Topic   Sampling

10  

LEARNING OUTCOMES
By the end of this topic, you should be able to:
1. Define what is sampling and its concept;
2. Identify the advantages and disadvantages of sampling;
3. Describe sampling terminologies;
4. Identify sample size and selection method; and
5. Differentiate between probability sampling and non-probability
sampling techniques.

10.1 CONCEPT OF SAMPLING

Sampling is a process of selecting samples from a group or population to become
the foundation for estimating and predicting the outcome of the population as
well as to detect the unknown piece of information. A sample is the sub-unit of
the population involved in your research work. There are a few advantages and
disadvantages associated with the sampling process.
124  TOPIC 10 SAMPLING

(a) Advantages of Sampling

Among the advantages are that sampling can save cost and human
resources during the process of research work. In ICT, sampling does
not cause much constraint such as heavy use of tools and technology
in predicting the research output.

(b) Disadvantages of Sampling

A researcher may not find the information about the population being
studied especially on its characteristics. The research can only estimate
or predict them. This means that there is a high possibility of error
occurence in the estimation made. Sampling process only enables a
researcher to make estimation about the actual situation instead of
finding the real truth. If you take a piece of information from your
sampling population, and if your reasoning is correct, your findings
should also be accurate to a certain degree.

When selecting a sample, it is very important for a researcher to consider

the possibility of error during the selection process. In the field of ICT,
sampling has little significance because the main purpose of ICT research is
to explore or describe diversity in technology, phenomenon and issues.
Another factor is the nature of ICT research which focuses on qualitative
approach. Qualitative approach does not make an attempt to quantify or
determine the extent of diversity. A researcher can select a sample and
describe his/ her inquiry based on the research problem.Then, the study
proceeds based upon the obtained sample.

You must always remember that qualitative research has a characteristic called
saturation point. Saturation point is where a researcher reaches the limit of
obtaining information after many attempts to get new information. When you
find you are not obtaining new information, it is assumed you have reached
the saturation point. Again, saturation point is subjective judgement which a
researcher always decide about it in the entire research process.
 TOPIC 10 SAMPLING  125

10.2 SAMPLING TERMINOLOGIES

In sampling, there are a few terminologies that a researcher should be familiar
with. For example, letÊs say you are working in a research project on computing
implementation for elderly and disabled citizens for a smart home system. You
are supposed to find out the average age of senior and disabled citizens involved
in your study.
(a) The community, families living in the town with smart homes form the
population or study population and are usually denoted by the letter N.
(b) The sample group of elderly people or senior citizens and disable people in
the vicinity of the smart home community is called sample.
(c) The number of elderly people or senior citizens and disabled people you
obtain information to find their average age is called the sample size and is
usually denoted by letter n.
(d) The way you select senior citizens and disabled people is called the
sampling design or strategy.
(e) Each citizen or disabled people that becomes the basis for selecting your
sample is called the sampling unit or sampling element.
(f) A list identifying each respondent in the study population is called
sampling frame. In case when all elements in a sampling population cannot
be individually identified, you cannot have a sampling frame for the study
population.
(g) Finally, the obtained findings based on the information of the respondents
are called sample statistics.

ACTIVITY 10.1

Identify the importance of sampling in a research work.

126  TOPIC 10 SAMPLING

10.3 SAMPLE SIZE AND SELECTION

Most of the new researchers always wonder about the sample size that needs to
be selected. You must remember that the larger the sample for your research, the
better outcome you can evaluate at the end of the research process. The larger the
sample, the more likely the sample mean and standard deviation will become a
representation of the population mean and standard deviation. For instance, in IT
survey, the sample size required depends on the statistical outcome needed for
the findings. The following are some guidelines to decide on how large a sample
should be:

 When the selected sample needs to be segregated into smaller clusters

involving comparisons of clusters, a large sample would be appropriate.
 The longer the duration of a study, the higher the number of subjects that
will drop out. To reduce attrition, a researcher should keep demands on
subjects to the minimum, to fully inform the subject about the study and
research, and make frequent communication with subjects to maintain the
interest.
 A larger sample is needed when the population is highly heterogeneous on
the variables being studied so that different characteristics can be identified.
If members of the population is less, then a small sample size would do to
obtain the necessary characteristics.

