Professional Documents
Culture Documents
REFERENCES:
LECTURE NOTES OF DR. GLENN SIA SU, PH.D.
PRACTICE OF RESEARCH BY DR. REBULLIDA ET AL.
RESEARCH METHODOLOGY BY GOMEZ AND GOMEZ
RESEARCH DESIGNS IN HTTPS://EXPLORABLE.COM/RESEARCH-DESIGNS
RESEARCH METHODS BY LINDA M. WOOLF
INTRODUCTION TO STATISTICS AND PROBABILITIES BY MIKE DONNELL
PARAMETRIC VS NON-PARAMETRIC IN HTTP://WWW.CSSE.MONASH.EDU.AU/
BIOLOGY STATISTICS MADE SIMPLE USING EXCEL BY NEIL MILLAR
COURSE MODULE IN BIO 180 GENERAL BIOSTATISTICS BY ATA SANTIAGO, GL SIA SU AND JG VICENCIO
THESIS AND DISSERTATION MANUAL OF FORMAT AND STYLE BY BORJA ET AL. THRU UP MANILA NGOHS
WHAT IS RESEARCH?
Descriptive Research
Primarily descriptive
Aims to portray selected characteristics of the subject accurately to
study their stability, distribution & occurrence
Provide necessary background for the formulation of a more precise
problem for subsequent more specific study & for the development of
hypotheses.
Provide the means for obtaining initial ideas about the interrelationships
among phenomena, & the types of determinants that might relevantly be
measured.
Exploratory research
Get extensive information about a phenomena, discovering new
dimensions & uncovering additional aspects of it.
Provides an opportunity to study the entire structure of problem under
investigation.
Supplies the needed knowledge & experience that will aid in setting up
fruitful hypotheses for more detailed studies.
RESEARCH
Explanatory-Experimental Research
Cause & effect sequences among phenomena
Seeks to find explanation of phenomena (effects) that one observes around him & their many
possible causes.
Aims to find out why things are as they are.
Illustrate 4 possible relationships
No relationship: If A occurs, B may or may not occur (B does not depend on A in any way)
With relationship: If A occurs, then B will occur.
Positive relationship: If A increases, then B increases
Negative relationship: If A increases, then B decreases
Evaluative research
Aims to assess the consequences of activities undertaken to advance some valued goals.
Seeks to determine whether activity or program being evaluated produced or is now producing
the desired effects.
Based on nature of problem and its related research goals.
Classified as experimental or non-experimental.
Experimental Research:
Experiment is a method designed for collecting data under
controlled conditions.
In operation when there is actual human interference with
the conditions that determine the phenomenon under
study.
Investigator intentionally manipulates 1 or more of the
independent variables (presumed causes) and observes the
changes in the dependent variables (presumed effects).
DIFFERENT TYPES OF RESEARCH
Experimental Research:
Independent variables can all be different characteristics or
can refer to different degrees or levels of one or more
independent variables.
Undertaken to determine a particular characteristic of a
population or to compare the effects of 2 or more
procedures called treatments
Absolute experiment common in physical sciences.
Comparative experiment commonly employed in
agriculture, education, psychology, sociology, social work &
other related fields.
Non-experimental Research
No manipulation of the independent & control variables
by the investigator as is characteristic of
experimentation.
Investigator seeks out those units or subjects which
exhibit the desired effects he is interested in, or who
have already been exposed to the stimuli under study or
to different levels of the independent variable.
Not concerned with analysis of effects of certain
experimental treatments on individuals but focuses on
reconstruction of process that occurred prior to
investigation
Differentiated to clinical non experimental research and
non experimental field research
Major Classification of Nonexperimental Research:
Case Study, Census & Survey
Case Study: Means of organizing social data with the view to preserve the
wholeness of the social object being studied.
Social object may be person, family, a social group or process or
culture.
A wide range and depth of experience
Advantage: Uniqueness and peculiarities rather than generalizations.
Disadvantage: Time and Cost
Generalizations cannot be made from a few cases to the population.
Census: Complete enumeration involves gathering of data on one or more
characteristics from all the units of the population.
Inherent difficulties in operational procedures
Used to obtain population data and the like at regular intervals of
time.
