDARD DEVIA-
variance-balanced design a type of
TION is an estimate of the deviation in
CROSSOVER DESiGN-in which the possi-
the larger population.
ble sequences of treatments are arranged
in such a manner that all PAIRWISE C O M - variance explained an indication of

405
variance inflation factor

how well variation in one variable (or set ability within each of the smaller, more
of variables) can be accounted for by the homogeneous groups is less than that
variation in another variable. For exam- for the undivided larger group.
ple, if the CORRELATION COEFFICIENT
variance ratio a comparison of two in-
between a PREDICTOR VARIABLE (X) and
dices of variance, expressed as a quo-
a DEPENDENT VARIABLE (y) is 0.25, then
tient. The F RATIO, which compares
25% of the variability in y is explained
effect variances to error variances, is an
by the variability in x.
example.
variance i n f l a t i o n f a c t o r (VIF) an variance-ratio test see F TEST.
index of the degree to which the vari-
abiUty of an estimated REGRESSION CO- variance-reduction model see VARI-
EFFICIENT is increased because of in- ANCE-PRESERVATION MODEL.
terrelationships among the variables in variance-stabilizing transforma-
an ordinaty LEAST SQUARES REGRESSION t i o n in REGRESSION ANALYSIS Or ANALY-
model. A variance inflation factor exists SIS OF VARIANCE, any of a class of
for each of the k PREDICTOR VARIABLES mathematical processes that convert data
in the model. A VIF of 1 means that s h o w i n g H E T E R O G E N E I T Y OF V A R I A N C E
there is no correlation among the k^^ to a different form, so as to enable the
predictor and the remaining predictor application of specific analytical tech-
variables (and hence the variance is not niques or to simplify considerations.
inflated at all), a VIF exceeding 4 war-
rants further investigation, and a VIF ex- variate n. 1. a specific value of a particu-
ceeding 10 indicates serious MULTI- lar VARIABLE. 2. a RANDOM VARIABLE it-
COLLINEARITY requiring correction. seff.
v a r i a t i o n n. fluctuation: the degree of
variance-preservation model a
VARIANCE or DISPERSION of values that
method of conducting a META-ANALYSIS is obtained for a specific variable.
to determine EFFECT SIZES that accounts
for variation among study designs. For v a r i a t i o n coefficient see COEFFI-
example, some studies included in the CIENT OF VARIATION.
analysis may involve a single INDEPEN- v a r i a t i o n r a t i o a measure of VARIABIL-
DENT VARIABLE (fl) with a single level
ITY for numerical values that represent
while others may involve two indepen- membership in specific categories (i.e.,
dent variables (b, c) with two levels each. for CATEGORICAL DATA). It indexes the
To ensure that single-factor effect-size proportion of cases that deviate from the
measures are comparable to multiple- MODE.
factor effect-size measures, a researcher
may use a variance-preservation model v a r i m a x r o t a t i o n a statistical proce-
to treat the multiple-variable design as dure applied within FACTOR ANALYSIS
two single-variable designs (i.e., the and PRINCIPAL COMPONENTS ANALYSIS
treatment effect for a is measured at to simplify solutions and enhance inter-
Level 1 of b and again at Level 2 of b). pretation of the results. It is a type of OR-
Variances within each level are then THOGONAL ROTATION intended to make
compared to ensure that they do not dif- each factor have a small number of large
fer significantly between the two types FACTOR LOADINGS and a large number of
of,designs. Alternatively, a researcher zero (or small) factor loadings. Thus, fol-
may use a variance-reduction model, in lowing a varimax rotation, each original
which the participants in the single- variable tends to be associated with a
variable design are stratified into two small number of factors, and each factor
homogeneous subpopulations (e.g., represents only a small number of vari-
males and females), such that the vari- ables. See FACTOR ROTATION.

406
 venue sampling

h (distance in km)
variogram

v a r i o g r a m n. a graphical representation scores of n individuals on a particular

of the degree of correlation between val- measure are arrayed. 3. in MATRIX ALGE-
ues separated by particular distances, BRA, a column or row of a matrix.
providing information about whether vector analysis a branch of mathemat-
values that are closer together in geo- ics concerned with the srady of entities
graphic space are more similar than ob- having both magnitude and direction in
servations that are farther apart. three-dimensional space.
Consider the generic example above, in
which the diamonds represent observed vector product see PRODUCT VECTOR.
data points, the solid curve represents V e n n d i a g r a m a visual depiction of el-
estimated data values, and the dashed ements and relations. Consider the fol-
line represents VARIANCE. lowing generic example.
One may use a variogram, for instance,
to examine whether adolescents living
in close proximity have more similar
perceptions of their neighborhoods than
do more widely dispersed adolescents.
varying-coefficient model a model
that provides information on the associ-
ation between a DEPENDENT VARIABLE
and a PREDICTOR VARIABLE over time.
An extension of MULTIPLE REGRESSION
techniques to data that have been col-
lected over multiple occasions (i.e., LON-
GITUDINAL DATA), vatying-coefficient
models substimte functional coefficients Circles represent the elements of a set,
that reflect the time-dependent relation- and the union and intersection between
ships between the variables of interest or among the circles represent relation-
for the constant coefficients used in typ- ships between the sets (i.e., the degree to
ical multiple regression. which they are mutually inclusive or ex-
clusive). [John Venn (1834-1923), Brit-
vector n. 1. a mathematical entity with ish logician]
magnitude and direction. Compare SCA-
LAR. 2. in MULTIVARIATE ANALYSIS, a one- venue s a m p l i n g see TIME-LOCATION
dimensional arrangement in which the SAMPLING.

