Professional Documents
Culture Documents
Research Method Nov 2012
Research Method Nov 2012
METHODOLOGY
1
I. INTRODUCTION TO
RESEARCH
2
INTRODUCTION…
RESEARCH:-
a way of examining your practice…
Research is undertaken within most professions.
More than a set of skills, it is a way of thinking:
examining critically the various aspects of your
professional work.
It is a habit of questioning what you do, and a
systematic examination of the observed
information to find answers with a view to
instituting appropriate changes for a more effective
professional service.
3
INTRODUCTION cont’d
1.1. What is research?
Research is the systematic collection, analysis
and interpretation of data to answer a certain
question or solve a problem. OR
5
INTRODUCTION cont’d
Philosophies means approaches e.g. qualitative,
or highlight something
7
1.2. Types (Categories) of research
8
2. Applied research
is problem oriented & directed towards
9
The philosophical approach to research
is basically of two types:
1. Empirical research
2. Theoretical research
11
2. Quantitative
Involves studies that make use of statistical
analyses to obtain their findings.
Key features include formal and systematic
measurement and the use of statistics.
It deals with objective measurable
behaviors and attitudes.
In most cases, health science research
deals with information of a quantitative
nature.
12
Characteristics of research
empirical evidence;
involves gathering new data from
14
Goals of Scientific Research
It is ultimately concerned with improving the
health of people and communities,
by enhancing the efficiency and effectiveness
of the health system as an integral part the
overall process of socio-economic
development, with full involvement of all
partners.
15
Research can be used for the purposes of:
a. Description:
discovering the distribution of certain
attributes about some population i.e
magnitude of health problem.
The researcher is concerned not with
why the observed distribution exists, but
merely with what the distribution is.
16
b. Explanation: why variables are relate,
simultaneous examination of two or more variables
i.e relationship b/n factors and disease outcome or
efficacy of an intervention(services)
17
Major differences among exploratory, descriptive, and explanatory research.
Exploratory Research Descriptive research Explanatory research
18
summary
There is only one type of research: good research.
Bad research does not deserve the name of research.
Badly done research is not only a waste of time,
money and effort.
It can be considered unethical if it exposes research
subjects to the inherent risks of experimentation with
no reward to them, to others or to their communities.
Good research ideas come from the knowledge,
work and attitudes of researchers.
19
II. Study designs
Classification of Study design
- Observational studies
the investigator simply observes the
natural course of an event
the investigator measures but does not
intervene.
- Interventional studies
the investigator assigns study subjects
to exposure and non-exposure, then
follows to measure for disease
occurrence.
the investigator manipulates the
intervention or exposure.
Algorithm for classification of types study design
1. Observational study designs
time
Compare
Rates
Outcome present
Not exposed
to risk factor
Outcome absent
PRESENT FUTURE
Incidence
among
Exposed exposed
(8/15)
Incidence
among
Unexposed
un-exposed
(5/15)
Follow-up period
Retrospective cohort studies
Outcome
Exposure occurrence Study starts
time
Retrospective assessment of exposure and outcome
Baseline measurement
Random allocation
• Loss to follow-up
START
• Non-participation
A. Non-probability sampling
1. Convenience sampling
2. Quota sampling
3. Purposive sampling
4. Snowball sampling
1. Convenience Sampling:
For convenience, the study units that are available at
the time of data collection are selected
Many clinic-based studies
2. Quota sampling:
- is done until a specific number of units (quotas) for
different categories of populations have been selected.
3. Purposive sampling
The investigator assumes that they are typical of the
study population
People who are assumed to provide rich information
are selected
4. Snowball sampling
dispersed
2. Systematic sampling
two stages.
Zone PSU
Woreda SSU
Kebele TSU
Sub-Kebele
HH
Sample Size Calculations
Common questions:
“How many subjects should I study?”
Too small sample = Waste of time and resources
PRECISION COST
∆
d2
For population <10,000, use finite population
correction
nf= ni
1+ni/N
Single proportion…cont.
