This course has been designed to equip the student with the basic research design including research methods in health sciences. The course aims to impart basic knowledge on different types of study design
1. Unit 3: Research design
Ashok Pandey
Associate Research Fellow
PRI
2. Basic Fundamental Steps
Research Ideas
Research Questions
Background/Significance
Study Objectives & Hypotheses
Study design
Study Sample
Data to be collected
Analysis Strategies
Data Collection Methods
Ashok Pandey (MPH/BPH, DGH) 2
3. Research design
• Observational study- type design, example
• Interventional study- type design, example
Qualitative research, Meta analysis, small topic
Foundation of qualitative and quantitative research
design
Identify different types of study design including
observational, pre-experimental and experimental
designs and their inherent threats their internal
and external validity
Ashok Pandey (MPH/BPH, DGH) 3
4. Study Design
• A study design is a specific plan or protocol
• for conducting the study,
• which allows the investigator
• to translate the conceptual hypothesis
• into an operational one.
• Methodology of planning and Collection of Data to fulfill the
aims of Research
• In most of the Research observations are made on a few units
and findings from ‘a few’ are generalized to a ‘large group’
4
Ashok Pandey (MPH/BPH, DGH)
5. Meaning of research design
A research design is the arrangement of
conditions for collection and analysis of data
in a manner that aims to combine relevance
to the research purpose with economy in
procedure.
Ashok Pandey (MPH/BPH, DGH) 5
6. Research design have
following parts
Sampling design
Observational design
Statistical design
Operational design
Ashok Pandey (MPH/BPH, DGH) 6
7. Depends on
State of Knowledge of the Problem
Type of Research Questions
Ashok Pandey (MPH/BPH, DGH) 7
8. Research Questions
1. What is the nature / magnitude problem?
2. Who is most effected?
3. How do the effected people behave?
4. What is the effective measure to reduce health damage?
5. What are the pollutants that cause air/water/noise pollution?
6. What are the sources of Indoor pollution in rural Nepal?
7. What is the major cause of poor respiratory health?
8. Are certain factors associated with the problem?
9. Will the removal of particular factors prevent and reduce the
problem?
10. What is the effect of a particular strategy / intervention?
11. Which of two alternate strategies give better results or cost
effective?
Ashok Pandey (MPH/BPH, DGH) 8
10. Decision Tree
Observational Study
Intervention Done
No Yes
Experimental Study
Comparison Group
Descriptive Study Analytic Study
Cohort Study
Cross-Sectional Study
Case-Control Study
Randomization
NRCT Study RCT Study
Direction of Study
E O
No
No
Yes
Yes
E O
E = O
10
Ashok Pandey (MPH/BPH, DGH)
11. Epidemiological Study Design
Observational Studies
Descriptive Studies
Analytic
Cross-Sectional
Case-Control
Cohort
Experimental / Interventional studies
As per Control: RCT/NRCT
As per Blinding: Single /Double Blind
As per Design: Simple/Cross-over
As per Area: Field/Clinical/Lab
11
Ashok Pandey (MPH/BPH, DGH)
15. Descriptive Type of Observational Study
• Other Name Case-Series/Population
• Unit of Study Case/Individuals
• Study Question What is happening
• Direction Of Inquiry
• Study Design
☻☻☻☻☻☻ desired information
☻☻☻☻☻☻ about cases/individuals is collected
15
Ashok Pandey (MPH/BPH, DGH)
16. Case-Series
Advantages
• Easy to do
• Excellent at identifying unusual situation
• Good for generating hypotheses
Disadvantages
• Generally short-term
• Investigators self-select (bias!)