Gay and Airasian (2003) also offered similar guidelines in their work for selecting
sample size in a research study. In their work, it is mentioned that beyond a
certain point (at about 5,000 respondents or more), the population size is almost
irrelevant and a sample size of 400 should be adequate. To some extent, the size
of sample depends on how homogenous or heterogenous the population is and
how alike or different its members are with respect to the characteristics of
particular research interest.

10.3.1 Selecting a Sample

The objective of selecting a sample is to achieve maximum accuracy in your
estimation within a given sample size and to avoid bias in the selection of the
sample. This is important as bias can attack the integrity of facts and jeopardise
your research outcome.

Table 10.1 explains how bias can occur in sample selection.

 TOPIC 10 SAMPLING  127

Table 10.1: Reasons Bias Occurs in Sample Selection

No. Reason(s)
1. Sampling done using non-random method (we will see sampling designs in the
next section) which can be influenced by human choices.
2. Sampling frame like list, indexing and records which serve as the platform of
selection does not cover the sampling population accurately or completely.
3. A section of sampling population refuses to co-operate.

There are also factors that may influence the degree of certainty in inferences
drawn from a sample for research study. As we know, the size of samples
influence findings such that large samples have more certainty than those based
on smaller ones. Therefore, the larger the sample, the researcher will obtain more
accurate findings.

Another factor is the extent of variation in the sampling population where the
greater the variation in the population will have greater uncertainty with respect
to its characteristics. Therefore, it is crucial for a researcher to bear these in mind
especially when selecting a sample for her/his respective research work.

SELF-CHECK 10.1

1. Describe the guidelines on selecting samples.

2. Identify how bias can occur during sample selection process.

10.4 SAMPLING TECHNIQUES

Sampling techniques often depend on research objectives of a research work.
Generally there are two types of sampling techniques that are widely deployed.
These techniques are:

(a) Probability Sampling

This sampling technique includes sample selection which is based on

random methods. The techniques that are based in this category are
random sampling, stratified sampling, systematic sampling and
cluster sampling.
128  TOPIC 10 SAMPLING

(b) Non-probablity Sampling

This sampling techniques is not based on random selection. Some

examples are quota sampling, purposive sampling and convenience
sampling.

10.4.1 Probability Sampling

The techniques in probality sampling are as follows:

(a) Random Sampling

Random sampling is used to increase the probability of the sample selected.
By deploying this technique, each member of a population stands a chance
to be selected. LetÊs say you are interested to survey the usage of e-
commerce application in business-to-consumer (B2C).

The sample you select needs to represent the types of e-commerce

application and its usage. Due to financial and time constraints you are
unable to survey the usage of all types of e-commerce application across the
Malaysian network (N= 100,000). Therefore you decide to confine the study
to e-commerce application for merchandise products in Malaysia (n=10,000)
which is called the accessible population.

From this accessible population, a sample of 100 e-commerce application is

retrieved. How do we randomly select sample? It is understood that
random sample is a procedure in which all individuals in the defined
population have an equal and independent chance to be selected in the
sample design. In the above example, the number of e-commerce
application on merchandise products across Malaysian network is 10,000
and you may intend to draw a sample of 100. When you select the first
application, it has 1:10,000 chances of being selected. Once the first
application selected, the remaining will be 9,999 so that each application
has 1:9,999 of being selected as second case. Therefore, once each case is
selected, the probability of being selected next changes because the
population of selection has become one case smaller each time.

(b) Stratified Sampling

In some IT surveys, a researcher may want to ensure individuals with
certain characteristics are included in the sample to be studied. For such
cases, stratified sampling is used. In this sampling design, a researcher will
attempt to stratify population in such a way that the population within a
 TOPIC 10 SAMPLING  129

stratum is homogeneous with respect to the characteristics on the basis of

which it is being stratified. You must bear in mind that it is important for
the characteristics chosen as the basis of stratification, are clearly
identifiable in the population. For example, it is much easier to stratify the
population on the basis of gender rather than age or income group.

(c) Systematic Sampling

Systematic sampling also known as Âmixed samplingÊ category since it has
both random and non-random sampling designs. A researcher has to begin
by having a list names of members in the population, in random approach.
Figure 10.1 below shows the order of the sampling.