407
verbal protocol

45 1
40
35 -
3
30
a. 25
o 0
a. 2
15
10
5
0
Democrat Republican Independent No Affiliation

Political Party

vertical line graph

verbal protocol a method of eliciting tions with less defensive justification

verbal reports of individuals' thoughts as than memory alone.
they perform a set of specified research
VIF abbreviation for VARIANCE INFLA-
tasks. Participants are asked to describe
TION FACTOR.
whatever they are seeing, thinking,
doing, and feeling as it occurs during task v i o l a t i o n of assumptions a situation
performance. Rather than capraring in which the theoretical ASSUMPTIONS
mere summaries of a task's results, the associated with a particular statistical or
method provides the researcher with in- experimental procedure are not fulfilled.
sight about the cognitive processes in- Common assumpfions for statistical
volved in task completion and the tests include NORMALITY of the distribu-
representations on which they operate. tion, equal VARIANCES within the cells
of treatment levels, HOMOGENEITY OF
verisimilitude n. the appearance of VARIANCE, and LINEARITY. Research de-
being true. In scientific investigation, a signs also need to meet certain assump-
theory or model is said to have verisimil- tions, such as RANDOM SAMPLING and
itude if it can be shown to be more con- RANDOM ASSIGNMENT, Sample REPRE-
sistent with empirically verified fact SENTATIVENESS, and the like. Because vi-
than its predecessors or competitors. olation of assumptions introduces bias,
vertical axis see Y-AXIS. the validity of assumptions must be con-
firmed prior to data analysis to ensure
vertical line graph a way of display- that the methods and strategies chosen
ing data that is similar to a BAR GRAPH are appropriate and will yield valid re-
but uses discrete lines of varying height sults.
to represent the different values of a
violation-of-expectation method
variable. For example, the above hypo-
a technique for studying infant cogni-
thetical illustration shows the political
tion, based on habituation and dis-
affiliations of Americans.
habituation procedures, in which
video-recall technique a research increases in an infant's looking time at
procedure in which participants review an event or other stimulus are inter-
a video recording of their behavior in a preted as evidence that the outcome he
previous situation and report what they or she expected has not occurred. For ex-
were thinking, feeling, or otherwise ex- ample, while a baby watches, a re-
periencing at that time. The video-recall searcher may repeatedly return a toy to a
technique is intended to reduce distor- blue box. If the researcher sometime
tion in SELF-REPORTS, the premise being later retrieves the same toy from a
that reimmersion in the situation gener- nearby red box (after a CONFEDERATE
ates a more representative set of recollec- surreptitiously moved it) and the baby

408
 volunteer bias

Employee overall performance

during past year
excellent

visual analogue scale

looks longer at that red box, it is as- function in newborn infants might have
sumed that he or she has some under- items requiring the infants tofixateon a
standing of object permanence and was target of black and white concentric cir-
not expecting the toy to be there. cles or track a colorful object as it moves
visual analogue scale a psychometric past their eyes.
instrument used to evaluate subjective voice key an electronic device that in-
characteristics that extend over a range terfaces between a microphone and a
of continuous values: Respondents spec- computer, used for recording vocal re-
ity their level of the characteristic of in- sponse times in language-production
terest by indicating a position along a tasks, such as word or picture naming.
continuous line anchored at its end When a stimulus is presented the voice
points by word descriptors. For example, key is activated and begins to monitor
a visual analogue scale for performance the sound level from the microphone; if
might have poor on the left end with a the sound level then exceeds a specific
blank line across to excellent on the right threshold level, indicating speech, the
end, as shown above. voice key records the amount of time
The respondent would mark a place that has elapsed since it was activated.
along that line to indicate his or her per- volunteer bias any systematic differ-
ceived performance level. Also called un- ence between individuals who volunteer
numbered graphic rating scale. to be in a study versus those who do not,
visual test any test intended to measure which may potentially render the result-
or srady vision, involving stimuli that ing group or sample of participants un-
are viewed. For example, a test of visual representative of the larger population.

409
 Ww
W 1 . symbol for the COEFFICIENT OF CON- (M) and females (F) from 1 to 27 accord-
CORDANCE. 2. symbol for the statistic ing to their performance on a task:
obtained from the WILCOXON RANK-
MMMFFFMMMMFFMMMFFFFFFFMMFMM
SUM TEST.
In this data set there are nine runs. If
symbol for the statistic obtained from the two samples are from the same pop-
the CRAMER-VON MISES GOODNESS-OF- ulation, then the males and females will
FIT-TEST. be well mixed and the number of runs
thus will be large (e.g., close to 25); if the
wait-list c o n t r o l group a group of re- number of runs is small, as it is in this ex-
search participants who receive the ample, the ordering cannot be caused by
same intervention or treatment as those chance fluctuation and the NULL HY-
in the EXPERIMENTAL GROUP but at a
POTHESIS thus is rejected. Indeed, visual
later time. Wait-list CONTROL GROUPS inspection of the data shows that the
commonly are used in therapy outcome males tend to cluster toward the left-
and similar studies to account for the hand side of the scale and the females
potential influence of elapsed time upon toward the right-hand side. Also called
treatment effectiveness; they may also runs test; two-sample runs test.
be used to address the ethical ramifica- [Abraham Wald; Jacob Wolfowitz (1910-
tions of withholding treatment from in- 1981), U.S. psychologist]
dividuals.
Ward's method an approach to HIER-
W a l d d i s t r i b u t i o n see INVERSE GAUS- ARCHICAL CLUSTERING that attempts to
SIAN DISTRIBUTION. [Abraham Wald minimize the SUM OF SQUARES of any
(1902-1950), Hungarian-born mathe- two (hypothetical) clusters that may be
matician] formed at each step of the data analysis.
In other words, Ward's method reveals
Wald's test a PARAMETRIC statistical the cost in increased sum of squares that
procedure used to evaluate the signifi- arises from merging clusters, such that
cance of individual coefficients ((i) in a only those clusters that yield the small-
LOGISTIC REGRESSION model. It yields est increases should be merged to
estimated values of population PARAME- achieve the optimum grouping of cases,
TERS that are compared to a CHI-SQUARE [foe H. Ward (1927-2011), U.S. statisti-
DISTRIBUTION having one DEGREE OF cian]
FREEDOM in order to determine their
theoretical occurrence under the NULL wash-out p e r i o d the time frame allot-
HYPOTHESIS. [Abraham Wald] ted for an administered drug to be elimi-
nated from the body or for a previously
W a l d - W o l f o w i t z test a nonpara- administered intervention to become
metric test of the NULL HYPOTHESIS that ineffective. Wash-out periods are partic-
two samples have been taken from iden- ularly important in medical and other
tical populations, based on whether or clinical research since the CARRYOVER
not the number of rans or sequences in EFFECT between treatments might other-
an ordering is random. For instance, wise confound the estimates of treat-
consider the following ordering of males ment effects.
 Wherry's formula

wave n. an individual measurement ses- weights of 4, 3, 2, 1, and 0, respectively.