Sampling technique
In complex samples (cluster, multistage) increase the sample
size to account for design effect
Suppose that you are interested to know the
proportion of infants who breastfed >18 months
of age in a rural area. Suppose that in a similar
area, the proportion (p) of breastfed infants was
found to be 0.20. What sample size is required to
estimate the true proportion within ±3% with 95%
confidence. Let p=0.20, d=0.03, α=5%
Suppose there is no prior information about the
proportion (p) who breastfeed
Assume p=q=0.5 (most conservative)
Then the required sample size increases
Significance Difference
Between Two Groups
Using power of a study to determine sample
size = significant difference = Hypothesis
testing
Action Reality
(Conclusion)
Ho is true Ho is false
93
In simpler terms, the researcher has detected a
difference between the groups that in actuality
does not exist; the difference between the
groups occurred by chance (or error).
94
In simpler terms, the researcher has failed to
detect a difference that actually exists between the
groups.
Which type of error is more serious—Type I or
Type II? The answer to this question often
depends on the context in which the errors are
made.
95
Type II error(false negative)
•Concluding there is no difference between
the groups being studied when, in fact,
there is a difference.
•There is an inverse relationship between
Type I and Type II errors, which means that
by decreasing the probability of making a
Type I error, the researcher is increasing
the probability of making a Type II error.
96
Power (1 - β) = 50%, Zβ = 0.00
Power (1 - β) = 75%, Zβ = 0.67
Power (1 - β) = 80%, Zβ = 0.84
Power (1 – β) = 90%, Zβ = 1.28
Power is one-sided and Zβ is always
one-sided
Most studies are recommend at power
of 80%.
General formula using power of a study to determine sample size
(Comparative CS, CC, Cohort) proportions
∆ = p1-p2
Let p1=0.35, p2=0.25, and Δ=p1-p2
=0.35-0.25 =0.10
Sample size calculations shown above are
based on SRS
Stratification will increase precision
compared with others the same size,
smaller sample can be selected
Cluster sampling will reduce precision
compared to SRS of the same size, larger
sample will be needed (design effect)
Research Errors
Observations are made on a group of people, and
inferences are made about r/s or associations of
various ‘exposures’ to ‘outcomes’.
The inferences reached are always subject to
uncertainty due to the variation of characteristics
across the population.
The accuracy of the inference depends on:-
the accuracy of the information collected and
the representativeness of the subjects observed to the
larger group of subjects in the population,
as well as on the accuracy of the statistical methods
used to draw the inference.
101
Common sources of error that need to be
controlled result from problems with ‘reliability’
and ‘validity’.
Our inference should have high reliability and
high validity.
The reliability and validity of inferences depend
on the
reliability and validity of the measurements (are we
measuring the right thing, and with accuracy?)
reliability and validity of the samples chosen (have
we got a true representation of the population that
we are drawing inferences from?).
102
The reliability of a sample is achieved by
selecting a large sample, and the validity is
achieved by ensuring the sample selection is
unbiased.
The researcher evaluate whether study results
are valid by assessing the presence of three
alternative explanations:-
1. Bias
2. Confounding
3. Chance/random error
103
Bias
categories:
104
A. Selection(inclusion/ exclusion) Bias
any error that arises in the process of identifying
the study populations.
can occur whenever the identification of
individual subjects for inclusion in the study on
the basis of either exposure (cohort) or disease
(case-control) status
Non-response
Loss to follow-up
Volunteer / Compliance bias
105
B. Observation or information bias
any systematic error in the
measurement of information on exposure
or outcome. Include:
Interviewer bias:
Recall bias
Social desirability bias-
106
Recommendations to minimize bias at the time of data
collection & study design are:
108
Confounding
109
R/s b/n low level of mother’s education and
Ex.
malnutrition in under five children
Control for Confounding Variables
a) Randomization :-means that subjects are
assigned to the study groups through a
random process.
It generally ensures that compared groups
are balanced on both known and unknown
confounders, if the sample size is
sufficient.
111
b) Restriction:- means that admissibility
criteria for enrollment are limited.
EX.: Limiting the study population to one
gender or only one race.
112
EX.:- Consider a cohort study on the relationship
b/n exercise and the risk of colon cancer.
A literature review reveals that potential
confounders of this association include age,
gender, and obesity.
Thus, when an exposed subject (an exerciser) is
enrolled who is a 55-year-old male with a normal
body mass index; an unexposed subject (a non
exerciser) would be enrolled who is a 55-year-old
male with a normal body mass index.