Major Classification of Nonexperimental Research:
Survey: Based on a sample of the population considered.
Economize the use of limited resources in gathering data on relatively
large population or universe units such as housheolds, business
establishments, farms, persons & so on.
Results provide estimates of the characteristics of the population, and if
the sample is chosen so that every element of the population has a known
chance of being included in the sample, the precision of the estimates can
be measured in relation to the data that would be obtained if the entire
population were enumerated.
Difference between sample estimates and true values of the
characteristics of the population under study can be minimized by a
proper survey design & by an adequate sample size.
Reliability of estimates of population characteristics made from a sample
survey depends on effective implementation of a sound & operationally
practical survey design or plan.
RESEARCH PROCESS
Step 1: Identification and Formulation of the Research Problem
Step 2: Determination of the Researchability of the Problem and the Value
of the Research
Step 3: Intimate Acquaintance with the Research Problem
Step 4: Formulation of the Hypotheses & Statement of the Assumption of
the Study
Step 5: Choice of Empirical Variables
Step 6: Determination of Possible Research Obstacles
Step 7: Choice of Research Method
Step 8: Preparation of a Detailed Research Design
Step 9: Collection of Data
Step 10: Analysis and Interpretation of Data
Step 11: Drawing of Conclusions & Recommendations
Step 12: Reporting the Research Findings
RESEARCH PROCESS
Step 2: Determination of the Researchability of the Problem and the Value of the
Research
Research topic must be within the limits of ones resources.
Problem must be manageable.
Step 3: Intimate Acquaintance with the Research Problem
Efforts should be exerted to acquire a wide & deep understanding of the nature of the
problem studied.
RESEARCH PROCESS
Step 4: Formulation of the Hypotheses & Statement of the Assumption of the Study
Hypothesis is an assertion or proposition about an unknown phenomenon; e.g., about the
relationship of 2 or more variables, the validity of which is to be tested in the course of the
research.
Formulation of hypotheses used certain assumptions.
Statements are taken to be true, without verification in the study in which it is made.
May be theoretically derived or based on commonly accepted as facts or findings of previous
research.
Hypotheses can be formulated by means of a functional relationship between variables which
have been operationally defined.
RESEARCH PROCESS
Summary:
Identify the problem
Ask a clear question
Limit the scope of the research
Say what is known and not known about the problem
Identify the skills and information needed to answer the question
Select the methods needed to collect & analyze the data
Explain the kind of results expected and how they will be presented
Specify the timetable and budget
PREVIOUS DISCUSSION
Introduction
Definition of Research.
Truth, accuracy, validity and reliability.
Role and Importance of Research.
Characteristics of Good Research (4 Elements).
Outline, sampling design, data gathering plan and plan for the analysis of
collected data.
Different Types of Research
Basic or Applied (Research problem)
Descriptive, Exploratory, Explanatory-Experimental and Evaluative (General)
Experimental and Non-Experimental (Nature of problem and related research
goal).
The Research Process
12 Steps
Possible Obstacles
THE RESEARCH QUESTION
FOCUSING ON RESEARCH PROBLEM
If there is no focused research problem
There is no research.
We do not get approval nor funding.
We lurk around with an unfinished academic degree.
Or give up on the idea of research as necessary to professional success.
Choose your research adviser.
Select a person who is competent in your topic and has the available
time and commitment to assist you in the entire research process.
Ask yourself if you can have a meaningful interaction with your research
adviser.
Consider the teamwork and contribution of members of your panel to
your research.
RESEARCH QUESTION
Objective of study
Uncertainty you want to resolve.
Begins with a general concern researchable issue.
RESEARCH QUESTION
Feasible
Number of subjects
Technical expertise
Time and Money
Manageable scope
Interesting
To the investigator
CHARACTERISTICS OF A GOOD RESEARCH QUESTION
Novel
Confirms or refutes previous finding
Extends previous findings
Provides new findings
Ethical
Relevant
To scientific knowledge
To clinical, health and environmental policy
To future research directions
STEPS YOU CAN FOLLOW
1. Choose an interesting topic where you already have some information. Avoid
topics that you are just encountering for the first time.
2. Survey existing studies to discover the many sides of the topic.
3. Consider what is still unknown about the topic.
4. Ask questions and see if these have been studied and answered in previous
research.