sion in panel sradies and other LONGI- One would multiply the number of A
TUDINAL DESIGNS in which the same grades a student obtained by 4, the num-
participants are measured repeatedly ber of B grades by 3, and so forth, and
over time. then divide the resulting sum by the
total number of grades to obtain the stu-
wavelet analysis a mathematical dent's overall weighted average. Also
method used in TIME-SERIES ANALYSIS called weighted mean.
to decompose variation into simpler, in-
dividual wavelets, or functions that weighted k a p p a an index of interrater
have specific frequencies, amplirades, agreement that takes into account the
and temporal characteristics. The pro- degree of disparity between the categori-
cess generates information about peri- zations assigned by different observers.
odicity within the series. Thus, different levels of agreement con-
tribute more or less heavily to the overall
weak inference a conclusion based value of kappa than others. For example,
upon correlations drawn from observa- if two raters differ by two categories, that
tional studies or QUALITATIVE RESEARCH, difference is assigned more importance
as opposed to experimental hypothesis (i.e., given a greater WEIGHT) in the anal-
testing. Compare STRONG INFERENCE. ysis than if they differ only by one cate-
weak l a w o f large numbers see LAW goty. See also COHEN'S KAPPA.
OF LARGE NUMBERS.
weighted mean see WEIGHTED AVER-
weight n. a coefficient or multiplier used AGE.
in an equation or statistical investiga- weighted m u l t i d i m e n s i o n a l scal-
tion and applied to a particular variable i n g (WMDS) see INDIVIDUAL-DIFFER-
to reflect the contribution to the data. ENCES SCALING.
For example, a weighted sample is one
in which different values are applied to w e i g h t i n g n. the process of multiplying
its different constiraent subgroups to re- test items, subtests, tests that are part of
flect their representation within the a test battety, or other measures that are
larger population from which it was components of a total score by a value or
taken. Thus, if a population is 50% male WEIGHT other than 1. If all components
and 50% female but the sample sradied were to be weighted by 1, the result
is 40% and 60%, respectively, different would be equal weighting, which is es-
multipliers could be used to adjust the sentially no weighting.
individual subsample results to match W e l c h test see ASPIN-WELCH-SATTER-
the makeup ofthe population. Similarly, THWAITE TEST.
a weighted least squares regression is a
version of ordinary LEAST SQUARES RE- Wherry's f o r m u l a an equation used
GRESSION in which different variables to estimate SHRINKAGE, or the degree to
contribute differentially to the analysis which REGRESSION COEFFICIENTS de-
process according to their relative im- rived from one sample are applicable to
portance. See WEIGHTING. another sample. It is given as

weighted average an average calcu- R' 1-^(1 -R\

lated to take into account the relative
importance of the items making up the
average: Different values or VVEIGHTS are
assigned to different data points to reflect
their relative confribution. For example,
in examining grade point average, one
might give grades A through F the
where N is the sample size, p is the num-
ber of predictor variables, and R^ is the
COEFFICIENT OF MULTIPLE DETERMINA-
TION. The formula yields an index called
the ADJUSTED R^ that itself has been ad-
justed according to the number of par-

411
whisker

ticipants and the number of predictors C O X O N SIGNED-RANKS TEST used to de-

that were used to generate the regression termine whether two DEPENDENT SAM-
equation. [R. J. Wherry Sr. (1904-1981), PLES have been drawn from the same
U.S. statistician] population. Each member of a sample
has two scores from some RATIO SCALE
w h i s k e r n. see BOX-AND-WHISKER PLOT. (e.g., a pre- and postscore), which allow
whole-interval recording a strategy for the computation of a difference
for observing behavior that provides in- score. The difference scores are then
formation about the specific timing and ranked and computations performed to
durafion of the behavior. In whole- determine whether the median of the
interval recording, the length of an obser- difference scores equal 0. The test takes
vafion session is identified (e.g., 1 hour) into account the direction of the differ-
and then broken down into smaller, equal- ences and gives more weight to large dif-
length time periods (e.g., 10-minute inter- ferences than to small differences. The
vals). An observer then records whether Wilcoxon matched-pairs signed-ranks
the behavior of interest occurs through- test is a NONPARAMETRIC equivalent of
out an entire interval, counts the total the DEPENDENT-SAMPLES T TEST. [Frank
number of intervals in which the behavior Wilcoxon]
was present, and calculates what percent-
W i l c o x o n rank-sum test a statistical
age of intervals that number represents.
test of centrality for ranked data that
See also PARTIAL-INTERVAL RECORDING.
compares the median values of two IN-
wide-range test a brief screening in- DEPENDENT SAMPLES to determine
strument administered to gauge ability whether they have been drawn from the
quickly and determine if a more compre- same population. In this NONPARA-
METRIC equivalent of the T TEST, one
hensive test is needed. Its items vaty
combines the data points from the dif-
widely in difficulty so as to measure per-
ferent groups into a single pool and
formance at very broad levels. For exam-
ranks their values in ascending order.
ple, a wide-range test of vocabulary The ranks that have been assigned are in
given to ninth-grade students might in- rarn used to determine the test statistic,
clude simple words typically known by W, which is evaluated for statistical sig-
elementary-school students as well as nificance. The calculations involved in
complex words generally familiar only the Wilcoxon rank-sum test are nearly
to college students. identical to those for the MANN-WHIT-
W i l c o x o n - M a n n - W h i t n e y test a NEY u TEST. [Frank Wilcoxon]
NONPARAMETRIC TEST, used when data
are rank-ordered, to determine whether W i l c o x o n signed-ranks test a NON-
two INDEPENDENT SAMPLES have been PARAMETRIC statistical procedure used
drawn from the same population, based to determine whether a single sample is
on comparison of their median values. derived from a population in which the
It combines the MANN-WHITNEY U TEST
median equals a specified value. The
and WILCOXON RANK-SUM TEST into data are values obtained using a RATIO
SCALE, and each is subtracted from the
a single statistical procedure. [Frank
hypothesized value of the population
Wilcoxon (1892-1965), Irish-born U.S.
median and the difference scores are
statistician; Henty Berthold Mann
then ranked. The test takes into account
(1905-2000), Ausfrian-born U.S. mathe-
the direction of the differences and gives
matician; Donald Ransom Whitney more weight to large differences than to
(1915-2001), U.S. statistician] small differences. The symbol for the
Wilcoxon matched-pairs signed- test statistic is T. Also called Wilcoxon
ranks test an extension of the WIL- T test. [Frank Wilcoxon]