113
Chance/Random Error
Unlike bias & confounding random errors
are considered unsystematic because they
arise from unpredictable process.
Evaluation of the role of chance is mainly
the domain of statistics and it involves:-
Hypothesis Testing
Confidence interval estimation
114
Chance… cont’d
1. Hypothesis Testing (test of statistical
significance)
When we test a hypothesis, we typically want to
disprove a null hypothesis.
quantifies the degree to which sampling variability
may account for the observed results using P-
value
For medical research, the P value 0.05 is set
conventionally to indicate statistical significant.
115
Steps in testing for statistical significance
116
3. Calculate test statistics
The test statistic is a measure of the
difference b/n population parameter
proposed by HO & its actual sample
estimate, expressed in terms of standard
deviation
In statistics, a result is called significant if
it is unlikely to have occurred by chance
117
•The P value is the probability that a difference
as large as we have observed could have
occurred simply by chance alone, assuming that
there is truly no association b/n exposure &
outcome under consideration (i.e., H0 is true).
118
4. Evaluate the evidence against HO
The test statistic is compared with its
frequency distribution to determine the
probability of obtaining this value by random
chance when HO is true (i.e the p-value)
The level of significance () is used to derive
a decision rule, a procedure defining the
conditions under which HO is accepted or
rejected
119
5. State the conclusion: HO is accepted
or rejected in accordance with the
decision rule
If HO is rejected, the results are said to
be statistically significant ( i.e unlikely to
be due to chance) at chosen .
If HO is not rejected, the results are
reported as not statistically significant
( i.e likely to be due to chance) at
chosen
120
2. Estimation of Confidence Interval
The confidence interval represents the range
within which the true magnitude of effect lies
within a certain degree of assurance.
It is more informative than just P value
because it reflects on both the size of the
sample and the magnitude of the effect.
Components of Interval estimation:-
Estimator
Reliability coefficient
Standard error
Estimator + (reliability coefficient )x (standard error)
121
Confidence Intervals
Used for estimating the true value of the
population parameter
122
Lower limit = Point Estimate - (Critical Value)
x (Standard Error of Estimate)
123
CI tells us how precise our estimate is likely
to be
124
Example
A prevalence of 10% from a sample size of 20
would have a 95% CI of 3% to 23%,
which is not very precise or informative.
But, a prevalence of 10% from a sample of size
400
would have a 95% CI of 7% to 13%,
which may be considered sufficiently
accurate.
125
95% CI commonly used
Sometimes 90% and 99%
The 95% CI is calculated in such a way
that, under the conditions assumed for
underlying distribution, the interval will
contain true population parameter 95% of
the time.
Loosely speaking, you might interpret a
95% CI as one which you are 95%
confident contains the true parameter.
126
CIs can also answer the question of whether
or not an association exists or a treatment is
beneficial or harmful. (analogous to
p-values…)
e.g., if the CI of an odds ratio includes the
value 1.0 we cannot be confident that exposure
is associated with disease.
127
C.I. for a population proportion
128
Hence,
129
Example:
A study on dental health practice. Of 300
adults interviewed, 123 said that they
regularly had a dental check-up twice a
year. What is the 95% C.I. for the true
proportion in the population?
(0.36, 0.46).
130
IV. Data collection methods
Data collection techniques – allows us to
systematically collect information.
questionnaire
Written questions are presented to be
One-to-one
confidential topics.
Interviewee can be either “key informants” or
“ordinary people”.
C. Key informant interviews
….)
Stage 2: Data collection
Selection of candidates
Training
dropouts
RESEARCH INSTRUMENT/ QUESTIONNAIRE DESIGN
The quality of research depends to a large
extent on the quality of the data collection
tools.
Research instruments include
Questionnaire, Checklist
information on:
facts with which the researcher is not very familiar,
sensitive issues.
For example:
‘What is your opinion on the services provided in the
ANC?’ (Explain why.)
2. Close-ended questions:
have a list of possible options or answers from which the
respondents must choose.
the respondents express their opinions or attitudes by
choosing rating points on a scale.
EX: What is your opinion on the following
statement: ‘Women who have induced abortion
should be severely punished.’
