Scientific Research
The Fountain of
Knowledge
Professor Syed Amin Tabish
FRCP (London), FRCP (Edin.), FAMS, MD (AIIMS)
Postdoc Fellowship, Bristol University (England)
Doctorate in Educational Leadership (USA)
Scientific Research
Research conducted for the purpose of
contributing towards science by the
systematic collection, interpretation and
evaluation of data and that, too, in a
planned manner is called scientific
research
It is a systematic collection, analysis and
interpretation of data to answer a question
or solve a problem
Classification of Scientific Research
According to data collection techniques:
Observational
Experimental
According to causality relationships:
Descriptive
Analytical
According to relationships with time:
Retrospective
Prospective
Cross-sectional
Classification of Scientific Research
According to the medium through
which they are applied:
Clinical
Laboratory
Social descriptive research
Classification Of Research
I. Descriptive research
Case series
Surveillance studies
II. Analytical research
Observational studies: cohort, case control and cross-
sectional research
Interventional research: quasi-experimental and clinical
research
Analytical observational research
OBSERVATIONAL
Cohort Studies (Prospective, Retrospective and
Ambidirectional)
Case-Control studies
Cross-Sectional studies
INTERVENTIONAL RESEARCH (Experimental
Studies)
1. Quasi-Experimental Research
2. Clinical Research
Quantitative Research
It is numerical, non-descriptive, applies statistics
or mathematics, and uses numbers.
It is an iterative process whereby evidence is
evaluated.
The results are often presented in tables and
graphs.
It is conclusive.
It investigates the what, where, and when of
decision-making.
Qualitative Research
It is non-numerical, descriptive, applies to
reason, and uses words.
Its aim is to get the meaning, and feeling and
describe the situation.
Qualitative data cannot be graphed.
It is exploratory.
It investigates the why and how of decision-
making
Types of Questions
Academic “Basic”
To add to our scientific
knowledge
Applied “Practical”
To solve our practical
problems
Research: the Fountain of Knowledge
A research problem refers to a difficulty that a researcher or
a scientific community or an industry or a government
organization or a society experiences. It may be a theoretical
or a practical situation. It calls for a thorough understanding
and possible solutions.
Research provides the basis for many government policies.
For example, research on the needs and desires of the
people and on the availability of revenues to meet the needs
helps a government to prepare a budget.
It is the fountain of knowledge and provides guidelines for
solving problems.
Step 1. Define a Research Area
Areas for
research are
very broad
and
overlapping
Selection depends on
Researcher’s
interest
Actual need
Available resource
Step 2: Select a Research Topic
Researcher can
not study every
topic in the
selected area
Magnitude of
problem
Seriousness
Preventability
Curability
Feasibility
Step 3. State the Research
Objectives?
State objectives at
the beginning of
study
State them clearly
Objectives are stated
in two forms
Goal (general
objectives)
Specific objectives
Research Objectives Should Be
Closely related to research questions
Covering all aspects of the problem
Very specific
Ordered in logical sequence
Achievable (take in account time & resource)
Mutually exclusive (no repetition, no overlaps)
Stated in action verb that can be evaluated
Step 4. Develop a Research
Question
Before start, you make sure that
You has a research question
Question is clear and specific
Reflect the objective (s)
Has no answer by common sense
Has no answer in literature
Finding an answer will solve the problem of
the study
At this stage, review of
literature is very
important to ensure that
research question has
no answer
Review of literature is important
Identify valid question
Refine question(s)
Avoid un-needed research
Avoid duplication
Avoid pitfalls of previous studies
Provide scientific background
Give rationale for study
Examples of Questions (1)
Description
What is the incidence
of disease “D”?
What is the prevalence
of disease “D”?
What is the rate of
risk factor “F” in the
community?
Examples of Questions (2)
Etiology
What is the cause of
disease “D”?
Is exposure “E”
associated with disease
“D”?
What is the risk factor
“F” associated with
the disease “D”?
Examples of Questions (3)
Diagnosis
Is test “T1” better
than test “T2” in
diagnosis disease “D”?
What the value of test
“T” in diagnosis
disease “D”?
Examples of Questions (4)
Therapy
Is drug “A” better
than drug “B” in
treatment of disease
“D”?
Is surgery more
effective than
conservative
treatment for disease
“D”?
Examples of Questions (5)
Prognosis
What is the five-year
survival of patients
with disease “D”?
What is the five-year
survival of patients
having disease “D”
after intervention “I”?
Step 5: Formulate a Hypothesis
It is a statement
of research
question in a
measurable
format
Hypothesis; must be:
Based on scientific
background
Translation of research
question
Reflect the study design
Use the study variable
Test only one relationship
Stated in measurable terms
Example for Formulating a Sound
Hypothesis
A prospective cohort study was designed
to answer the research question: “ is
hypercholesterolemia (HC) a risk factor
for coronary artery disease (CAD)?”