412
 within-groups degrees of freedom

W i l k s ' s l a m b d a (symbol: A) a statistic and lowest remaining values before any

used in MULTIVARIATE ANALYSIS OF calculations are performed. Typically, an
VARIANCE to determine whether there equal number of high and low values are
are significant differences between the replaced, comprising from 10% to 25%
means of groups on a combination of of the total distribution. [Charles P.
DEPENDENT VARIABLES. It may be con- Winsor]
verted to an F RATIO to obtain a SIGNIFI-
W i s h a r t d i s t r i b u t i o n a generaliza-
CANCE LEVEL or used to calculate ETA
tion of the univariate CHI-SQUARE DIS-
SQUARED, which is a measure of the pro-
TRIBUTION to two or more variables. It is
portion of variance in the combination
often used as a model for the distribu-
of dependent variables that is explained
tion of the sample COVARIANCE MATRIX
by the INDEPENDENT VARIABLE. [Samuel
for normal random data pertaining to
Stanley Wilks (1906-1964), U.S. mathe-
multiple variables. [John Wishart (1898-
matician]
1958), British statistician]
Wilks's multivariate outlier test
w i t h d r a w a l design an experimental
a statistical procedure for identitying ex-
design in which the treatment or other
treme measurements or observations
intervention is removed during one or
within MULTIVARIATE data. Essentially,
more periods. A typical withdrawal de-
it is a sequential application of WILKS'S
sign consists of three phases: an initial
LAMBDA: Potential OUTLIERS are studied
condition for obtaining a baseline, a
in a forward fashion, with the most ex-
condition in which the treatment is ap-
treme point being removed at each step,
plied, and another baseline condition in
until the test fails to indicate the pres-
which the treatment has been with-
ence of outliers. [Samuel Stanley Wilks]
drawn. Often, the baseline condition is
Williams's agreement measure represented by the letter A and the treat-
(symbol: /) a numerical index that mea- ment condition by the letter B, such that
sures the degree of similarity between this type of withdrawal design is known
the judgments of a specific rater and the as an A-B-A DESIGN. A fourth phase of re-
judgments of the whole set of raters. applying the intervention may be
Specifically, it measures whether the added, as well as a fifth phase of remov-
rater of interest agrees with the set as ing the intervention, to determine
often as others in that set do. Williams's whether the effect of the intervention
agreement measure is similar to CO- can be reproduced (see A-B-A-B DESIGN;
HEN'S KAPPA but distinguished by its A-B-A-B-A DESIGN).
focus on a particular rater. [George W.
Williams, U.S. biostatistician] within-groups analysis of vari-
ance see WITHIN-SUBJECTS ANALYSIS OF
Winsorized mean a measure of CEN- VARIANCE.
TRAL TENDENCY that is less sensitive to
OUTLIERS than is a standard mean. To within-groups degrees of freedom
obtain a Winsorized mean, one replaces in an ANALYSIS OF VARIANCE for a
the highest and lowest values from a set WITHIN-SUBJECTS DESIGN, the number
of data with less extreme values, sums of scores or observations obtained for a
the values in the modified set, and calcu- treatment level minus one, added across
lates the average. [Charles P. Winsor aU treatment levels. For example, in a
(1895-1951), U.S. statistician] 60-participant study examining three
dosages or levels of a drug treatment, the
W i n s o r i z i n g n. an IMPUTATION strat- DEGREES OF FREEDOM would be calcu-
egy, intended to reduce the influence of lated by determining the number of par-
OUTLIERS, in which extreme values in a ticipants observed within each level
data set are replaced with the highest minus one and them summing those

413
within-groups design

values: (20 - 1) + (20 - 1) + (20 - 1) = 19 +
19 + 19 = 57. The within-groups degrees
of freedom is used to calculate the
W I T H I N - G R O U P S M E A N S Q U A R E . Also
called w i t h i n - c o n d i t i o n s degrees o f
f r e e d o m ; w i t h i n - s u b j e c t s degrees
o f f r e e d o m ; w i t h i n - t r e a t m e n t s de- the influence of the treatment under in-
grees o f f r e e d o m . Compare BETWEEN-
G R O U P S DEGREES O F F R E E D O M .
w i t h i n - g r o u p s d e s i g n see WITHIN-
SUBJECTS D E S I G N .
w i t h i n - g r o u p s m e a n s q u a r e an
index of random variability or error in
an ANALYSIS OF VARIANCE. It Is Calcu-
lated as the WITHIN-GROUPS SUM OF dependent-groups a n a l y s i s o f v a r i -
SQUARES divided by the WITHIN-GROUPS
DEGREES OF F R E E D O M , and it forms the
denominator of the F RATIO. Also called
mean square w i t h i n ; within-
c o n d i t i o n s m e a n square; w i t h i n -
groups variance; w i t h i n - s u b j e c t s
m e a n square; w i t h i n - s u b j e c t s v a r i -
ance; w i t h i n - t r e a t m e n t s error;
w i t h i n - s u b j e c t s degrees o f f r e e -
w i t h i n - t r e a t m e n t s m e a n square;
w i t h i n - t r e a t m e n t s variance. Com-
pare B E T W E E N - G R O U P S M E A N S Q U A R E .
w i t h i n - g r o u p s s u m o f s q u a r e s an
index of variability i n an ANALYSIS O F
V A R I A N C E that is used to determine the
WITHIN-GROUPS MEAN SQUARE. It Is Cal-
culated by adding together the squared
deviations of the individual observa-
tions (scores) on the D E P E N D E N T VARI-
ABLE from the relevant group mean. Also
called w i t h i n - c o n d i t i o n s s u m o f
squares; w i t h i n - s u b j e c t s s u m o f
squares; w i t h i n - t r e a t m e n t s s u m o f
squares. Compare BETWEEN-GROUPS
SUM OF SQUARES.
w i t h i n - g r o u p s v a r i a n c e see W I T H I N - lated-samples design; dependent-
GROUPS MEAN SQUARE.
within-subjects analysis o f vari-
a n c e a variation of the standard A N A L Y -
SIS OF V A R I A N C E that is applied to data
from a study in which the independent
variable has multiple levels and each
participant experiences each treatment
level or is otherwise measured more w i t h i n - s u b j e c t s f a c t o r the indepen-
than once (see WITHIN-SUBJECTS D E -
SIGN). Because such designs involve re-
cording multiple responses from the
same person, it is necessary to examine
how each individual varies in his or her
responses, so as to separate such unique
fluctuation from variation that is due to
vestigation. For example, a researcher
studying how amount of daily walking
(e.g., none, 30 minutes, 60 minutes, 90
minutes) affects quality of sleep might
have participants walk each length of
time across consecutive weeks and then
evaluate the results using a within-
subjects analysis of variance. Also called
ance; dependent-samples a n a l y s i s
o f variance; w i t h i n - g r o u p s a n a l y -
sis o f variance; repeated measures
analysis o f variance; treatments-
by-subjects a n a l y s i s o f v a r i a n c e .
Compare BETWEEN-SUBJECTS A N A L Y S I S
OF VARIANCE.
d o m see W I T H I N - G R O U P S DEGREES O F
FREEDOM.
w i t h i n - s u b j e c t s d e s i g n an experi-
mental design in which the effects of
treatments are seen through the com-
parison of scores of the same participant
observed under all the treatment condi-
tions. For example, teachers may want
to give a pre- and postcourse survey of
skills and attirades to gauge how much
both changed as a result of the course.
Such a design could be analyzed with a
DEPENDENT-SAMPLES T TEST, a WITHIN-
SUBJECTS ANALYSIS OF VARIANCE, OF an
ANALYSIS OF COVARIANCE. Also Called
correlated-groups design; corre-
groups design; dependent-samples
design; related-measures design;
related-samples design; repeated
measures design; treatment-by-
subjects design; w i t h i n - g r o u p s de-
sign. Compare BETWEEN-SUBJECTS D E -
SIGN.
dent variable under study i n a W I T H I N -
 worm plot