• no controls
16
Ashok Pandey (MPH/BPH, DGH)
18. Cross-sectional Study
• Data collected at a single point in time
• Describes associations
• Prevalence
A “Snapshot”
18
Ashok Pandey (MPH/BPH, DGH)
19. Cross-Sectional Study
• Other Name Prevalence Study
• Unit of Study Individual
• Study Question What is happening
• Direction of Inquiry
• Study Design
Population
Diseased
Not
Exposed
to Factor
Exposed
to Factor
Non-
Disease
Exposed to
Factor
Not
Exposed to
Factor 19
Ashok Pandey (MPH/BPH, DGH)
21. Case-Control Studies
Start with people who have disease(Cases)
Match them with controls that do not have
disease (Match Confounding)
Look back and assess exposures
21
Ashok Pandey (MPH/BPH, DGH)
22. Controls
A control is a standard of comparison
(confounded with variability but without
effect)
for
• Effects
• Variability
22
Ashok Pandey (MPH/BPH, DGH)
23. Case-Control Study
• Other Name Retrospective Study
• Unit of Study Cases/Control
• Study Question What has happened
• Direction of Inquiry=
• Study Design
Cases
Not
Exposed
Exposed
Control
Exposed
Not
Exposed
23
Ashok Pandey (MPH/BPH, DGH)
24. Objective of a Case-Control Study
To find out association
To assess Risk Ratio
24
Ashok Pandey (MPH/BPH, DGH)
25. Case-Control Studies: Strengths
• Good for rare outcomes: cancer
• Can examine relation of exposures to disease
• Useful to generate hypothesis
• Fast
• Cheap
• Provides Odds Ratio
25
Ashok Pandey (MPH/BPH, DGH)
26. Case-Control Studies: Weaknesses
• Cannot measure
–Incidence
–Prevalence
–Relative Risk
• Can only study one outcome
• High susceptibility to bias
26
Ashok Pandey (MPH/BPH, DGH)
27. Nested case-control studies
• Nested case-control studies are not very
different from classic case-control studies.
• They have a special name because of the way
they are conducted.
• A nested case-control study is a case-control
study situated within a prospective cohort
study.
Ashok Pandey (MPH/BPH, DGH) 27
29. Example of Nested Case Control Study
•A nested case control study examined the
relationship between serum organochlorides
and breast caner.
•Study subject were drawn from a cohort of
over 57,000 female members of Kaiser
Permanente Medical Care Program who went
multiphasic examination in late 1960s, at which
time blood samples were collected and stored.
•The cohort was followed upto 1990.
•150 women who developed breast cancer during the
followup period were then randomly selected and
individually matched to 150 women in the cohort
who had remained free of breast cancer.
Ashok Pandey (MPH/BPH, DGH) 29
30. Cohort Study
• Begin with disease-free individuals
• Classify patients as exposed/unexposed
• Record outcomes in both groups
• Compare outcomes using relative risk
30
Ashok Pandey (MPH/BPH, DGH)
31. Cohort Study
• Other Name Prospective Study / Follow-up Study/Incidence Study
• Unit of Study Individual
• Study Question What is happening
• Direction of Inquiry
• Study Design
•
Cohort
Cohort
Exposed to
Factor
Not Non
Diseased
Not
Exposed to
Factor
Diseased
Diseased
Non-Diseased
31
Ashok Pandey (MPH/BPH, DGH)
32. Cohort Study: Strengths
• Can measure multiple outcomes
• Can adjust for confounding variables
• Can calculate Attributed Risk
32
Ashok Pandey (MPH/BPH, DGH)
34. Measurements of Association
•Significance Test
•Relative Risk
•Attributable Risk
•OR
•Significance Test
•OR
Cohort Study Case Control Study
34
Ashok Pandey (MPH/BPH, DGH)
35. Measures of Association
Significance Test – to test significance of difference
in exposure between control and Cases
Odds ratio - ratio of the odds of contracting
disease in given exposure
Relative Risk – Ratio between incidence among
exposed and incidence among non-exposed
Attributed Risk – percentage of difference between
incidence among exposed and non-exposed with
incidence among exposed
RR or OR of 1 indicate no effect of exposure (equal odds)
35
Ashok Pandey (MPH/BPH, DGH)
36. Experimental Studies
Clinical trials provide the “gold standard”
of determining the relationship between
factor and the event
36
Ashok Pandey (MPH/BPH, DGH)
37. Types of Experimental Study
As per Randomization:
• Randomized Control Trials (RCT)
• Concurrent Parallel Design (RCT)
• Sequential RCT Design
• RCT with External Control
• Non – Randomized Trials
37
Ashok Pandey (MPH/BPH, DGH)
38. Randomized Control Trials (RCT)
• Before and After Comparison
• Comparison with Placebo
• Comparison Of two medicine/procedure/tests
• Comparison Of > two medicine/procedure/tests
38
Ashok Pandey (MPH/BPH, DGH)
39. Experimental Study
• Other Name Intervention Study
• Objective To know the effect of intervention
• Unit of Study Individual meeting entry criteria
• Study Question What is happening after intervention in both
groups
• Direction of Inquiry I E
• Study Design 1(Intervention with Placebo)
Group 1/cases Intervention
Negative
Outcome
Positive
Outcome
Group
2/control
Placebo
Positive
Outcome
Negative
Outcome
39
Ashok Pandey (MPH/BPH, DGH)
40. Experimental (Interventional)
To determine whether one or more variables (e.g. a
program or treatment variable) causes or affects one or
more outcome variables
Test the hypothesis
To provide "scientific proof”
To test the effectiveness and efficiency of on-going / new
health services / programs for improving the health of the
community
To study the efficacy of drugs/vaccines for the treatment
and prevention of diseases or health problems
Ashok Pandey (MPH/BPH, DGH) 40
42. Pre-Experimental Study
Types of Pre-experimental Study
• Before and After Study
• After Only Study / Post Test Only Study
• Pre-test Post-test Design
– One group pre-test post-test design
– Two groups pre-test post-test design
– Etc.
