1. Dr. Hinda Hassan Khideer Mahmood
MBBS(Uof K), MRCP,Nephrology Fellowship(SMSB)
2. At this stage, the researcher needs to ask 2
questions,
What is the information need to be collected to meet the
objectives?
How much information is required?
In other words, the variety of variables to be
collected and the amount of data to be collected.
3. Study variables are actually the factors that influence
the problem or the problem itself, transformed in a
way that could be measured
4. Factors Variables
1. Long waiting time Waiting time
2. Absence of drugs Availability of drugs
3. Lack of supervision Frequency of supervisory visits
4. Poor knowledge of causes of disease Knowledge of causes of disease
Example of factors changed into variables
5. A variable is defined as, a quality, characteristic or
constituent of a person, object or phenomenon that is
changeable and measurable. The variable can take
different values within a population e.g. age or within a
person e.g. blood pressure.
DATA (singular: datum) refers to the complete
set of observations or measurements recorded in
the course of a research process
6. In research, data comprise observations on one or
more variables. Studying for example a group of
patients with heart failure: The variables to be studied
may include: age, sex, weight, height, degree of leg
swelling and level of jugular venous pressure.
7. Data may take different forms according to the
observation made, but can broadly be divided into:
Categorical (Qualitative)
Numerical (Quantitative)
8. are where the data fall into a
few clearly defined categories, e.g.
dead or alive, male or female.
They are either Nominal or
Ordinal.
9. When categories are not ordered but simply have
names they are called nominal categories: Example:
blood groups: A, B, AB and O. There is no reason to
suspect that group A is any better than B. Another
example is marital status: married, widowed, single.
Race/ethnicity
Binary (dichotomous) data –Yes/No
–Cure—Yes/No
–Gender—Male/Female
10. The categories are ordered in some way. Examples
include disease stages e.g. in Hodgkin's lymphoma:
stage I, II, III, IV, or Stage I being restricted to one side
of the diaphragm (i.e. early) while in stage IV the
disease, is disseminated all over the body (very severe).
11. are variables which take numerical values.
Numerical variables may either be:
Discrete data:
Occur when the variable can only take certain whole
numerical value. Discrete data cannot have a value
with fraction or decimal place (i.e. whole figure).
Example: Number of children per family 2 or 3.
Number of migraines headache attacks in a patient is 2
per week.
Continuous data:
Continuous data can take a whole range of values e.g.
age, blood pressure, temperature, blood glucose. The
temperature may be 37.4C, 37.5C, 38C, 38.2C,
38.5C, etc.
12. Why is it important to distinguish between data
types? We often use very different statistical methods
depending on whether the data are categorical or
numerical.
Continuous data can sometimes be turned into
categorical data to make it easier to understand. We
might for example categorize blood pressure into
hypertensive, or normotensive, or the blood glucose
response as diabetic, glucose intolerant or normal.
13. In health research we often look for causal explanations,
so it is important to make a distinction between dependent
and independent variables.
Dependent variable: describes or measures the problem
(e.g. disease outcome, performance) under study.
Independent variable: describes or measures the factor
that is assumed to cause or at least influences the problem.
In other words, independent variables can be thought of
as interventions or treatments (it can be manipulated),
some patients may receive the treatment at varying
dosages. Independent variables are sometimes called
predictors and dependent variables are called outcome.
14. Whether a variable is dependent or
independent is determined by the statement
of the problem and the objectives of the
study. It is therefore important when
designing a study to clearly state which
variables are the dependent and which are
the independent ones.
Cause Effect /outcome
(independent variable) (dependent
variable)
15. After the selection of variables, they should be
clarified and defined. This is done to assure if different
investigators performed the study, similar findings will
be obtained and to assure if the same investigator
repeated the study, similar results will be found. There
are two aspects that should be considered in defining
variables
1. Formulation of an operational definition
2. The specific scale of measurement used in data
collection
16. Example :
Q1: Health education involving active
participation by mothers will produce more
positive changes in child feeding than health
education based on lectures. What is the
independent/dependent variable?
A: Independent variable: types of health education.
Dependent: change in child feeding.
17. in Indonesia wilson et al 1991reported improvement
in hand washing behavior in the village after a
programme in which 465 mothers were given soap and
an explanation of the fecal-oral route of transmission
what is the independent/dependent variable?
Independent :health education and soap
Dependent :change in hand washing behavior
18. Consider this scenario:
A researcher is interested in a risk factors for coronary
artery disease, we might take a work sample of men
who enjoy themselves playing cards in Buri club in the
evenings, and another group (the fitness group) go
exercising by walking from Buri to Riyadh every
evening
We might ask the questions: How protective is exercise
for coronary artery disease?
If we look at the fitness group (who walk everyday)
we probably find that they have much less heart
disease than the card players and it would be tempting
to claim that exercise is highly protective against heart
disease - tempting but probably wrong.
19. The two groups may have a lot of different habits
that are associated with heart disease. It is possible
that the fitness group are highly conscious about their
weight and on the habit of eating low fat foods . Maybe
that the playing-cards group get tense during playing
and smoke heavily. On the other hand, there may be
some people who started the habit of exercising (by
walking) only when they knew that they have heart
disease.
20. All these factors which may contribute to the outcome
and are associated with the tested exposure are called
CONFOUNDING factors.
Confounding is a distortion where one exposure is
associated with another exposure that is also a risk factor
for the disease in question (in the given scenario… eating
habits and exercise, smoking and lack of exercise).
This can cause incorrect conclusions to be drawn. It is
important to remember that to be a confounding factor,
something must be associated both with exposure and with
outcome.
22. Example:
A relationship is shown between the low level of the mother's
education and malnutrition in under years old children.
However, family income is related to the mother's
education as well as with malnutrition.
mother's education malnutrition
(Independent variable) (Dependent variable)
Family income
(Independent variable)
23. Family income is therefore a potential confounding
variable. In order to give a true picture of the relationship
between mother's education and malnutrition, the family
income should also be considered and measured. This
could either be
1. incorporated into the research design, such as by
selecting only mothers with a specific level of family
income,
2. or it can be taken into account in the analysis of the
findings, with mother's education and malnutrition
among their children being analyzed for families with
different categories of income.
24. In many studies variables such as age, sex, education
level, socio-economic, marital status, religion and ethnic
(tribal) status are collected. These are called background
variables. These background variables are often related to
a number of independent variables so that they may
influence the problem indirectly. (Hence, called
background variables).
Background variables are confounders and should be
considered when planning the study and/or when
analyzing the data.
25. The art of good research is to think about what
might be confounding factors, then to design studies
that get around them and control them as much as
possible.