SUBJECTS ANALYSIS OF VARIANCE. ThlS ables that is readily subject to revision

variable has multiple levels to which upon further experimentation. See HY-
each participant is exposed. For exam- POTHESIS.
ple, if a researcher is interested in job
w o r m plot a visual examination of
performance differences as a function of
the differences between two GROWTH
workshift length (e.g., 8 hours, 12 hours,
CURVE distributions, used for assessing
24 hours) and has each participant work
the GOODNESS OF FIT of the data. It is a
each shift length during the study, then
general diagnostic tool for the analysis
workshift length is a within-subjects fac-
of RESIDUALS. The vertical y-axis dis-
tor. Also called within-subjects vari-
plays the difference between each obser-
able. vation's location in the theoretical and
within-subjects mean square see empirical distributions, while the hori-
WITHIN-GROUPS MEAN SQUARE. zontal A;-axis depicts the z SCORES for dif-
ferent age groups, as in the following
within-subjects sum o f squares see
generic example.
WITHIN-GROUPS SUM OF SQUARES.
within-subjects variable see WITH-
IN-SUBJECTS FACTOR.
within-subjects variance see WITH-
IN-GROUPS MEAN SQUARE.
within-treatments error see WITH-
IN-GROUPS MEAN SQUARE.
W M D S abbreviation for weighted multi-
dimensional scaling. See INDIVIDUAL-
DIFFERENCES SCALING.
w o r k decrement in an experiment, a
decline in the magnitude of responses as
a function of frequency of the response.
w o r k i n g hypothesis a provisional but A well-fitting plot for a given age group
empirically testable statement about the is linear with zero SLOPE and zero curva-
relationship between two or more vari- ture.

415
X a letter used to symbolize a variable of ample, in the ordered pair (4, 8), the x-
interest. For example, in a REGRESSION coordinate is 4 and the corresponding
EQUATION, X may represent a particular vertical Y-COORDINATE value is 8. See
predictor or INDEPENDENT VARIABLE. CARTESIAN COORDINATE SYSTEM.

x' symbol for x PRIME. x-intercept n. in an equation represent-

ing a straight-line relationship between
X (X bar) symbol for SAMPLE MEAN. two variables, the value of variable x
when the value of variable y equals zero.
X-axis n. the horizontal axis on a graph. For example, in the general LINEAR
See ABSCISSA. EQUATION format x = a + by,a represents
the A:-intercept.
x-coordinate n. the horizontal value in
a pair of graph coordinates (x, y), which X prime (symbol: x') the predicted or EX-
indicates how far to move left or right PECTED VALUE of a given variable of in-
from the origin along the X-AXIS. For ex- terest, X.

416
 Yy
y symbol for a variable of interest. For ex- called acquiescence bias; acquies-
ample, in a REGRESSION EQUATION, ycent response set; response acqui-
may represent a particular outcome or escence. Compare NAY-SAYING.
DEPENDENT VARIABLE.
y-intercept n. in an equation represent-
Y symbol for Y PRIME. ing a straight-line relationship between
Yates's correction for continuity two variables, the value of variable y
an adjustment made to a CHI-SQUARE when the value of variable x equals zero.
TEST of data from a CONTINGENCY TABLE For example, in the general LINEAR EQUA-
having only two columns and two rows TION format y = a + bx,a represents the y-
of information. The Yates's correction intercept.
yields a more conservative chi-square
statistic and improves the test's accuracy yoked-control group a CONTROL
by accounting for the fact that it uses a GROUP in which each participant is
CONTINUOUS DISTRIBUTION to approxi- paired with a participant in another
mate a DISCRETE DISTRIBUTION. To group (see MATCHED-PAIRS DESIGN); the
apply the correction, one takes each paired individuals are then exposed to
value within the table (i.e., the observed precisely the same experiences except
frequency) and subtracts the value pre- for the specific treatment or other condi-
dicted to occur by a theoretical model tion under study. This procedure is in-
(i.e., the expected frequency). One then tended to make the control group as
takes the difference so obtained and sub- similar as possible to the experimental
tracts 0.5, squares the total, and divides group. For example, in a study of the ef-
the resulting number by the expected fectiveness of rewards upon children's
frequency. When such a correction is ap- learning, matched individuals in the
plied, the term Yates-corrected chi- control and experimental groups would
square test or Yates chi-square test is complete the same tasks under the same
used to refer to the test itself. [Frank conditions but only those in the experi-
Yates (1902-1994), British statistician] mental group would be praised for good
performance.
y-axis n. the vertical axis on a graph. See
ORDINATE. Youden's index a measure of the ca-
y-coordinate n. the vertical value in a pacity of a diagnostic test to correctly
pair of graph coordinates (x, y), which identify individuals with a certain illness
indicates how far to move up or down and to correctly identify those who do
from the origin along the Y-AXIS. For ex- not have the illness. It ranges from 0 to
1, with values closer to 1 indicating both
ample, in the ordered pair (4, 8), the y-
greater SENSITIVITY and SPECIFICITY of
coordinate is 8 and the corresponding
the test. Also called Youden's J . [Wil-
horizontal X-COORDINATE value is 4. See
liam John Youden (1900-1971), Austra-
CARTESIAN COORDINATE SYSTEM.
lian-born U.S. chemist and statistician]
yea-saying n. answering questions posi-
tively regardless of their content, which y prime (symbol: y') the predicted or EX-
can distort the results of surveys, ques- PECTED VALUE of a given variable of in-
tionnaires, and similar instraments. Also terest, y.