Examine the Following
Hypothesis
1. All subjects having HC will develop CAD
2. Dietary habits affect the risk of CAD
3. Rats given high fat diet will develop CAD
4. Atherosclerosis is associated with a high risk of
CAD
5. Increased serum cholesterol and triglycerides
and decreased HDL lead to increased risk f
CAD
6. CAD is high among hypercholesterolemic
subjects
7. Subjects with HC have a higher risk of
developing CAD compared to subjects without
HC
8. The prevalence of CAD among HC subjects is
higher than that in subjects without HC
9. The probability of HC in CAD patients is
higher than in subjects without CAD
Hypothesis (1) is refuted because it is not biologically
plausible
Hypothesis (2) is refuted because it is not related to
research question
Hypothesis (3) is refuted because it does not reflect the
study design (cohort not experiment)
Hypothesis (4) is refuted because it does not use the
study variables
Hypothesis (5) is refuted because it examine more than
one relationship. It is better to use separate hypothesis
for each studied association
Hypothesis (6) can be refuted because it is
not stated in a measurable terms.
Expression “high” is very subjective
Hypothesis (8) & (9) an be refuted because
they do not reflect the study design.
Although they fulfill all other criteria
Hypothesis (7) is a sound hypothesis
STUDY DESIGNS
Classification
depends on role of
investigator in
controlling factors
under study
Observational
Ecological
Cross sectional
Case control
Cohort
Experimental/interven
tion
Lab experiments
Clinical trials
Community interventions
Quasi-experimental
1. Ecological Study
Sampling units are
groups
Good source for
new hypothesis
Problems
Ecological fallacy
Temporal
relationship
2. Cross Sectional Study
Data collected at a
single point in time
No control group
Prevalence rate can be
calculated
Develop new
hypothesis
“Snapshot”
Prevalence vs. Incidence
Prevalence
The total number of
cases at a point in
time
Includes both new
and old cases
Incidence
The number of new
cases over time
Example of a Cross-Sectional
Study
Study of association between garlic
consumption & CAD in the Family
Practice Clinic
Cross-sectional Study
Sample of Population
Garlic Eaters Non-Garlic Eaters
Prevalence of CAD Prevalence of CAD
Time Frame = Present
Cross-Sectional Study
Strengths
Quick
Cheap
Weaknesses
Weak evidence of
association
Lack of
representativeness
Absence of temporal
relationship
No control of
confounders
Prevalence-incidence
bias
3. Case-Control Study
Start with people who have disease
Compare them with controls that do
not
Look back and assess exposures
Case-Control Study
Patients with CAD
Patients w/o CAD
Present
Past
High Garlic Diet
High Garlic Diet
Low Garlic Diet
Low Garlic Diet
Cases
Controls
Do not eat
Eat garlic
ill not ill
49 49 98
4 6 10
Presentation of Case-control Study
Strengths
Good for rare outcomes: cancer
Can examine many exposures
Useful to test hypothesis
Fast & easy
Cheap
Provides Odds Ratio
Minimal ethical problems
No risk to participants
No attrition problems
Intuitively
If the frequency of exposure is higher
among cases than controls
Then the incidence rate will probably be
higher among exposed than non
exposed.
Cohort Study
Begin with disease-free individuals
Classify individuals as exposed/unexposed
Record outcomes in both groups
Compare outcomes using relative risk
Strengths
Provides incidence data
Establishes time sequence for causality
Eliminates recall bias
Suitable for rare exposure
Allows for accurate measurement of
exposure variables
Strengths
Can measure multiple outcomes
Can adjust for confounding variables
Can calculate relative risk
Dose response can be calculated
Weaknesses
Expensive
Time consuming
Problems of attrition
Cannot study rare outcomes (diseases)
Some ethical problems
Weaknesses
Exposure may change over time
Disease may have a long pre-clinical phase
Change of diagnostic criteria
Experimental Studies
Clinical trials provide the “gold
standard” of determining the
relationship between garlic and
cardiovascular disease prevention.
Strengths
Best measure of causal relationship
Best design for controlling bias
Can measure multiple outcomes
Weaknesses
Ethical issues
Unawareness of patients
Withholding beneficial treatment
Inability to refuse (prisoners)
Potential side effects
Co-intervention & contamination
Feasibility problems
Small Group Tasks
Elect a leader and a speaker
Read your scenario
Design a study to answer the research question
Your funds are not unlimited!
Briefly present your study to the large group
Time:
10 min for study design
10 min for group presentations
Selection of Research Design
Selection of the best
research design
depends on many
factors
Purpose of study
State of existing
knowledge
Characteristics of study
variables
(exposure/outcome)
Latency
Feasibility
Research: The Fountain of Knowledge
A research problem refers to a difficulty that a
researcher or a scientific community or an industry or a
government organization or a society experiences. It
may be a theoretical or a practical situation. It calls for
a thorough understanding and possible solutions.
Research provides the basis for many government
policies. For example, research on the needs and
desires of the people and on the availability of revenues
to meet the needs helps a government to prepare a
budget.
It is the fountain of knowledge and provides guidelines
for solving problems.