In a simple sense, sampling refers to the method used
to select a given number of people (or things) from a
It may be termed as the process of selecting a few (a
sample) from a bigger group (the sampling population)
to become the basis for predicting the prevalence of an
unknown piece of information, situation, or outcome
regarding the bigger group.
- Kumar, 2005
The simplest rationale for sampling is that it may not
be feasible because of time or financial constraints, or
even physically possible, to collect data from
everyone involved in an evaluation. Sampling
strategies provide systematic, transparent processes
for choosing who will actually be asked to provide
- Mertens and Wilson, 2012
4. Sampling Terminology
Population is any group of individuals that has one or
more characteristics in common and that are of
interest to the researcher.
Sampling Units are non-overlapping collections of
elements from the population that cover the entire
A Sampling Frame is a list of sampling units.
A Sample is a small proportion of the population that
is selected for observation and analysis.
Inferential Statistics refers to a statistical technique
that is used for drawing conclusions about population
Statistics is known as the measurements of the
characteristics of the elements included in the
When these measurements are related to the
parent population, these are referred to as
Sampling Errors are those errors which arise on
account of sampling. This occurs due to certain
amount of inaccuracy in the collection of
information. It is inversely proportional.
7. Sample Size
A sample should be of optimum size which fulfils the requirements of
efficiency, representativeness, reliability and flexibility. It should be small
enough to avoid unnecessary expenditure. And should be large enough
to avoid sample errors.
Borg and gall (1979) suggest that, as a general rule, sample size should be
There are many variables.
Only small differences or small relationships are expected or predicted.
The sample will be broken down into subgroups.
The sample is heterogenous in terms of variables under study.
Reliable measures of the dependent variables are unavailable.
8. Factors affecting size of sample
The nature of
Kind of data
9. Some Findings About Sample Size
The larger the sample, the smaller the magnitude of sampling
Survey studies typically should have larger samples than are
needed in experimental studies because the returns from
surveys are from who, in sense, are volunteers.
Researcher should select large enough samples so that the
subgroups are of adequate size.
Subject availability and cost factors are legitimate
considerations in determining appropriate sample size.
10. Characteristics of
e of the
11. Steps in Sampling Procedure According to
W. C. Cochran
Formulation of the
objectives of the
Clear definition of
population to be
Use of appropriate
No overlapping in
the choice of a
adequate size of the
analysis of samples
12. Types of Sampling Strategies
Probability or Random
Non- Probability or Non-
Random Sampling Strategies
13. 1) Simple Random Sampling
• 2) Systematic Random
3) Multi-Stage Random
4) Stratified Random Sampling
• 5) Cluster Random Sampling
Non-Probability or Non-
Probability or Random
1) Convenience Sampling
• 2) Purposive Sampling
3) Quota Sampling Strategy
• 4) Snowball Sampling
14. 1) Probability or Random Sampling
The word ‘probability’ refers to the study of likelihood of
events, e.g., what are the chances of winning a toss?
In probability sampling, the sample is selected in such a way
that each unit within the population or universe has an equal
and independent chance of being included in the sample.
It deliberately attempts to seek proper representation of the
wider population instead of seeking representation of a
particular group or section of the wider population .
15. a) Simple Random Sampling Strategy
A sample selected by
randomization method is
known as simple random
Methods for selection of
The use of table of
The use of computer
1) Define the
2) List population
3) Decide the
4) Selection of
Simple to apply
Analysis of data is
Analysis of data has a sound
A complete list of the
population might not be
It is also possible that the
subpopulations of interest
might not be represented in
17. b) Systematic Random Sampling Strategy
It is a commonly used strategy, in which units of population
are arranged in some systematic manner. This strategy can
be used if the complete and up to date list of the sampling
units are available.
The mechanics of taking a systematic sample is simple. This
consists of selecting only the first unit at random, the rest
being automatically selected according to some pre-
determined pattern involved.
Make a list of all participants and
give them number 1-100.
Use any method of simple
randomization (lottery or random
method) to select the first
participant. For example participant
For selection of other participants,
select every nth participant.
n = 100/ 20 = 5
Therefore, select every 5th individual
after participant 6 till you obtain
desired number of par
Steps for Systematic
Easier to draw, without
More precise than simple
random sampling as more
evenly spread over
If the units are arranged in a
specific pattern, that could
result in choosing a biased
sample. E.g. Files kept in
20. c)Multistage Random Sampling
This method consists of a combination of sampling
strategies. It is choosing a sample from the random sampling
schemes in multiple / various stages (K.M.T. Collins, 2010).
Here the population is regarded as made of a number of
primary units each of which is further composed of a
number of secondary stage units and so on, till the
researcher ultimately reaches the desired sampling unit in
which he is interested.
21. For example, to get a sample of crop fields
growing wheat in Punjab:
1) Select a State (Punjab)
2) Get a sample of districts
3) Sample of villages from each selected district
4) Finally a sample of crop fields from each selected village.
It is convenient, less time-consuming and less expensive
method of sampling.
This kind of sampling is more flexible and enables existing
divisions and subdivisions to be taken into account.
23. 4) Stratified Random Sampling
This type of sampling requires dividing the population into
homogenous groups, each group containing subjects with
similar characteristics. These subgroups are known as strata.
The strata are the partition of the population , which is more
homogenous than the complete population.
The members of a stratum are similar to each other and are
different from the members of another stratum in the
characteristics that are to be measured.