Ashok Pandey (MPH/BPH, DGH) 42
43. Before-After study
– Compare same subjects before and after intervention
– No randomization
– No control group
Ashok Pandey (MPH/BPH, DGH) 43
44. After only study
Before
Study Pop.
(Existing
Records)
After Study
Population
Compare
Intervention
Ashok Pandey (MPH/BPH, DGH) 44
45. Pre-test and Post Test
• Study Population (All the staffs from Institution A)
• Sampling
• Study Group Assess the Research Knowledge
• (20 participants) (Pre Test)
• Manipulation (Intervention) Compare
• (Research Training Package)
• Study Group Assess the Research Knowledge
• (20 participants) (Post Test)
Ashok Pandey (MPH/BPH, DGH) 45
46. Quasi-experimental Study
– The treatment allocation is based on subject’s
preference.
– Treatment group or area
– Easier to get subjects enrolled in the study and to
obtain formal consent.
– Might be difficult to get appropriate control for
each eligible case.
Ashok Pandey (MPH/BPH, DGH) 46
47. Types of Quasi-experimental Study
• Control Group Design
(Non-equivalent and Equivalent)
(Single and Double)
– Pre-test Post-test with one Control Group Design
– Pre-test Post-test with two Control Groups Design
(Solomon four group designs – two experimental and two
control groups)
• Non Randomized Controlled Trial (NRCT)
• Historical Controlled Trial (HCT)
• Community Trial
• Field Trial Ashok Pandey (MPH/BPH, DGH) 47
49. Non Randomized Controlled Trial (NRCT)
– The treatment allocation is based on subject’s
preference.
– Treatment group or area
– Easier to get subjects enrolled in the study and to
obtain formal consent.
– Might be difficult to get appropriate control for
each eligible case.
Ashok Pandey (MPH/BPH, DGH) 49
50. Historical Controlled Trial (HCT)
– New intervention used in series of subjects and the
results are compared with the outcome of a series of
comparable subjects treated in the past.
– Controls were selected from the records or data bank.
Ashok Pandey (MPH/BPH, DGH) 50
51. Community Trial
– Community as a whole is studied
– Is useful when the outcome of interest is so
common that is difficult to identify a high-risk
group, e.g. Education Intervention Program
– Mostly directed towards the change of health
behaviour among community
Ashok Pandey (MPH/BPH, DGH) 51
52. Field Trial
– Intervention field and usually among general
population
– Before the Disease Occurrence
e.g. Vaccination Program
Ashok Pandey (MPH/BPH, DGH) 52
53. Matching
• Ensures that subjects in each group
– Are pretty equivalent on some characteristic or all
important confounds
– Should be related to the dependent measure
• Disadvantages
– Expensive and time consuming
– May not be possible
– Matching on some variables establishes
equivalence on others
Ashok Pandey (MPH/BPH, DGH) 53
55. True Experimental Design
• Groups
should be
equivalent
at beginning
So,
• Observed
differences
must result
from the
treatment
Ashok Pandey (MPH/BPH, DGH) 55
56. Masking (Blinding)
• Attempt to eliminate biases & preconceptions
• Single-blind
– Subject masking
– Use of placebo
• Double-blind
– Subject masking and researcher maskingllectors and data
analysts
• Triple-blind
– Subject masking, researcher masking and study sponsor
masking
Ashok Pandey (MPH/BPH, DGH) 56
57. Experimental Designs (summing up)
Pre-Exp Quasi-Exp True Exp
Presence of a control
group?