5. Choose at least one question you can manage to research given a period of time,
resources, access to and availability of data.
6. Consider how you can contribute to confirming existing theories. How your
research can generate and add to knowledge.
7. State your research problem in a sentence showing what you want to describe,
explain or predict.
8. Start developing your hypothesis and identifying your variables. The variables
will be needed at the early stage of defining the problem.
DEVELOPING THE RESEARCH QUESTION AND THE
STUDY PLAN
Problems Solutions
Not feasible Narrow the question
Too broad Specify a smaller set of
variables
Not enough subjects
Expand & modify criteria
Lengthen time frame
Collaborate with colleagues
Methods beyond skill of
investigator Consult experts
Learn the skill
RESEARCH QUESTION AND STUDY PLAN:
PROBLEMS AND SOLUTIONS
Solutions
Problems Consider less costly study
Not feasible designs
Too expensive Fewer subjects &
measurements
Less extensive measurements
Fewer follow-up visits
Not interesting, novel or Consult with mentor
relevant
Modify research question
Uncertain ethical suitability Consult with institutional
review board
Modify research question
RESEARCH QUESTION AND STUDY PLAN:
PROBLEMS AND SOLUTIONS
Solutions
Write research question at an
early stage
Problems Get specific like:
Study Plan is vague How subjects will be
sampled?
How variables will be
measured?
PRIMARY AND SECONDARY QUESTIONS
Advantages
Efficiency in results
Disadvantages
Increasing complexity in designing, implementing and drawing statistical
inferences from the study with multiple hypothesis.
SENSIBLE STRATEGY!!!!
Categories of variables:
Independent presumed to cause, effect, influence or stimulate
the outcome.
Dependent refers to the output, the outcome or the response
variable.
Control may produce changes which may be mistaken to be
the effect of the independent variables being considered
Controlled, held constant or randomized so the effects are
neutralized, cancelled out or equated for all conditions.
Confounding/extraneous/intervening variable observed
event/factor that is expected to possibly affect the relation
between the independent and dependent variable.
UNDERSTANDING VARIABLES
Stating the research objective means presenting what you want to achieve.
You determine the reasons why you are dealing with the research problem.
Objectives are usually expressed in the form of statememts of expected
outcomes.
Some helps in writing research objectives:
Your general objective should be related to the main problem.
You can divide the main problem into sub-problems. Specific objectives can be stated
in relation to sub-problems.
Your hypothesis will lead you stating expected outcomes.
Characteristics of a Good Research Objectives:
Phrased clearly, unambiguously and specifically.
Focus on a clear goal to give direction to the research process
Stated in measurable terms and should not involve value judgement.
HYPOTHESIS
FORMULATE TESTABLE HYPOTHESIS
Hypothesis must be expressed in a concrete and clear manner, and stated in
terms of observable and measurable behavior, allowing objective evaluation
of results.
Provides guidance in the search for evidence by way of:
Limiting area of investigation
Sensitizing the researcher to pertinent data and relationships
Providing a unifying concept
Null hypothesis
Alternate hypothesis
SIGNIFICANCE OF THE RESEARCH
DEFENDING THE SIGNIFICANCE OF THE RESEARCH
Target population large set of people throughout the world to which results will
be generalized.
Example: All water bodies in Laguna
Study sample subset of the target population available for study
Example: Rivers in Laguna
INFERENCES IN GENERALIZATION
Infer
Design
Target Intended
Population sample
Advantages:
Probability samples are the only type of samples where the results can be generalized
from the sample to the population.
Allows the researcher to calculate the precision of the estimate as well as the
sampling error.
Disadvantages:
More difcult and costly to conduct.
NON-PROBABILITY SAMPLING METHODS
Judgmental/Purposive
A representative sample of the population is selected based on a researchers
"expert" judgment.
Prior knowledge and research skill are used in choosing the respondents or
elements to be sampled.
Selection of patients in a clinical trial by a medical specialist
Choice of participants based on a pre-test questionnaire or focus-group
discussion (FGD).
Haphazard/Accidental
Also known as convenience sampling, is a sampling design where the samples
are selected by an arbitrary method that is easy to carry out.