1. strongly agree
2. agree
3. no opinion
4. disagree
5. strongly disagree
STEPS IN DESIGNING A QUESTIONNAIRE
Step 1:-Content
- decide which question need to measure variables & to
Attitude/opinion/belief questions
V. DATA
PROCESSING
AND
ANALYSIS
151
1.Introduction
Data processing:-
Transforming or process the data
into a format suitable for analysis.
152
When it has to be started?
Should start in the field, with checking for
completeness of the data and performing quality
control checks.
Data of small samples may even be processed
and analyzed as soon as it is collected.
153
What should it include?
Data processing and analysis include:
Sorting data
Performing quality-control checks
Data processing, and
Data analysis.
154
I.SORTING DATA
An appropriate system of arranging, organizing/
the data to facilitating subsequent processing and
analysis.
If you have different study populations (for
example village health workers, village health
committees and the general population), you
obviously would number the questionnaires
separately.
Eg. Case - control studies.
/Users vs nonusers Fp service/
155
II. PERFORMING QUALITY CONTROL CHECKS
Most important data checks,
1.Accuracy of data
2.Checking outliers
3. Handling missed values
Accuracy
Verify the accuracy of sample of completed data.
Entered twice & compare
156
Performance quality…..
Checking outliers
Outliers:-values that are considerably
large/small, impossible/inconsistent with
information gathered.
Occur:-error on measuring , recording
157
Performance quality…..
Handling missed values
If an inconsistency is clearly due to a mistake
made by the interviewer on recording , it may
still be possible to check with the her/him and
to correct the answer.
If the inconsistency is less clearly a mistake in
recording, it may be possible (in a small scale
study) to return to the respondent and ask for
clarification.
158
Performance quality…..
Note:
A decision to exclude data of doubtful quality
is ethically correct and it testifies to the
scientific integrity of the researcher. You
should keep track of any questions you had to
exclude because of incompleteness or
inconsistency in the answers, and discuss it in
your final report.
159
III. DATA PROCESSING
Data processing in involves:
categorising the data,
coding, and
summarising the data
manual:-data master sheets & without
master sheets,
by computer :-data entry and
verification.
160
DATA PROCESSING…..
1. Categorising
categorical & numerical variables
Decisions have to be made concerning
how to categorise responses.
For categorical variables that are
investigated through closed questions or
observation (for example, observation of
the presence or absence of latrines in
household), the categories have been
decided upon beforehand.
161
Data processing…..
The numerical variables are often better
collected without any pre-categorization.
If your categories had been wrongly chosen
you cannot reclassify the data anymore.
162
Data processing…..
2. Coding
is assigning a separate (non-overlapping)
numerical, letter code for separate answers
and missing values.
163
Data processing…..
Coding……
Grouped as:
1. Pre-coding ;-when questionnaire being written so
that coding categories & column location appear in
the questionnaire itself.
2. Post -coding:-…. After respondents have answered
questions.
necessary mainly for open ended questions for
which response categories can't anticipate.
3. Recoding:-changes (including adding, combining &
removing) earlier coding to facilitate meaningful
analysis.
done after data have been collected.
164
Missing values
Missing values occur when measurements were
not taken, or respondents did not answer questions
so forth.
165
Code book preparation
Code book is essentially a list of each
variables, the column in which the
variable has been entered, and the codes
associated with the value of the variables
166
Examples of code book
168
3. Summarising the data using:
169
For example, the straightforward answers of the smoking
questionnaire for male smokers could be processed as follows
Table 13.1: Master sheet for smokers (males)
170
Data master sheets…..
171
172
(2) Compilation by hand (without using
master sheets)
When the sample is small (say less than
30) and the collected data is limited, it
might be more efficient to do the
compilation manually.
Certain procedures will help ensure
accuracy and speed.
173
3/Computer compilation
Before you decide to use a computer, you
have to be sure that:
It will save time or that the quality of the
analysis will benefit from it.
174
Computer compilation……
Computer compilation consists of the
following steps:-
I. Choosing an appropriate computer
program
II. Data entry
III. Verification or validation of the data
IV. Programming (if necessary)
V. Computer outputs/prints
175
Computer compilation……
Choosing an appropriate computer
program
The most widely used programs are:
Epi Info
SPSS, which is a quite advanced Statistical
Package for Social Sciences (SPSS Inc.).