In some cases,
but usually not
Often Always
Random selection of
subjects from a
population?
No No Yes
Random assignment of
subjects to groups?
No No Yes
Random assignment of
treatments to groups?
No No Yes
Degree of control over
extraneous variables?
None Some Yes
Ashok Pandey (MPH/BPH, DGH) 57
60. What isa Clinical trial?
• prospective ethically designed investigation in It
human subjects to objectively
discover/verify/compare the results of two or more
therapeutic measures(drugs)
Aclinical trial is aplanned experiment that involves
volunteers/patients
Aim to compare the response to new treatment with
that of an existing one orplacebo
Clinical trial is just a part of New Drug Discovery
Process.
•
•
•
Ashok Pandey (MPH/BPH, DGH) 60
61. Why are clinical trialsimportant?
• Clinical trials translate results of basic
scientific research into better ways to
prevent, diagnose,or treatdisease
• Themore people take part, the fasterwe can:
-Answer critical researchquestions
-Findbetter treatments andwaysto preventdisease
Ashok Pandey (MPH/BPH, DGH) 61
63. RandomizedClinicalTrials
• Subjectshaveequal chanceto be assignedto one of two or
more groupsjust like tossingof coin.
– One group gets the most widely accepted treatment
(standardtreatment)
– Theother getsthe new treatment beingtested
– All groupsare asalike aspossible; removesthe probability of bias.
Ashok Pandey (MPH/BPH, DGH) 63
65. Openlabel trial Blindedclinicaltrial
Doctor and patient know
which drug isgiven
Single Blind: the patient
doesn’t know whichtreatment
he/she isgetting
Double Blind: neither doctor
nor patient knows
Ashok Pandey (MPH/BPH, DGH) 65
66. Prospective Retrospective
Patients are enrolled for
the study before any
treatment begins
Progressof patients is
monitored during course
of treatment
Pastcaserecords & other
statistical data are usedfor
analysis
Investigator has to rely on
methods employed & data
available.
Ashok Pandey (MPH/BPH, DGH) 66
67. Placebostudy
• Placebo
-It is an inert medicament given in thegarbof medicine.
-Itresemblesthe active drug in physicalproperties but
doesnot haveanypharmacologicalactivity.
• Thenew treatment istested against aplacebo.
Ashok Pandey (MPH/BPH, DGH) 67
68. PilotStudy
• Asmall study that helpsto develop abiggerstudy.
Advantage:
• to find outpossibledifficulties
• to help with designof the bigger,more pivotal study.
Ashok Pandey (MPH/BPH, DGH) 68
69. Cross-overstudy
• Two types of treatment are studied in the same
group.
• Before giving 1sttreatment baseline observations are made
for certainperiod – “Run-in period”.
• When one treatment is over, before starting 2nd
treatment some time is allowed for the effect of
treatment to completely wash out – “Wash-out period”.
Ashok Pandey (MPH/BPH, DGH) 69
70. • Advantages
• No.of subjects required isless.
• Minimizes chancesof subject
variation.
Crossover design
Standard Placebo T
est
Placebo T
est Standard
T
est Standard Placebo
* Awash out period of aweek between two weeks of
therapy
Group Week 1 Week 2 Week 3
1 Standard Placebo Test
2 Placebo Test Standarad
3 Test Standard placebo
Ashok Pandey (MPH/BPH, DGH) 70
71. PhaseI
• 25-100
• Healthy volunteers; exception are cytotoxic
drug and antiretroviral drug.
• Doneby trained clinicalpharmacologist
• Nonblinded or openlabeled
Ashok Pandey (MPH/BPH, DGH) 71
79. Qualitative vs Quantitative
Qualitative
Research
Quantitative
Research
Type of questions Probing Limited probing
Sample Size small large
Info. Per
respondent
much varies
Admin Requires skilled
researcher
Fewer specialist
skills required
Type of Analysis Subjective,
interpretative
Statistical
Type of research Exploratory Descriptive or
causal
Ashok Pandey (MPH/BPH, DGH) 79
80. Qualitative Research Quantitative Research
Researcher may only know roughly in
advance what he/she is looking for
Researcher knows clearly in advance what
he/she is looking for
The design emerges as the study unfolds. All aspects of the study are carefully
designed before data is collected.