417
Yule-Simpson effect
Yule-Simpson effect see SIMPSON'S
PARADOX.
Yule's paradox see SIMPSON'S PARA-
DOX. [George Udny Yule (1871-1951),
British statistician]
Yule's Q a measure of the strength of
the relationship between two DICHOTO-
418
MOUS VARIABLES, such as a person's sex
(male or female) and graduation from
college (yes or no). It ranges from -1 to
+1, with values close to either negative
or positive 1 indicating a very strong re-
lation and values around 0 indicating
little to no relation. Also called Yule's
coefficient of association. [George
Udny Yule]
z-axis n. the third dimension in a CARTE- counts of zero (i.e., no boxes sold) be-
SIAN COORDINATE SYSTEM or graph. It is cause several teams did not go out to seU
perpendicular to both the horizontal x- due to inclement weather in their neigh-
AXis and the vertical Y-AXIS. borhoods; however, zero counts also
could occur because of nonsuccess in
Zelen's design a type of RANDOMIZED selling. Consequently, the, number of
CLINICAL TRIAL in which patients are as- zeros may be inflated and the number of
signed to experimental conditions be- teams not selling boxes cannot be ex-
fore they have consented to participate plained in the same manner as the num-
in the research. In a typical trial, patients ber of teams selling more than one box.
are first informed of the nature of the
study and then decide whether to partic- zero m a t r i x see NULL MATRIX.
ipate, which may result in certain biases
that Zelen's design attempts to over- zero-order correlation a simple asso-
come. There are two versions of the de- ciation between two variables that does
sign, both of which remain ethically not control for the possible influence of
controversial: In the single-consent other variables. For example, consider
strategy, patients assigned to the experi- the relationship between success selling
mental treatment are told there is an al- computers and knowledge of how the
ternative (the control) available and Internet works. A zero-order correlation
offered the option to switch or remain in would examine the direct relationship
the experimental group; those in the between these two variables without
control group are not informed of their taking into account other explanatoty
srady participation and their agreement information, such as education level,
to participate thus is not obtained. In sales experience, and so forth. The actual
the double-consent strategy, both the index of the magnirade or degree of
experimental and control groups are of- such a relationship is called the zero-
fered the option to switch from the order correlation coefficient; it ranges in
treatment to which they initially were value from -1 to +1, with the former in-
assigned. Also called randomized con- dicating a perfect negative relationship
sent design. [Marvin Zelen (1927- ), and the latter a perfect positive relation-
U.S. biostatistician] ship. Compare PART CORRELATION; PAR-
TIAL CORRELATION.
zero-inflated Poisson regression
(ZIP) a model for analyzing a data set zero-sum game in GAME THEORY, a
with an excessive number of zero out- type of game in which the players' gains
comes. An alternative to regular POIS- and losses add up to zero. The total
SON REGRESSION for explaining outcome amount of resources available to the par-
variability, it assumes that with proba- ticipants is fixed, and therefore one
bility p the only possible observation is player's gain necessarily entails the oth-
zero (0), and with probability 1 - p, a ers' loss. The term is used particularly in
POISSON VARIABLE Is obsetved. For ex- analyses of bargaining and economic be-
ample, consider a study of 100 teams havior but is sometimes also used in
selling boxes of cookies on a weekend. In other sociocultural contexts (e.g., poli-
this data set, there could be excessive tics).

419
zero-truncated Poisson distribution

zero-truncated Poisson distribu- of scores with different underlying char-

t i o n see TRUNCATED POISSON DISTRIBU- acteristics by converting them into val-
TION. ues that have a distribution with a mean
ZIP abbreviation for ZERO-INFLATED
of 0 and a standard deviation of 1. It is
POISSON REGRESSION. distinct from FISHER'S R TO Z TRANSFOR-
MATION, which is specific to PRODUCT-
z score the STANDARDIZED SCORE that MOMENT CORRELATION COEFFICIENTS.
results from applying a Z-SCORE TRANS-
FORMATION to raw data. For purposes of z test a type of statistical test that com-
comparison, the data set is converted pares the means of two different samples
into one having a distribution with a to determine whether there is a signifi-
mean of 0 and a standard deviation of 1. cant difference between them (i.e., one
For example, consider a person who not likely to have occurred by chance).
scored 30 on a 40-item test having a Generally, this involves comparing the
mean of 25 and a standard deviation of mean from a sample of a population to
5, and 40 on an 80-item test having a the mean for the whole population but
mean of 50 and a standard deviation of may also involve comparing the means
10. The resulting z scores would be +1.0 of two different populations. The z test is
and -1.0, respectively. Thus, the individ- based on the NORMAL DISTRIBUTION and
ual performed better on the first test, on is used when a population's STANDARD
which he or she was one standard devia- DEVIATION is known or the sample is
tion above the mean, than on the sec- large (greater than 30). The equivalent T
ond test, on which he or she was one TEST is used with unknown standard de-
standard deviation below the mean. A viations or smaller samples. Different
tabular or graphical arrangement of sev- variations of the basic z test exist, such
eral z scores is called a z-score distribu- as the z test for a population proportion,
tion. which compares differences between
PROPORTIONS as opposed to means; the
z-score t r a n s f o r m a t i o n a statistical 2 test for two dependent samples,
procedure used to convert raw data into which compares differences between
z SCORES, dimensionless quantities that two samples whose members have been
may be interpreted without reference to matched on certain characteristics; and
the original units of measurement. It is the z test for two independent samples,
performed by subtracting each data which compares differences between
point from a reference value (the sample two distinct, unrelated groups of partici-
average) and dividing the difference by pants.
the STANDARD DEVIATION of the sample.
The primary purpose of the z-score z t r a n s f o r m a t i o n see FISHER'S R TO Z
transformation is to allow comparisons TRANSFORMATION.
Appendixes
 Abbreviations and Acronyms