Friends as a sample of college students.
Ambush interviews of random people in the area.
Households of relatives or friends as sampling sites.
Quota
Dividing the population into predetermined classes then obtaining haphazard
samples of a xed size (quota) within each class.
Obtaining the RH bill opinion of 2O people per region in a municipality.
A researcher desires to obtain a certain number or respondents from different
income categories.
PROBABILITY SAMPLING METHODS
SIMPLE RANDOM SAMPLING
The most basic type of sampling design wherein every element in the population has
an equal chance of being included in the sample.
Steps to follow:
1. Prepare an exhaustive list (sampling frame) of all members of the population of
interest.
2. Decide on the size of the sample.
3. Select the necessary number of sampling units (lottery method, a table of random
numbers, or the RAN function of a calculator).
Advantages:
Simple design, easy to analyze.
Guarantees unbiased estimates of population parameters.
The average of the sample estimates over all possible samples is equal to the population
parameter.
Disadvantages:
Not cost efficient.
Requires a sampling frame or listing of all elementary units of the population which
might be costly and tedious to prepare.
Even though the probability is very small, it is possible that non-representative
samples can be drawn from the population.
PROBABILITY SAMPLING METHODS
SYSTEMATIC SAMPLING
The researcher selects samples at regular intervals (every 1st , 2nd , 3rd or kth ].
The sampling interval is computed as k = N/n where k is the sampling interval, N is the
population size, and n is the desired sample size.
Example:N = 20; n = 10
Sampling interval = 20/10 = 2 or every 2nd unit
Randomly draw a number from 1 to 20 = 3 for instance
Samples are selected every 2nd unit (example: households 3, 5, 7... and so on) until n = 10 is
reached.
Advantages:
Easier compared to simple random sampling .
A systematic sample may spread the members selected for measurement more evenly
across the entire population.
In some cases, systematic sampling may have better representativeness and more
precise.
It can allow the researcher to draw a probability sample without complete prior
knowledge of the sampling frame.
Disadvantages:
Can lead to difculties when the variable of interest is periodic (with period n).
For example, when conducting a sample of nancial records that follow a calendar schedule.
Setting the sampling interval to 7 would mean that all observations would fall on the same day of
the week.
This introduces bias into the sample since an inappropriate interval was chosen.
Periodic or list effect - a foregoing source of sampling error.
For example, if we used a very long list such as a telephone directory for our sampling frame.
PROBABILITY SAMPLING METHODS
STRATIFIED RANDOM SAMPIING
A modication of simple random sampling
Is used when we want to guarantee that each stratum (subgroup) constitutes an
appropriate portion or representation in the sample.
It involves categorizing the members of the population into mutually exclusive
and collectively exhaustive groups.
Simple random sampling is then independently carried out for each group.
Advantages:
Can be used to improve the accuracy of the sample estimates when there is prior
knowledge that the variability in the data is not constant across the subgroups.
Enable the researcher to determine the desired level of sampling precision for each
stratum.
Stratied random sampling produces an unbiased estimate of the population mean with
better precision compared to simple random sampling with the same total sample size
n.
Disadvantage:
May require a very large n if reliable estimates for each stratum are desired.
PROBABILITY SAMPLING METHODS
CLUSTER SAMPLING
A method of sampling in which the element selected is a group (rather than an
individual), called a cluster.
For instance, the clusters could be city blocks. Cluster sampling.
Advantages:
Cost-efficient.
Samples are easier to select if a sampling frame is available.
Disadvantages:
May require complicated analyses .
Sample size must be large enough to obtain representative estimates; of the
Parameters.
RESEARCH DESIGNS
DIFFERENT RESEARCH METHODS
There are various designs which are used in research, all with specific
advantages and disadvantages. Which one the scientist uses, depends on
the aims of the study and the nature of the phenomenon:
Case Study
Study of only one case of an outcome
Case Series
Study of 2 or more cases of the same outcome
Describe clinical features of a rare disease, a disease defying diagnosis, a
new disease, interesting cases.
Study only cases, no comparison groups
Relationship of the disease with an exposure factor cannot be
determined.