176
Statistical Methods for
Analysis of Categorical Data
177
Categorical Variables
Non-numerical
– Sex: Male, Female
– Marital Status: Married, single, widowed, divorced
– Yes/No
Can be nominal or ordered
Numeric (continuous) variables can be condensed to
categorical variables
Analysis is based on the frequencies associated with
each category of the variable
178
Categorical Variables analysis…
1. Chi-squared Test
Widely used in the analysis of contingency
tables.
Allows us to test for association between
categorical variables.
The Ho for this test is there is no
association between the variables.
Consequently a significant p-value implies
association.
179
180
TEST STATISTIC:
nad bc
2
2
(a c)(b d )( a b)(c d )
181
Test Statistic: 2-test with d.f. = (r-1)x(c-1)
O E ij
2
2 ij
i, j E ij
i raw total j column tot al Ri C j
th th
Eij
grand total n
Oij=observed frequency, Eij=expected frequency of the cell at the
juncture of I th raw & j th column
182
Assumptions of the 2 - test
185
P‐Value
• When P is below the cut off level(α), say 0.05, the
result is called statistically significant( and below
0.01, it may be called highly significant); when
above 0.05 it is called not significant.
• It is important to distinguish between the significance
level and the p – value.
• The significance level α is the probability of making a
type I error. This is set before the test is carried out.
• The P – value is the result observed after the study is
completed and is based on the observed data
186
Another way to state conclusion:
Reject H0 if P-value < ,
Accept H0 if P-value .
187
2. Measures of Association
The chi-squared test tells us whether or
not an association exists.
It is useful, however, to know how strong
the association is.
The relative risk and the odds ratio are
appropriate measures of the strength of
the association.
188
Relative Risk
Consider the 2 X 2 contingency table in
which the outcome is the presence or
absence of disease and the other variable
is the presence or absence of a factor.
189
Relative Risk
2X2 Table
Disease
Factor
Present Absent
Present p1 1-p1
Absent p2 1-p2
190
Suppose the probability of getting the disease
when the factor is present p1 and the
probability of getting the disease when the
factor is absent is p2.
191
Relative risk (RR) is simply the risk of
getting the disease when the factor is present
relative to the risk of getting the disease
when the factor is absent.
192
Cohort/Experimental Design
ILL NOT ILL
Incidence
a
Exposed a b a+b
a+b
Not exposed c c
d c+d
c+d
193
Cohort study
1st Give Breast Cancer
Birth Yes No Total
Total= a/a+b
RR 96 6072 6168
c/c+d
194
a/a+b = 31/1628 = 0.019
b/b+d = 65/4540 = 0.014
195
In general: RR > 3 indicates strong
association
RR = 1.5 – 2.9 moderate
RR = 1.1 – 1.4 weak
196
Odds Ratio
Another commonly used measure of
association is the odds ratio (OR).
The odds in favor of an event happening
(such as getting a disease) is the probability of
the event happening relative to the probability
of the event not happening.
OR: Indicates odds of diseased relative to the
exposure status.
Odds of disease in exposed = a/b or a:b
Odds of disease in unexposed = c/d or c:d
197
OR = odds in favor of disease when the factor is present
odds in favor of disease when the factor is absent
OR: is the chance of being exposed (or diseased) as opposed
to not being exposed (or diseased).
It is possible to calculate either exposure or disease odds
ratio, which are exactly the same
198
Case-Control/Cross-sectional Designs
Cases Controls
Exposed
a
b
Not exposed c d
Total a+c b+d
10/2/2011 262
200
The odds ratio is estimated by:
201
When the CI for the RR or OR
includes 1, we conclude that there is
no association
202
VI. HEALTH RESEARCH
FINDING
COMMUNICATION
203 Prepared by Taye Gari(Public health Lecturer) Saturday, July 13, 2024
Definition
Process of sharing information and
knowledge about research findings to
all relevant parties.
Why?