ACE alternating conditional expectation

ACES active control equivalence study
ACF autocorrelation function
AD average deviation
AEq age equivalent
AH alternative hypothesis
AIC Akaike's information criterion
AID automatic interaction detector
AML asymmetric maximum likelihood
ANCOVA analysis of covariance
ANOVA analysis of variance
ARIMA model autoregressive integrated moving-average model
ARMA model autoregressive moving-average model
AR model autoregressive model
ATI aptirade-treatment interaction
AUC area under the curve
AWS test Aspin-Welch-Satterthwaite test
BCfl or BCa bias-corrected accelerated percentile interval
BIC Bayesian information criterion
BLUE best linear unbiased estimator
BMA Bayesian model averaging
BRR balanced repeated replication
CART analysis classification and regression tree analysis
CDF cumulative distribution function
CF cumulative frequency
CFA confirmatoty factor analysis
CFI Bentier comparative fit index
CGF cumulant generating function
CHAID chi-square automatic interaction detector
CI confidence interval
CLT central limit theorem
Co comparison stimulus
CTT classical test theoty
CuSum chart cumulative sum chart
DA data augmentation
DAG directed acyclic graph

423
Abbreviations and Acronyms

DFBETAS differences in beta values

DFFITS difference in fits
DIG deviance information criterion
DIF differential item functioning
DS dual scaling
Duncan's MRT Duncan's multiple range test
DV dependent variable
e exponent; error
ECM algorithm expectation-conditional maximization algorithm
ECME algorithm expectation-conditional maximization either algorithm
EDA exploratory data analysis
EFA exploratory factor analysis
EMA ecological momentary assessment
EM algorithm expectation-maximization algorithm
EPSEM equal probability of selection method
erf error function
ESM experience-sampling method
exp exponential function
FA factor analysis
FDA functional data analysis
FU plot follow-up plot
GAM generalized additive model
GAMM generalized additive mixed model
GARCH generalized autoregressive conditional heteroscedasticity
GCR greatest characteristic root
GEE generalized estimating equation
GIGO garbage in, garbage out
g inverse generalized inverse
GLM general linear model; generalized linear model
GLMM generalized Unear mixed model
GLS generalized least squares
GMM generalized mixed model
HLM hierarchical linear model
HSD test honestly significant difference test
ICC intraclass correlation coefficient; item characteristic curve
IDA initial data analysis
UD independent and identically distributed
INDSCAL individual-differences scaling
IPA interpretive phenomenological analysis
IQR interquartile range
IRB institutional review board
IRT item response theory
ITI intertrial interval
 Abbreviations and Acronyms

IV independent variable
KMO test Kaiser-Meyer-Olkin test of sampling adequacy
K-R 20 Kuder-Richardson formula 20
K-R 21 Kuder-Richardson formula 21
LCA latent class analysis
LLR log-likelihood ratio
In natural logarithm
LOCF last observation carried forward
loess local regression
log logarithm
lowess local regression
LR likelihood ratio; logistic regression
LSD least significant difference
LSD test Fisher least significant difference test
LTA latent transition analysis
MA moving average
MAD mean absolute deviation; median absolute deviation
MA model moving-average model
MANCOVA multivariate analysis of covariance
MANOVA multivariate analysis of variance
MAPE mean absolute percentage error
MAR missing at random
MARS multivariate adaptive regression spline
MCA multiple classification analysis
MCAR missing completely at random
M C M C method Markov chain Monte Carlo method
M C M L estimation Monte Carlo maximum likelihood estimation
MCP minimum convex polygon
MDA multiple discriminant analysis
MDL minimum description length
MDS multidimensional scaling
MI modification index
MIMIC model multiple indicators-multiple causes model
MLE maximum likelihood estimation
ML-EM estimation maximum likelihood-expectation maximization estimation
MOE margin of error
MTMM multitrait-multimethod matrix; multitrait-multimethod
model
MVUE minimum variance unbiased estimator
NCE normal curve equivalent
NFl normed fit index
1
NH null hypothesis
NHST null hypothesis significance testing

425
Abbreviations and Acronyms

NNT number needed to treat

NS not significant
OLS ordinary least squares
OR odds ratio
O-X-O one-group pretest-posttest design
PCA principal components analysis
PDF probability density function
PGF probability generating function
PLS partial least squares
PMF probability mass function
PMLE penalized maximum likelihood estimation
P-P plot probability-probability plot
PRE proportional reduction of error
p value probability level
QDA quadratic discriminant analysis
QI quasi-independence
Q-Q plot quantile-quantile plot
R response or respondent
rad radian
RBD randomized block design
R correlation multiple correlation coefficient
RCT randomized clinical trial
RDD random-digit dialing; regression-discontinuity design
REG random event generator
REGWQ test Ryan-Einot-Gabriel-Welsch multiple range test
REML restricted maximum likelihood
RMS root mean square
RMSE root-mean-square error
RMSR root-mean-square residual
RNG random number generator
ROC curve receiver-operating characteristic curve
RRT randomized-response technique
RSM response surface methodology
RT reaction time
S subject
SDT signal detection theory
SEM structural equation modeUng
SIMEX simulation and extrapolation procedure
SQRT sequential probability ratio test
SSA smallest space analysis
STARIMA model space-time autoregressive integrated moving-average model
STARMA model space-time autoregressive moving-average model
SVD singular value decomposition

426
 Abbreviations and Acronyms
Acronyms

TAR threshold autoregression

autoregression
TTl
TTI trait-treatment interaction
UMP test uniformly most powerful
powerful test
test
VIF variance inflation factor
factor
WMDS
VyMDS weighted multidimensional scaling
scaling
ZIP
ZIP zero-inflated
zero-inflated Poisson
Poisson regression