DIFFERENT RESEARCH METHODS
There are various designs which are used in research, all with specific
advantages and disadvantages. Which one the scientist uses, depends on
the aims of the study and the nature of the phenomenon:
Type of observational study where cases of a disease are compared with a similar
group who do not have the illness.
Groups are compared to history of exposure to some factor.
Undertaken to study rare diseases or conditions and determine sensitivity and
specificity of certain tests or criteria.
Straight forward : Goes with natural direction of exposure coming before the
development of disease.
Measure magnitude of disease relative to the exposure factor.
Subjects:
Entire population
Select people who are exposed and getting a comparable group of non-exposed
individuals
CROSS-SECTIONAL STUDIES
A type of observational study that involves the analysis of data collected from
a population, or a representative subset, at one specific point in time.
Aims to provide data on the entire population under study.
can be used to describe, not only the odds ratio, but also absolute risks and
relative risks from prevalences (sometimes called prevalence risk ratio.
May be used to describe some feature of the population, such as prevalence of
an illness, or they may support inferences of cause and effect.
Study relationships of an outcome to a specified exposure as the 2 variables
occur at the individual level.
Give a description of the occurrence of one factor with the presence of an
outcome.
Measure: Prevalence of both outcome and the exposure factor.
Simultaneous existence of 2 factors in an individual.
CROSS-SECTIONAL STUDIES
Inherent problem: Cannot determine which among the exposure or the outcome
occurs first (temporal ambiguity)
Formulate possible hypothesis regarding causation which can be tested by
analytical study designs.
Not expensive and easy to conduct.
DIFFERENT RESEARCH METHODS
There are various designs which are used in research, all with specific
advantages and disadvantages. Which one the scientist uses, depends on
the aims of the study and the nature of the phenomenon:
Type:
1. One-group posttest design only (one-shot case study) very poor design.
Difficult to determine if there was a change or not.
2. One-group pretest-posttest design attempt to detect changes by
comparing measurements before and after intervention.
Any change in measurement may still not be due to intervention.
QUASI EXPERIMENTAL DESIGN
Utilize control groups but falls short of being a true experiment because of:
Absence of random allocation in assignment of treatment/intervention.
QUASI EXPERIMENTAL DESIGN
Types:
1. Non-equivalent control group design control may have same
characteristics as experimental except that there was no randomization.
Variables may influence effect of exposure on the outcome.
REPLICATION
Commonly known as replicates
Number of experimental units in a single treatment at one time = number of replicates for
that treatment (e.g. 5 mice per treatment, 3 plants per treatment, etc.)
The mean of the observations from each replicate can be subjected to statistical tests or
inferences.
BLOCKING
Groups of experimental units sharing a common level of an extraneous variable.
Based on agricultural eld experiments where blocks such as plots of land share the same soil
conditions.
e.g. In comparing the effects of soil characteristics to plant yield: Experimental units were 3
plots of land and the treatments/ independent variable is variety of a particular plant. Plant
yield is the outcome of interest.
EXPERIMENTAL STUDIES
Source: Santiago, ATA, Sia Su GL and Vicencio, JG. ND. Course Module in
Bio 190 General Biostatistics. Department of Biology CAS UP Manila.
RESEARCH DESIGNS
Planning a research.
Represents the plan by which the investigator intends to pursue his/her
objectives.
Minimizes the effect of competing variables on the relationship between the
exposure factor and the outcome.
Embodies strategies by which desired results are obtained with minimum
expenditures of time, money and effort.
DATA ANALYSIS
HOW TO ANALYZE YOUR DATA?
RESEARCH PROCESS
Measurements
Sample or Population Data.
Organize and summarize.
Collection and interpretation of quantitative data and the use of probability
theory to estimate population parameters.
Descriptive statistics (Summarize and describe).
Important characteristics of a set of measurements.
Methods of tabulation, graphical presentation, computation of averages.
DATA
A collection of facts (measurements) from which conclusions may be
drawn.
VARIABLE
A characteristic that changes or varies over time and/or for different
individuals or objects under consideration.
Can assume any of a set of values.
Generates data.
TYPES OF DATA
Data
Qualitative Quantitative
Nominal Ordinal
NOMINAL VARIABLE
Is for mutual exclusive, but not ordered, categories.
Compare five different genotypes.