Figure 10.1: Example of systematic sampling

This sampling method is good as long as the list does not contain any
hidden order. Systematic sampling is frequently used in ICT research and
survey, especially in selecting specified number of records from computer
documents.

(d) Cluster Sampling

In cluster sampling, the unit of sampling is not referring to an individual
entity but rather a group of entities. For example, in an organisation there
are 25 departments and in each department there are an estimated 20 IT
administrators. You need a sample of about 100 staff but this would mean
going to many departments if random sampling approach is used. Using
cluster sampling, you may select 5 departments randomly from a total of 25
departments. You study all the staff in the 5 departments you chose. The
advantage that can be highlighted here is: it saves cost and time especially if
the population is scattered. The disadvantage is that it is less accurate
compared to other techniques of sampling discussed.
130  TOPIC 10 SAMPLING

10.4.2 Non-Probability Sampling

In some research scenarios, it is not possible to ensure that the sample will be
selected based on random selection. Non-probability sampling is based on a
researcherÊs judgement and there is possibility of bias in sample selection and
distort findings of the study. Nonetheless, this sampling technique is used
because of its practicality. It can save time and cost, and at the same time, it is a
feasible method given the spread and features of a population. Some common
sampling methods are quota sampling, purposive sampling and convenience
sampling.

(a) Quota Sampling

The main reason directing quota sampling is the researcherÊs ease of access
to the sample population. Similar to stratified sampling, a researcher needs
to identify the subgroups and their proportions as they are represented in
the population. Then, the researcher will select subjects based on his/ her
convenience and judgement to fill each subgroup. A researcher must be
confident in using this method and firmly state the criteria for selection of
sample especially during results summarisation.

(b) Purposive Sampling

This sampling method is selected on the basis that members conform to
certain stipulated criteria. You may need to use your own judgement to
select cases to answer certain research questions. This sampling method is
normally deployed if the sample population is small and when the main
objective is to choose cases that are informative to the research topic
selected. Purposive sampling is very useful in the early stages of an
exploratory study. One of the disadvantages of this technique is that the
sample may have characteristics different from population characteristics.

(c) Convenience Sampling

Using this sampling method, a researcher is free to use anything that they
could find in the research outline. The sample is selected based on
preferences and ease of sampling respondents. This sampling is easier to
conduct and less expensive. However, it has poor reliability due to its high
incidence of bias. In ICT, convenience sampling seems to be dominant
especially in cases of organisations that conduct web surveys, mail their
responses to a survey questions and SMS their opinions to a question.
Although convenience sampling can cater to a lot of data, it is not reliable in
terms whether the sample represents the real population or not.
 TOPIC 10 SAMPLING  131

SELF-CHECK 10.2

1. What are the sampling techniques of probability sampling?

2. Distinguish between quota and purposive sampling techniques.

 Sampling is a process of selecting samples from a group or population to

become the foundation on estimating and predicting outcome of the
population.
 Two main techniques of sampling: probability and non-probability.
 Probability sampling is based on random selection while non-probability
sampling is not based on random selection.
 Probability sampling consists of random sampling, stratified sampling,
systematic sampling and cluster sampling.
 Non-probability sampling consists of quota sampling, purposive sampling
and convenience sampling.
 In cluster sampling, the unit of sampling does not refer to an individual
entity but a group of entities.

Non-probability sampling Sampling

Probability sampling Sampling design
Purposive sampling Sampling population
Random sampling Sampling statistics
132  TOPIC 10 SAMPLING

You wish to study the impact of Service Oriented Architecture

adoption among software architects for software testing methodology.
There are 300 software architects in 15 departments in the unit you
choose. You plan to obtain 1,000 software architects using the cluster
sampling technique. Describe the steps you would take in selecting the
sample.

Internet Resources

Easton, V. J. & McColl, J. H. (2007) Statistics Glossary: Sampling [Electronic

Version ]http://www.stats.gla.ac.uk/steps/glossary/sampling.html

Galloway, A. (1997). Sampling: A Workbook [Electronic version]

http://www.tardis.ed.ac.uk/~kate/qmcweb/scont.htm

Trochim, W. K. (2007). Research method tutorials [Electronic version]

http://www.socialresearchmethods.net/kb/sampling.php