Researcher is the data gathering
instrument
Researcher uses tools, such as
questionnaires or equipment to collect
numerical data.
Data is in the form of words, pictures or
objects
Data is in the form of numbers and
statistics
Subjective - interpretation of events is
important ,e.g., uses participant
observation, in-depth interviews etc.
Objective - seeks precise measurement &
analysis of target concepts, e.g., uses
surveys, questionnaires etc.
Qualitative data is more 'rich', time
consuming, and less able to be
generalized.
Quantitative data is more efficient, able to
test hypotheses, but may miss contextual
detail.
Ashok Pandey (MPH/BPH, DGH) 80
81. What is meta analysis?
Quantitative approach for systematically
combining results of previous research to
arrive at conclusions about the body of
research.
81
Ashok Pandey (MPH/BPH, DGH)
A systematic review of literature to address the
question:
on the basis of the research to date, how big is a
given effect
82. Four Steps of Meta Analysis
• Identify your studies
• Determine eligibility of studies
– Inclusion: which ones to keep
– Exclusion: which ones to throw out
• Abstract Data from the studies
• Analyze data in the studies statistically
82
Ashok Pandey (MPH/BPH, DGH)
83. Forest Plot
The dotted line passes
across null hypothesis that
OR=1.0
The Risk Estimate of each
study is lined up on each
side of the dotted line, with
95% CI spread as the line
The diamond below is the
summary estimate
The two ends of the
diamond indicate 95% CI
It is called as a forest plot so that we don’t miss the wood for the trees!
The size of the black square box indicates weight of the study
83
Ashok Pandey (MPH/BPH, DGH)
84. Time Series Design
Collect data on the same variable at
regular intervals (weeks, months, years,
etc.).
Data often is an aggregrate measure of
a population, e.g., graduation rates,
free/reduced lunches, consumer price
index, etc.
Ashok Pandey (MPH/BPH, DGH) 84
85. Time Series Design
O1 O2 O3 O4 O5 X1 O6 O7 O8 O9
“The essence of the time-series design is the
presence of a periodic measurement process on
some group or individual and the introduction of
an experimental change into this time series of
measurements, the results of which are indicated
by a discontinuity in the measurements recorded
in the time series” .
— Campbell & Stanley (1963)
Ashok Pandey (MPH/BPH, DGH) 85
86. Time Series Design
Time series designs useful for:
•Establishing a baseline measure
•Describing changes over time
•Keeping track of trends
•Forecasting future short term trends
Ashok Pandey (MPH/BPH, DGH) 86
88. As the basis of Time series Analysis businessman can
predict about the changes in economy. There are
following points which clear about the its importance:
1. Profit of experience.
2. Safety from future
3. Utility Studies
4. Sales Forecasting
6. Stock Market Analysis
8. Process and Quality Control
9. Inventory Studies
10. Economic Forecasting
5. Budgetary Analysis
7. Yield Projections
11. Risk Analysis & Evaluation of changes.
12. Census Analysis
Ashok Pandey (MPH/BPH, DGH) 88
Importanceof TimeSeriesAnalysis:-
89. The change which are being in time series, They are
effected by Economic, Social, Natural, Industrial &
Political Reasons. These reasons are called components
of Time Series.
SECULAR TREND :-
SEASONALVARIATION :-
CYCLICALVARIATION :-
IRREGULAR VARIATION :-
Ashok Pandey (MPH/BPH, DGH) 89
Components of Time Series:-
90. I = Irregular
Fluctuation
Time Series Model
• AdditionModel:
Y = T + S + C + I
Where:- Y = Original Data
T = Trend Value
S = Seasonal Fluctuation
C = Cyclical Fluctuation
• MultiplicationModel:
Y = T x S x C x I
or
Y = TSCI
Ashok Pandey (MPH/BPH, DGH) 90
91. Advantages
Data easy to collect
Results easy to present in graphs
Ease of interpretation (look for
patterns in graph)
Can forecast short term trends
Ashok Pandey (MPH/BPH, DGH) 91
92. Disadvantages
Data collection method may change
over time.
Difficult to show more than one
variable at a time.
Needs qualitative research to
explain fluctuations.
Assumes present trends will
continue unchanged.
Ashok Pandey (MPH/BPH, DGH) 92