Research is not complete unless communicated
Accessibility
Feedback mechanism
Communication, Dissemination & Utilization of Research
Findings
Poor Dissemination,
Communication &
Utilisation of research findings
Big Gap!
findings
Detailed report (as in thesis or consultancy work)
Scientific paper for publication in a journal
A conference abstract
Present research findings orally
Poster presentation
211 Prepared by Taye Gari(Public health Lecturer) Saturday, July 13, 2024
Sections of a Research Report
Title
Acknowledgment
Abstract
Introduction including problem statement
Objective(general & specific)
Method & Material
Results
Discussion
Conclusion
Recommendations
Limitations
References or bibliography
1. Titles
Title should accurately reflect the content of the
paper.
Purpose
to provide a brief, informative summary that will attract the
target audience
The attractiveness of a report or a scientific paper begins
from its title
It is the basis for the initial impression and is powerful in
encouraging or discouraging reading.
How should a title be constructed?
Guidelines for producing a good title
The simpler the title, the better
Consider the target readership
Be brief – short titles are clearer and more
arresting
Avoid excessive adjectives and noun strings
Titles should almost never contain
abbreviations
Content of the title page
1. The Title
2. Name of Investigator/s/
3. Type of the research report
4. Advisor name
5. Year and city and country name
2. Acknowledgements
It is important to acknowledge the source of research
funding.
proceed.
3. The Abstract
Purpose
to highlight key points from the major sections of the article
It is a brief summary of the main points of the larger work
Second most read part of a paper
Could be the basis on which the work is judged.
Should be short, clear, informative and interesting
Should contain the whole essence of the paper
Should stand alone
Component of Abstract
Basic parts:-
Major purpose of the study (Why what was done was done) -
Introduction
Goal (objective)
Basic procedures (What was done) – Methods
undertaken.
Should be
Short
Impressive
Clearly tell the reader why you have undertaken the study
Important points to write an effective introduction
You must know your audience
Keep it short
Explain why you have done the study
Explain why it is important and how it is different from
previous work
Be convincing in describing that it adds value to what has
been done before
Try to make your first sentence impressive and catchy.
5. The Methods
Purpose
Respect chronology
6. The Results
Purpose
To describe the results of data analysis that
are relevant to the study purpose
Gives details of the methods that will be used and why they
are appropriate
To identify the major controversies that are important for your project
Why literature review?
It helps you find out what others have learned &
reported on your topic, & take account of this in the
design of your study.
“SMART”
Specific
Measurable
Achievable
Realistic
Time bound
How should objectives be stated?
1. General objectives
Is closely related to the statement of the
problem
Some times called purpose of the
research
Identifies in general terms what is to be
accomplished by the research and why
Types of research objectives
2. Specific objectives
smaller portions of the general objectives.
identify in greater detail the specific aims of the
research project, often breaking down what is to
be accomplished into smaller logical
components.
Identify questions that the researcher wants to
answer
specify what to do in a study, where & for what
purpose
Research Methods
Methods
What is variable?
Is a characteristic of a person, object or phenomenon,
which can be measured and take on different values
Example:
Age of a person
Weight (expressed in kilograms or in pounds);
Home - clinic distance (in kms or in minutes walking
distance);
Monthly income (expressed in Birr, or Dollars);
Number of children (1, 2, etc.)
Variables
Numerical Variables
Variables where it’s values expressed in numbers
The variable ‘age’ can take on different values since a
person can be 20 years old, 35 years old and so on
Categorical Variables
Variables whose values expressed in categories or
groups,
Example: The variable sex has 2 districts groups, male
or female.
Variables
If too many variables are included for the study the
research may become difficult.
Research methods or techniques
The type of study design chosen
depends on
The type of problem
What is sampling?
Sampling is the process involving the selection of a
finite number of elements from a given population of
interest, for purposes of inquiry
What is a sample
In research it is not always possible to study an entire
population.
A decision is often made, therefore, to study only a
small fraction of the population, or a “sample” of it,
from which conclusions can be drawn about the whole
population.
A sample should possess certain characteristics
A sample is a representative part of a population
What are the characteristics a sample should
possess?
Waste resource
Decreases precision
of the estimate
Narrows the range of
conclusions and
generalizations
Planning for data collection
A plan for data collection can be made in 2 steps:
Step 1
Listing the tasks that have to be carried out and who should be
involved
Make rough estimate of the time needed for the different part of
the study
identify the most appropriate period in which to carry out the
research
Step 2:
Scheduling the different activities that have to be carried out in
a work plan
Stage in data collection planning process
312
THE END
Thank You