Sampling

Sampling and Sampling Procedures
LECTURE 6
GENERAL CONSIDERATIONS
In most epidemiologic studies, we deal with a
sample of the population
 The study population may be:
 An entire population
 A group of the population with certain
characteristics
- age
- sex
- religion
- social status
 Hospital inpatients

THE SAMPLE
A sample is a portion of the population,
used for survey purposes when you can not
examine the entire population because the
resources are not adequate.
 A sample should be a representative group
of the whole population to be studied.
 When a sample is representative to the
whole population, generalization from the
sample can be made.

A number of techniques can make the sample
a representative group of the whole population.
 Only the simpler techniques will be considered
in this lecture, and when used with care will
meet the requirements of epidemiologic
studies.
 Complex sampling methods need the help of a
statistician.

SAMPLE SIZE

The precision with which the investigator make
generalization from a sample to the entire
population is related to the size of the sample.
THE SAMPLE
Sample selection needs a list of all members
(units) of the community from which a sample is
selected.
 The sampling units may be :
1. Individuals
2. Families
3. Blocks in a city
4. Class in a school
5. Whole school

SAMPLING PROCEDURES
Probability sampling:
 which is the most recommended method
because all sampling units have the same
probability of being selected and included in the
study.
 This method assures that the sample is a
representative of the whole population.
 Inferences from the sample can be made about
the whole population (generalization of results to
the whole population).
SAMPLING PROCEDURES
Sampling techniques include the following :
Haphazard sample:
 Ex- in an experiment, where the experimental
animal is chosen as the investigator can catch
the animal from the cage.
SAMPLING PROCEDURES
Selected sample:
 The individuals are selected according to the
investigator’ opinion that they are typical of the
population being studied and is appropriate for the
study objective .
 The sample may be selected because it is handy
sample.
 selection of groups known to be co-operative
 The disadvantage of this sample is that you can not be
confident that this sample represents the population, so
generalization cannot be made out of this sample.
SAMPLING PROCEDURES
Self-selected sample:
 The sample consists of individuals who
volunteered for an experiment.
 Such method can provide some information
about the population and can serve as a basis
for future adequate studies
 Generalization cannot be made out of this
sample.
SIMPLE RANDOM SAMPLE
A sample may be defined as random if every
individual in the population has an equal
chance of being included in the sample.
 Random selection is the basis of this sampling
technique.
 Any method of selection based on volunteering
or selection of groups known to be co-operative
is excluded and cannot be used for
generalization to the entire population.

Selecting a random sample
It is necessary to have a complete list of sampling
units of the total population:
 Individual persons
 House holds
 Schools
 cities
Selecting a probability sample
To select a simple random sample:
 Make a numbered list of the units in the
population (assign a number to each individual)
 Estimate the size of the sample
 Select the required number of sampling units
using a table of random numbers.
Selecting a random sample
If the size of the population is not too large:
 Write the numbers on small cards
 Place them in a bowl and mix them thoroughly
 The number representing the sample size is
selected from the bowl.
Selecting a probability sample
Selecting a random sample
Systematic sample ( more efficient):
 In this sampling method, every nth person in
the list is chosen and included in the sample
 If the list contains 10,000 units and the
researchers wants 1,000 units, he will select
one person every 10 persons (nth =10)
 The first person will be selected at random
between numbers from 1-10, then every
tenth unit will be selected.
CLUSTER SAMPLE:
This sampling method is used when a large
survey is planned to cover a district, governorate
or the entire country (two stage sampling).
 If there is no satisfactory list of the population is
available ( there is no list of a city’s population),
 a list of groups or clusters of individuals is
available, such as city districts, census
neighborhoods, villages or schools.
 a random sample of these clusters is selected
and all individuals in the cluster are included in
the study (one stage sampling).

CLUSTER SAMPLE:
An initial sample is taken from the units:
 Neighborhoods
entire district
 Villages
entire governorate
 Governorates
entire country
 Then a listing of individuals within the chosen
units (clusters) is made and a random sample
is taken from each (two stage sampling).
 The disadvantage to this method is that
disease may be accumulated in some clusters
giving misleading results.

Selecting a probability sample
Stratified random sample:
 If the severity of the disease differs according
one or more characteristics such as sex, age, or
socio-economic status, then the population to be
sampled should be divided into subgroups or
strata according to the characteristics that affect
the disease.
 Then random sample is selected within each
stratum and all individuals in each stratum are
included in the study .
Selecting a random sample
Stratified random sample:
To assess the prevalence of a disease like caries
which is unevenly distributed among age groups,
we have to obtain a sample stratified by age:
1. The study population is subdivided into age groups
such as 6-12, 13-18, 19-25 and 26 and over.
2. A certain number of each age group is then selected
randomly and all individuals in each age stratum are
included in the study .