B.

Com Business Management (Hons)

SAMPLING
 Sampling Defined
❖ In research it is often difficult to collect data from all
relevant research participants (elements) and under those
circumstances sampling becomes unavoidable.
❖ Sampling in research comes by default.
❖ Sampling is therefore a process of selecting a sufficient
and representative number of elements from the target
population so that by studying or analyzing the smaller
group (sample) the results can be generalized to the target
population with confidence.
❖ One attribute of a good sample is that it must be both big
enough and representative of the population from which it
has been drawn especially if results are to be used to make
inferences.
2 L MUTANDWA 4/27/2017
 Types of Sampling Designs

•Sampling Designs
•Probability Sampling Procedures

•Non Probability Sampling Procedures

3 L MUTANDWA 4/27/2017
 Probability Sampling Methods
⚫ This represents sampling procedures whereby all
research elements have a known chance of being
selected in the final sample.
⚫ It is an objective way of selecting research subjects and
therefore normally leads to representativeness if
applied correctly.
⚫ These methods are normally used in a quantitative
research where hypotheses are used thus deductive
approach.
⚫ They are also linked to explanatory and predictive
studies.
⚫ Major methods for this design include:
4 L MUTANDWA 4/27/2017
 Simple Random Sampling
❑ It is the most basic method of sampling.
❑ This is applicable where each element has an equal
chance of being selected in the final sample.
❑ It is applicable when the population is large and
relatively homogenous relative to the data required by
the researcher.
❑ The major drawback for research is that, it might lead
to selection of less well informed research subjects
because everyone has an equal chance of being
selected.

5 L MUTANDWA 4/27/2017
 Systematic Random Sampling
❖ This is the only sampling procedure which sacrifices
randomness or unpredictability.
❖ Only the first research subject is picked at random,
thereafter subsequent research subjects are picked
systematically at a uniform interval.
❖ It is applicable where the population is very large and
also homogenous.
❖ The weakness is that there is a remote chance of
systematic bias which might compromise the quality of
data collected for analysis thus negatively affecting the
research findings.
6 L MUTANDWA 4/27/2017
 Stratified Random Sampling
It is the only sampling procedure that is applicable to a
heterogeneous population.
It is not an end itself as it only assist the researcher to group
elements in the population into homogeneous sub-groups in
preparation to pick the final sample.
After the initial grouping, another sampling procedure,
simple random sampling, is then used to pick final
research subjects.
Stratified Random Sampling ensures that no sub-group is
left unrepresented.
It can further be split into proportionate and
disproportionate stratified random sampling.
7 L MUTANDWA 4/27/2017
 Cluster Random Sampling
✔ Under this sampling procedure, the population is
divided into clusters, where each cluster is similar in
profile to every other cluster.
✔ Clusters are then randomly selected for sampling or
units from these selected clusters are picked as sample
subjects.
✔ Cluster random sampling is normally used when the
population is relatively homogeneous, very large and
spread over a wide geographical area, for instance,
sampling the population of teachers, nurses or police
officers the whole country.
8 L MUTANDWA 4/27/2017
 Non Probability Sampling
Methods
⚫ In this sampling design the elements in the population have no
known chances of being selected as research subjects.
⚫ As a result sampling is done on the basis of some other factors
which has nothing to do with probability.
⚫ This means results from a study of such samples cannot be
generalized to the entire population of objects in question.
⚫ Non probability sampling procedures are therefore mainly used
when researchers are not concerned with generalizability of
findings but to just obtain some preliminary information quickly
and cheaply.
⚫ Representativeness of a sample is therefore not of importance
under these methods.
⚫ The methods are normally linked to qualitative studies that are
descriptive and exploratory in nature thus inductive.
⚫ Major methods under this design include:

9 L MUTANDWA 4/27/2017
 Convenience Sampling
❖ As the name implies, this method involves collecting
data from population elements who are conveniently
available to provide it.
❖ It thus represents a sample drawn to suit the
convenience of the researcher.
❖ It is most used during the exploratory phase of a
research and perhaps the best method of getting some
basic information quickly and efficiently.
❖ The method is used when the target population is
relatively homogeneous.

10 L MUTANDWA 4/27/2017
 Judgmental/Purposive Sampling
❑ This involves the choice of subjects which are in the best
position to provide the data required for research, for
instance, if the researcher wants to know what it takes for
women managers to make it to the top of their
organisations.
❑ The most suitable respondents to provide relevant data will
be women who have risen to the top of their companies’
hierarchy.
❑ The researcher’s judgment is what is used to determine the
most suitable research subjects but the basis of the
judgment must be given.
❑ The sampling method maybe the only useful one for
answering certain types of research questions.
L MUTANDWA 4/27/2017
11
 Quota Sampling
❑ Using this method the population is divided into
segments and a quota of each segment in the
population is used as a basis to select representatives
(subjects) from each segment.
❑ Quota sampling is almost similar to proportionate
stratified random sampling but only differs in the
sense that the predetermined proportions of
respondents are sampled on a convenience basis.

12 L MUTANDWA 4/27/2017
 Snowball Sampling
⚫ This a special sampling technique which is used when
the target population is very difficult to locate because
of various factors such as stigmatization and lack of
accessibility.
⚫ As a researcher you would therefore need to:
a) Make contact with the first element in the population.
b) Ask them to identify further cases.
c) Ask the new cases to identify further new cases and
so on.
d) Stop when either no new cases are given or when the
sample is large enough to produce enough data for
13 presentation and analysis.
L MUTANDWA 4/27/2017
 Cont..
⚫ The main problem is on making the initial contact,
once the researcher has done that, the particular subject
is used to identify further cases and the sample
snowballs.
⚫ It is the only sampling technique applicable when it is
difficult to identify a target population.
⚫ It thus calls for special efforts to locate and gain access
to the needed population elements.

14 L MUTANDWA 4/27/2017
 Note
For each sampling method used the researcher should
give justification for having chosen the particular
technique.
Sampling methods maybe chosen depending on
whether the population in question is homogeneous or
heterogeneous and this is relative to the data sought
by the researcher.

15 L MUTANDWA 4/27/2017
 The End

16 L MUTANDWA 4/27/2017