Code five genotypes with numbers but the order is arbitrary.
ORDINAL VARIABLE
The order matters but not the difference between values.
The amount of pain patients are feeling (on a scale of 1 to 10).
A score of 7 means more pain than a score of 5, and that is more than a score
of 3.
But the difference between the scores 7 and 5 may not be the same as that
between 5 and 3.
The values simply express an order.
OTHER CATEGORIES OF VARIABLES ACCORDING TO
SCALE OF MEASUREMENT
INTERVAL VARIABLE
Exact distance between 2 categories can be determined but the zero point is
arbitrary.
Difference between a temperature of 100 degrees and 90 degrees is the same
difference as between 90 degrees and 80 degrees.
RATIO VARIABLE
Has all the properties of an interval variable, and also has a clear definition of 0.0.
Exact distance between 2 categories can be determined but the zero point is fixed.
When the variable equals 0.0, there is none of that variable.
Variables like height, weight, enzyme activity are ratio variables.
CATEGORIES OF DATA ACCORDING TO HOW MANY
VARIABLES ARE MEASURED
UNIVARIATE DATA
Result when a single variable is measured on a single experimental unit.
e.g., record of a persons height.
BIVARIATE DATA
Result when two variables are measured on a single experimental unit.
e.g., record of a persons height and weight.
MULTIVARIATE DATA
Result when more than two variables are measured.
e.g., record of a persons height, weight, gender, and age.
ORIGIN AND DISPLAY OF DATA (DESCRIPTIVE STATISTICS)
NUMERICAL MEASURES
Can be calculated for either a sample or population measurements.
Calculate a set of numbers that will convey a good mental picture of the
frequency distribution.
DESCRIBING DATA WITH NUMERICAL MEASURES
5. Box plots
DESCRIBING DATA WITH NUMERICAL MEASURES
b. Samples of n measurements:
3. Standard deviation
a. Population:
b. Sample:
4. A rough approximation for s can be calculated as s R / 4.
The divisor can be adjusted depending on the sample size.
DESCRIBING DATA WITH NUMERICAL MEASURES
V. Box Plots
1. Box plots are used for detecting outliers and shapes of
distributions.
2. Q 1 and Q 3 form the ends of the box. The median line is in
the interior of the box.
3. Inner and outer fences are used to find outliers.
a. Inner fences: Q 1 - 1.5(IQR) and Q 1 + 1.5(IQR)
b. Outer fences: Q 1 - 3(IQR) and Q 1 + 3(IQR)
PARAMETRIC STATISTICS
Tests done for samples that are approximately normally distributed. Requires a mean
and variance. Requires the estimation of 1 or more population parameters.
Example: An estimate of the difference in the average between the 1st and
2nd populations.
NON-PARAMETRIC STATISTICS
Tests that can be done without the assumption of normality. These tests do not require
a mean and standard deviation. Does not involve estimation of a specific parameter.
Research participants are not passive sources of data, but individuals whose rights and
welfare must be protected.
ETHICAL PRINCIPLES
Principle of Beneficence
Research design must be scientifically sound
Risks are minimal and benefits are more
Risks: Physical or Psychosocial harm like breaches of confidentiality, stigma and
discrimination.
ETHICAL PRINCIPLES
Principle of Justice
Benefits and burdens of research be distributed fairly.
Equitable access to the benefits of research
Vulnerable populations should not be target for research if other populations would also
be suitable.
VULNERABLE PEOPLE
Children
Prisoners
Pregnant women, fetuses, embryos
People with impaired decision-making capacity
WHAT RESEARCH IS EXEMPT FROM
INSTITUTIONAL REVIEW BOARD APPROVAL?
TITLE PAGE
TABLE OF CONTENTS
INTRODUCTION
Background of the Study
Statement of the Problem
Research Objectives
Significance of the Study
Scope and Limitations
Definition of Terms
FORMAT AND STYLE
PRESENTATION OF RESULTS
Dummy tables and figures.
Properly labeled
LITERATURE CITED
APA Style
APPENDIX
Images
Accomplished IACUC Form (from NIH).
Test animal, plant samples.
Laboratory setup or layout.
Cage dimensions
Etc.
FORMAT AND STYLE