VARIABLES
MODERATORS:
PRESENTOR:
Prof. Najam Khalique
Dr. Ajitha A
Dr. Saleem Mohammad Khan
OUTLINE:
Definition
Types of variable
Selection of variables
Number of variables
Defining the variables
Clarifying the variables
DEFINITION OF THE VARIABLES:
The characteristics that are measured are referred to as
variables, whether they are measured
1. Numerically - age or height
2. Categories - sex
3. Presence or absence of disease
TYPES OF VARIABLES
1. DEPENDENT VARIABLE
2. INDEPENDENT VARIABLE
3. COMPOSITE VARIABLE
4. COMPLEX VARIABLE
1. DEPENDENT VARIABLE:
The variables in which changes are results of the level or
amount of the independent variable or variables.
Synonyms: effect, outcome, consequence, result, condition, disease
When an association between variables is studied, the variables
may be referred to as dependent and independent.
Eg. Prevalence of a disease in different age and sex groups
Dependent variable Independent variable
Whether transcendental meditation affects the blood pressure
Dependent variable
2. INDEPENDENT VARIABLE:
Variables that are manipulated or treated in a study in order to see what
effect differences in them will have on those variables proposed as
being dependent on them.
Synonyms: cause, input, predisposing factor, risk factor, characteristic
Eg. Hypertension Coronary heart disease
(independent variable)
Salt intake Hypertension
(dependent variable)
3. COMPOSITE VARIABLE:
A variable based on two or more other variables may
be termed a composite variable.
Eg. adiposity may be measured in terms of BMI
Incidence or prevalence rates
Sex ratios
4. COMPLEX VARIABLE:
Some variables are too complex to be easily measured as single
entities and are best broken up into component aspects that can be
regarded as separate variables and measured separately.
Eg. if we wish to study ECG findings we may, instead of making
a global and probably subjective appraisal of the ECG pattern,
give separate consideration to a series of different measurable
aspects, Q and QS patterns, S-T junction and segment depression
etc.
SELECTION OF VARIABLES
The variables to be studied are selected on the basis of their
relevance to the objectives of the investigation.
Depends on any of the following five grounds:
1. Whether it is important enough to warrant study as an
independent variable in its own right.
2. Whether it is a possible confounding factor.
Hypertension
Obesity Coronary heart disease
3. Whether it may be a modifier variable
4. Whether it may be an intervening cause
5. Whether it has a strong influence on the dependent
variable to warrant its inclusion, without necessarily
meeting the above criteria.
Eg. In a study of relationship between work accidents and age.
Type of work (confounding factor)
Age work accidents
( Independent variable ) (dependent variable)
Modifier variable: years of employment, department, sex
Intervening cause: Can a difference in the use of a protective
equipment explain it?
NUMBER OF VARIABLES
How many variables should be studied?
‘As many as necessary and as few as possible’
Unnecessary variables - increase cost
Impair participation
Complicate the analysis
DEFINING THE VARIABLES
Each of the variables measured in a study should be clearly
and explicitly defined.
Two kinds of definition:
1. Conceptual definition
2. Operational definition
1. CONCEPTUAL DEFINITION:
Defines the variable as we conceive it.
In effect, it is the characteristic we would like to measure.
Eg. ‘Obesity’ might be variously defined as ‘excessive fatness’
or ‘overweight’ or ‘a bodily condition that is socially regarded
as constituting excessive fatness’.
2. OPERATIONAL DEFINITION: (working definition)
It defines the characteristic we will actually measure.
It is phrased in terms of objectively observable facts and it is
clear and explicit enough to avoid ambiguity. Where
necessary, it states how the facts are obtained.
Eg. ‘Obesity’ might be operationally defined in different
surveys based on weighing or skinfold thickness
CLARIFYING THE VARIABLES
Once the variables have been selected, each of them should be
clarified.
Two aspects to be considered:
1. Operational definition - must be formulated, clearly defining
the variable in terms of objectively measurable facts, and stating, if
necessary how these facts are to be obtained.
2. Scale of measurement - to be used in data collection should
be specified
Eg. selected variables in a survey of illnesses among
hospitalized infants.
It contains two variables:
1. Age - based on the date of birth and date of admission
2. Social class - inferred from father’s occupation
REFERENCES
1. Abramson J, Abramson Z. 11 12. In: Survey methods in
Community Medicine: Epidemiological research,
programme evaluation, clinical trials. Edinburgh: Churchill
Livingstone; 2008. P.102-116.
2. Health Research Methodology: A guide for training in
research methods. Second edition. Manila: WHO, Regional
office for the Western Pacific; 2001
THANK YOU