427
 Entry Illustrations

Andrews plot 11
ANOVA summary table 11
array 14
back-to-back stem-and-leaf plot 18
bagplot 19
balanced Latin square 19
bar graph 20
bimodal distribution 27
biplot 29
box-and-whisker plot 32
bubble plot 34
Cartesian coordinate system 36
centile reference chart 40
chain graph 41
Chernoff faces 42
circumplex 44
classification table 46
classification tree 46
complete block design 53
component bar graph 54
compound bar graph 55
confidence band 57
contingency table 61
contour plot 63
correlation matrix 66
covariance matrix 67
cumulative frequency distribution 75
cumulative frequency polygon 76
cumulative relative frequency distribution 77
cumulative relative frequency graph 77
cumulative sum chart 78
cycle plot 79
decision tree 84
dendrogram 87
differences vs. totals plot 93

428
 Entry Illustrations

directed acyclic graph 95

directed graph 95
dual bar graph 105
Euler diagram 118
factor pattern matrix 130
flowchart 140
follow-up plot 141
forest plot 142
frequency distribution 145
frequency polygon 145
funnel plot 147
Graeco-Latin square 155
graphic rating scale 156
growth curve 159
hanging rootogram 160
histogram 166
historigram 167
hyperbolic distribution 169
identity matrix 170
intercept 178
inverse matrix 182
item characteristic curve 184
J curve 186
jittering 187
joint distribution 187
Kronecker product of matrices 191
Latin square 194
leptokurtic 195
line graph 198
Lorenz curve 202
main diagonal 204
matrix , 208
mesokurtic : 213
minimum volume ellipsoid 216
multidimensional scaling 223
multimodal distribution 224
multitrait-multimethod matrix 227
multivariate distribution 229
normal distribution 238
ogive 243
operating characteristic curve 245

429
Entry Illustrations

Pascal's triangle 255

path diagram 256
personality profile 258
pie chart 260
platykurtic 262
quadrant 282
quality control angle chart 284
radial plot 288
receiver-operating characteristic curve 298
regression Une 302
relative frequency distribution 306
relative frequency polygon 306
residual matrix 313
response surface methodology 314
rootogram 318
rugplot 320
scatterplot 325
scatterplot matrix 326
scree plot 328
singular matrix 339
skewness 340
skew-symmetrical matrix 340
stalactite plot 349
survival curve 366
symmetrical matrix 367
synthetic risk map 368
three-way table 380
trait profile 385
transition matrix 386
transpose 386
tree diagram 388
two-by-two table 392
undirected graph 398
U-shaped distribution 403
variogram 407
Venn diagram 407
vertical line graph 408
visual analogue scale 409
worm plot 415

430
Overview of Research Design Considerations

Ol S -S
QJ tl.
S Q >
Q -g EEs
is S E
<j i z
0) J - 0,
5 5 S .2 S -
fig
*S (0 OJ ro 0) ~
C3 Z cc
0)
Sa.

C < op O OJ

S< S
Q o mau 3

i CD s S s I
OS . .

S .5
ol
; UJ

5 '

c
>-
o <^
srtStu
i\ stud
idinal

c
>
'5) uo n
rve

c <b
o a

s .a - "2 E S .2
- c nj u
ra .y o
2 S g- E S^f
if c o
"1 2 w t 3 . c o o

2. " ^ 2. > " 2 w ^ S a .2 s.

2 o
- s "=
I m <
I ^ s H ^ 2 11 i . > 3 -S z
> I- . u oi
i < g i'l
) . O <
5 1.. : ^"^ S a,

431
 Symbols

Latin
adj R^ adjusted R^
C coefficient of contingency
d Cohen's d; Glass's d
d' d prime
D Cook's distance; difference score
D2 Mahalanobis distance
df degrees of freedom
E expected value
f an effect size index; frequency
F F ratio
likelihood-ratio chi-square
communality coefficient
Ho null hypothesis
H, alternative hypothesis
Wa alternative hypothesis
H hat matrix
In Williams's agreement measure
I identity matrix
k coefficient of alienation; number of units in a statistical analysis
K? D'Agostino test statistic
M sample mean
MS mean square
MSE mean squared error
MSR mean-square residual
MSSD mean-square successive difference
n number of scores or observations from a particular
experimental condition
N total number of cases (participants) in an experiment
P probability
q Studentized range statistic; probability of failure in a binary trial
Q Cochran Q test statistic; Yule's Q; quartile
r product-moment correlation coefficient; sample correlation
coefficient
r2 coefficient of determination
R multiple correlation coefficient
coefficient of multiple determination

432
 Symbols

adjusted 7?^
biserial correlation coefficient
''bis biserial correlation coefficient
Rc canonical correlation coefficient
^effect size effect-size correlation coefficient
''pb point biserial correlation coefficient
point biserial correlation coefficient
Spearman correlation coefficient
tetrachoric correlation coefficient
reliability coefficient
RK cross-validated multiple correlation
sample standard deviation
sample variance
5D standard deviation
S standard error
SEE standard error of estimate
SEM standard error of measurement; standard error of the mean
SS sum of squares
SSCP sum of squares of cross-products
SSE sum of squared errors
55, regression sum of squares
regression
T Wilcoxon signed-ranks test statistic
J2 Hotelling's test
TSS total sum of squares
U Mann-Whitney U test statistic
V Cramer's V
W coefficient of concordance; Wilcoxon rank-sum test statistic
W2 Cramer-von Mises goodness-of-fit test statistic
X a variable of interest
x' X prime
X sample mean
X data matrix
y a variable of interest
y' y prime
A
predicted value of an outcome or dependent variable
y

Greek
a alpha
P beta
t chi-square

r distribufion chi-square distribution

A change in a parameter; Glass's d

433
Symbols

e a small value
z?- proportion of shared variance
r| correlation ratio; a latent variable
T|2 amount of explained variance
TIG^ generalized eta squared
T|p2 partial eta squared
Y Goodman-Kmskal's gamma
K Cohen's kappa
X eigenvalue
A Wilks's lambda
|i population mean
co^ omega squared
<S) phi coefficient
Is)^ Cramer's V
K ratio expressing the circumference of a circle to its diameter;
probability of success in a trial
p population correlation coefficient; Spearman correlation
coefficient
o population standard deviation
population variance
OM standard error of the mean
S sum
T Kendall's tau
e ability parameter

434