24. Advantages of stratified Random Sampling by
Mendenhall, Ott and Schaeffer (1971)
It produces more homogenous data within each stratum
The logistics of collecting data within the specified strata
reduce the cost of the sample.
Since data are collected within each of the separately
defined strata, it is possible to obtain separate estimates of
population characteristics from each stratum.
25. 5) Cluster Random Sampling
When the population is large and widely dispersed gathering
a simple random sample poses administrative problems. So,
the subjects are selected in groups or clusters.
For example, to study the fitness level of students it would be
impractical to select students randomly. By cluster sampling,
the researcher can select a specific number of schools and test
all the students in those selected schools, i.e. a geographically
close cluster is sampled.
Cluster random sampling technique emphasizes first on
random selection of clusters from the large population of
clusters and then on including all the population members of
a selected cluster in the study sample.
27. 2) Non-Probability Random Sampling
In this sampling, the sample is selected in such a way that
the chance of being selected for each unit within the
population or universe is unknown.
It does not require random selection of sample. The sample
is selected in such a way that the chance of being selected
for each unit within the population or universe is unknown.
Because the selection of sample is subjective, there are no
statistical techniques that allow for the measurement of
28. a) Convenience Sampling Strategy
In this sampling strategy, the selection of units from the
population is based on easy availability or accessibility.
The researcher may try to include the people in their
research samples who are willingly available or conveniently
approached by them, e.g. members of their family.
The major disadvantage of this technique is that the
researcher has no idea how much representative the sample
29. b) Purposive Sampling
When the desired population is rare or vey
difficult to locate, purposive sampling May be
used as it targets a particular group of people.
E.g. studying juvenile delinquents.
Such sampling may satisfy researcher's need
but it does not pretend to represent the wider
population; it is deliberately and unashamedly
selective and biased (Cohen et al., 2007)
30. Variants of Purposive Sampling
To include those in a sample who otherwise may
be excluded from, or under represented because
there are few of them (Gorard, 2003).
Here the researcher deliberately seeks those people who
disconfirm the theories being advanced, thereby
the theory if it survives such disconfirming cases.
Selecting sample from as diverse a population as possible in
order to ensure strength and richness to the data, their
applicability and their interpretation.
31. c) Quota Sampling Strategy
It may be defined as a non – probability equivalent of
stratified sampling in which one attempts to represent
significant strata of the wider population without resorting to
randomization (Bailey, 1978).
Like stratified sample, a quota sample strives to represent
significant characteristics of a wider population; unlike
stratified sampling it sets out to represent these in
proportions in which they can be found in the wider
Example- sup[pose the wider population is were composed
of 55 % females and 45 % male, then the sample would
have to contain 55 % females and 45% males.
32. Steps of Quota Sampling Strategy
1) Identify those characteristics which appear in the wider population
which must also appear in the sample, i.e. divide the wider
into homogenous groups, e.g. male & female, rural & urban etc.
2) Identify the proportions in which the selected characteristics
appear in the wider population, expressed as percentage.
3) Ensure that the percentaged proportions of the characteristics
selected from the wider population appear in the sample.
33. d) Snowball Sampling
This type of sampling is basically sociometric. It is defined as
obtaining a sample by having initially identified subjects who
can refer you to other subjects with like or similar
characteristics (Eckhardt & Ermann, 1977).
A researcher a required to start with the identification of a
small number of participants who have the characteristics in
which he is interested.
These people are then used as informants to identify others
who qualify for inclusion ( Cohen et al., 2007).
34. Snowball Sampling is applied in the cases
The population is unknown.
The access to the population of the study is difficult.
One is to conduct ethnographic studies.
The topic of the research study is too sensitive ( e.g. to study
the teenage drug addicts)
35. Errors in Sample Research (Fienberg,
Errors of Bias
Errors of Bias
36. 1) Sampling Errors (SE)
Sampling errors results from the luck of the draw when choosing a
sample: one gets a few too many units of one kind, and not
enough of another.
With the probability samples, the SE can be estimated using (i) the
sample design and (ii) the sample data. As the sample size
increases, the SE goes down. If the population is relatively
homogenous, the SE will be small.
The errors arising in the outcomes of the research studies on
account of using samples instead of the total population are named
as sampling errors.
37. Types of Sampling Errors
The errors happen by chance, in
carving sample from the parent
population, are named as the
chance errors or random errors.
Unusual units in a population do
exist and there is always a
possibility that a one or two
abnormally large number of them
will be chosen .
This issue can be resolved by using
Errors of Bias in Sampling
It is a tendency to favour the
selection of units that have
The errors due to this
tendency cannot be corrected.
38. 2) Non-Sampling Errors
The bias associated with non-sampling errors is not
dependent o the size of the sample; rather, it is associated
with differences between those who respond to the
questions and those who do not.
Main factors of errors (Kothari,1990):
Errors due to observation.
Errors due to non-response and measurement errors.
Errors in processing & analysis of data.
Errors in preparation of reports.
Errors in coverage by using defective frame.
39. Types of Sampling Errors
The effect of these errors
approximately cancel out if fairly
large samples are used.
Errors of Bias in Sampling
These are brought about
intentionally by measuring
devices, the observer, and the
respondent of the study
through involvement of their
The effects of these errors
persist and not get cancelled
in the long run.
Sampling is the procedure a researcher uses to gather
people, places or things to study. The representative
proportion of the population is called a ‘sample’. Population
refers to the large group from which the sample is taken.
Probability and Non-Probability sampling are two strategies
of sampling. Sampling and Non-Sampling Errors may occurs
due to biased sample and wrong measurement.