Sampling Procedures

Probability Sampling Methods

• Probability sampling includes simple random sampling, stratified

sampling, systematic sampling, cluster sampling and multistage
sampling which permits randomization
Simple Random Sampling
• Simple random sampling gives an equal and independent chance of
selection for each element of the population as a sample element.
• Drawing a simple random sample includes enumerating all elements
in the population, specifying equal and independent chances of
inclusion of each element as a sample and random drawing of
numbers
• From a finite population of elements, simple random sampling with
replacement results in an equal probability of selection for all
elements in the population.
Merits and Demerits of Simple Random Sampling

Merits
• An equal and independent chance of selection for each element of the
population into the sample which controls for subjectivity and personal
error.
• It serves as a foundation or basis for all types of random samples.
Demerits
• Biased samples restricting the equal and independent chance of
selection of some elements
• Difficult to ascertain that the sample is representative of the
population.
Systematic Sampling
• Systematic sampling means the selection of the sample from the population
according to some rule or formula.
• The formula being the ratio of population size and the sample size.
• Systematic sampling involves drawing every nth element of the population
from a predetermined list of population.
• If the appropriate unit is unavailable the next is selected
• For example: Let sample size (n) = 200 and population size (N) = 2,000,
selecting each (N/n)th element means selecting every 10th element from the
list and continuing this process until 200 elements are selected is called a
systematic sample.
Merits and Demerits of
Systematic
Merits Sampling
• Systematic sampling is cost-effective
• The resultant sample may be comprehensive and representative of the
population
• It deploys inferential statistics and thus observations of the sample may
be used for generalizations.
Demerits
• systematic sampling requires complete knowledge of the population
and each of its elements
• It fails to ensure representativeness and tends to bear the over-risk in
drawing conclusions from the observations of the sample.
Stratified Random Sampling
• To stratify means to classify or to separate people into groups
according to some characteristics, such as position, rank, income,
education, sex, or ethnic background.
• These separate groupings are referred to as subsets or subgroups.
• For a stratified random sample, the population is divided into groups
or strata.
• A random sample is selected from each stratum based upon the
percentage that each subgroup represents in the population.
Merits and Demerits of
Stratified Sampling Method
Merits
• There is unbiased selection of samples with an equal chance of
selection to all the units of the population.
• It is an objective and a good representative of the population.
• The heterogeneity of the population is converted into homogeneity.
• Observations can be used for inferential purposes.
• Its results are more reliable.
• Since strata are homogenous, there is less chance of standard error.
Demerits
• Its not economical with respect to time, money and energy.
• It poses difficulty in deciding a relevant criterion for stratification since
normally, only one criterion can be used for stratification but
generally, more than one criterion seems to be relevant for
stratification.
• It is difficult to form strata.
• The selected sample may be a representative sample with reference
to the used criterion but not to the other.
Cluster Sampling
• Cluster sampling involves the selection of population elements not
individually but in aggregates/groups.
• With the existence of a large population spread over a large geographical
location, all the other three probability sampling methods would fail to
be cost-effective with respect to time, money, energy and other reasons.
• All the geographical divisions are based on territory, community,
neighbourhood etc.
• Cluster sample converts homogeneous population into heterogeneous.
Merits and Demerits of Cluster
Sampling
Merits
• Cluster sampling is easier, more flexible, practical, highly applicable and
economical to apply to large populations.
• It may be a good representative of the population.
• Elements can be readily substituted within the same random section.
• It is possible to estimate the characteristics of clusters of elements
drawn.
• Observations can be used for inferential purposes i.e. to generalize
research findings from a study sample to the population from which the
sample was obtained.
Demerits
• A lack of surety that each sampling unit included in an area sample
will be of equal size and thus offers little control over the size of each
cluster.
• It is also difficult to ensure that the clusters are mutually exclusive.
• Cluster sampling fails to be free of error and fails to be comprehensive
hence there’s no certainty if it represents the population.
Non-Probability Sampling Methods
• Non-probability sampling based on subjective selection of the sample
elements without ascertaining the equal probability for each element
of the population to be included as the sample element.
• The non-probability samples include convenience sampling,
judgement sampling, quota sampling, double sampling and snowball
sampling
Convenience Sampling
• Convenience sampling refers to the samples that are taken by their
frequent availability or because it is impossible to apply other
sampling methods.
• In convenience sampling sample elements are selected from the
population by virtue of their accessibility or convenience.
• The researcher remains least concerned about including people with
specific traits rather than selecting the available elements.
Merits and Demerits of
Convenience Sampling
Merits
• Fast data collection.
• Inexpensive data collection.
• Easy data collection
Demerits
• It fails to be the representative of population.
• Convenience sampling is biased and not free from error.
• It fails to apply parametric statistics and thus renders limited
generalizability.
Purposive Sampling
• Purposive sampling comprises handpicked samples assuring that
specific elements are included.
• The high degree of selectivity assures the representativeness of all
relevant strata under the study.
• Purposive samples are frequently called judgemental samples
because the judgement to include specific elements of a given
population is exercised by the researcher to seek the information.
Merits and Demerits of
Purposive
Merits
Sampling
• It incorporates cost-effectiveness, accessibility and convenience.
• It involves sufficient care and decision in the selecting elements.
• Purposive sampling assures the selection of specific elements in the sample.
• It employs the application of researcher’s knowledge with reference to the specific sample.
• This sampling plan involves the use of the best available knowledge of the sample subjects
and the control of significant variables.
Demerits
• It is not free from error.
• It fails to assure that the sample is truly random or representative.
• It requires extensive information about the population compared to other sampling
methods.
• Like the other non-probability sampling methods, it fails to use inferential statistics.
Quota Sampling
• Quota sampling is a frequently used non-probability sampling that
involves drawing the proportional elements from various target
population strata.
• It involves combining sample size decisions with particular characteristics
such as age, area, sex, class, etc., which enables the researcher to focus
on the most suitable elements based on these criteria.
• It involves both accessibility and judgement.
• Quota sampling is specific with respect to the size and proportion of
subsamples with subgroups chosen to reflect corresponding proportions
in the population.
Merits and Demerits of Quota
Sampling
Merits
• It is a method that involves ease, frequent usability and cost-
effectiveness as compared to other sampling methods.
• It guarantees the inclusion of members of different sub populations.
• Less data collector error
• Quota sampling is an improvement over the judgement sampling.
Demerits
• Requires adequate resources of more time, money, personnel and effort
• Over representation of population elements in a sample that are not
readily accessible, cooperative, articulate and visible
Respondent Assisted Sampling
• This sampling method involves sampling rare, low incidence and hard
to reach populations and studying social networks.
Merits and Demerits of Respondent
Assisted Sampling
Merits
• It does not have ethical challenges linked to other non probability sampling
methods such as snow ball sampling
Demerits
• Selection bias in the selection of sample elements that is determined by the
subjective choice of previously selected respondents
• It may require more resource requirements of time, money and effort.
• Selected sample elements are not independent of each other as the
selection of a member of the population is dependent on ties to those
previously interviewed.
Factors determining sample size and sampling procedures