10/6/2024

SAMPLING DESIGN
Once the researcher has clearly specified the problem
and developed the appropriate design and data-
CHAPTER 4 collection instrument, the next step is to select the
elements from which information is collected.
To effectively undertake research project, the researcher
must have the data, without data the analysis may not
SAMPLING be convincing.

Information (data) can be generated from either

DESIGN population or sample.

The basic idea of sampling is that by selecting some of

the elements in the population, we may draw
conclusions about the entire population.

SAMPLING DESIGN Why sample?

• It is the totality of object or phenomena
Population: under consideration for a specific study. Lower cost

• A census is a count of all the elements in a

Census: population. Greater accuracy of
results
• It is a proper subset or part of population. There are several reasons
for sampling;
Sample: • It is used to represent the population. Greater speed of data
collection
• is the procedure of selecting a sample from a
population.
Sampling: • aims at obtaining consistent and unbiased Availability of population
estimates of the population. elements.

SAMPLING CONCEPTS AND TERMINOLOGIES: Steps in sampling design:

• The unit of analysis or case in population, Define the target population:
Sampling • it is from which information is collected which
• Research elements are the objects on which the
element: provides basis for analysis, measurements are taken
• That element or set of elements considered for
Sampling selection in some stage of sampling
unit: Identify the sampling frame:
• It is the listing of the elements from which the actual
• It is the actual list of sampling units from which sample will be drawn
Sampling the sample is selected.
frame: • It is closely related to the population.
Select a sampling procedure:
• The choice of sampling method depends largely on what
Sampling • Size of the sample / size of population. the researcher can develop for a sampling frame.
ratio:

1
 10/6/2024

Steps in sampling design: Characteristics of good sampling:

Determine the sample size: It depends on The logic of the theory of sampling is the logic of induction.
• If the population is identical any sample would be sufficient.
• homogeneity of the population (dispersion variance),
• variation or heterogeneity in the population, more controlled sampling procedures are
• degree of confidence (how much precision they need), required.
• number of sub groups to be studied,
A good sample must be,
• cost and
• time factors. • as representative of the entire population as possible, and
• ideally it must provide the whole of the information about the population from which
the sample has been drawn.
Select the sample elements:
• The researcher needs to choose the elements that will be included in the
So, the ultimate test of a sample design is how well it represents
study. This depends upon the type of sample being used and the sampling the characteristics of the population it purports to represent.
method.
In measurement terms the sample must be valid
Collect the data from the designated elements:
• Appropriate methods of data collection methods are adopted to gather Validity of the sample depends on two factors: accuracy and
information from the sample elements to support the study, which leads
to further analysis and conclusion to the study precision.

Accuracy: Precision:
It is the degree to which bias is absent from the sample.
The degree to which the standard errors are minimized.
When the sample is drawn properly, some sample elements
underestimate the population values being studied and No sample will fully represent its population in all aspects.
others overestimate them.
A sample statistics may be expected to differ from its
parameters as a result of random fluctuations inherent in the
An accurate unbiased sample is one which the sampling process.
underestimations and overestimations are a balanced among
the members of the sample. This is referred to as the error of variances or sampling error.
Precision is measured by the standard errors of estimates,
There is no systematic variance with an accurate sample. a type of standard deviation measurement; the smaller
the standard error of estimates, the higher is the precision
of the sample.
Systematic variance is defined as the variation in measures
due to some known or unknown influences that cause the The ideal sample design produces a small standard error of the
scores to lean in one direction more than another. estimate.

Categories of sampling procedures: Classification of Sampling Methods

Sampling
• There are two categories of sampling procedures: Random Methods
(Probabilistic) and Non-random (Non-Probabilistic).
• Probability sampling: Probability Non-
• Is a sampling technique in which every number of the population Samples probability
will have a known, non-zero or equal probability of selection.
• Non-Probabilistic sampling:
• is a sampling technique in which units of the sample are selected Systematic Stratified Convenience Snowball
on the bases of personal judgment or convenience.

Simple
Cluster Judgment Quota
Random

2
 10/6/2024

Simple random sampling:

• Each population element has a known and equal chance
of being selected.
• Usually two methods are adopted
• The lottery method
• Table of random number:
• The disadvantage
• is too expensive to interview a national face-to face sample
based on simple random method,
• it requires listing of the entire population of interest.

Systematic sampling: Stratified random samples:

A systematic random sampling is a sample which contains
every ‘i' th element of the population.
is distinguished by the following two step
The first element is chosen randomly, the rest procedure;
systematically. • The parent population is divided into mutually exclusive and
It is one of the most widely used probability sampling exhaustive sub sets. …called strata or subpopulations
technique, its major advantages are its simplicity and • The division is mutually exclusive and exhaustive
flexibility, • A simple random sample of elements is chosen independently
it may result in error when the population is not uniform from each group or subset.
or homogenous or where there are systematic particularly useful when we have heterogeneous
arrangements, it is less expensive and easy to carry out.
populations.
Example: from 20 elements if we want to take four, then i
th element = 20/4 = 5th thus every 5th element of the
group of 20 elements after the 1st sample is taken is With stratification each stratum is homogenous
included in our sample. internally and heterogeneous with other strata.

The researcher chooses a stratified

The researcher must still decide
random sample due to the
whether to select;
following reasons;

To increase a sample’s a proportionate stratified

statistical efficiency sample, The sample drawn is
proportionate to the stratum’s
share of the total population.
To provide adequate data for
analyzing the various
subpopulations and a disproportionate stratified
sampling, those strata that are
very homogenous are sampled
To enable different research less
methods and procedures to be
used in different strata.

3
 10/6/2024

Cluster Sampling: Cluster sampling differs from stratified

sampling in several ways
If the total area of interest happens to be big one, a Stratified sampling Cluster sampling
convenient way in which a sample can be taken is to divide  We divide the population into a  We divide the population into
the area into a number of smaller non-overlapping areas and few sub groups, each with many subgroups, each with a
then to randomly select a number of these areas. many elements in it. few elements in it.
 The subgroups are selected  The subgroups are selected
according to some criterion that according to some criterion of
The total area can be divided into groups or clusters of is related to the variables under ease or availability in data
elements and some of the groups or clusters are selected study. collection.
randomly.
 We try to secure homogeneity  We try to secure heterogeneity
within subgroups and within subgroups and
heterogeneity between homogeneity between
subgroups. subgroups.
reduces cost by concentrating surveys in selected areas.  We randomly choose elements  We randomly choose a
from within each subgroup. number of subgroups, which
we then study in depth.

II. Non Probability sampling: Convenience sampling (accidental samples):

The probability of selecting the population elements is
unknown.

We cannot estimate any range within which to expect the Select anyone who There is no control
population parameter. is convenient, it to ensure
can produce, precision.
There are a variety of ways to choose persons or cases to ineffective, highly It is least reliable • Ex: The person on the
include in the sample. unrepresentative but cheap and easy street interviewed for
a television program.
sample and is not to collect.
Despite accepted superiority of probability sampling recommended, it
methods, has high bias and
systematic errors.
non probability sampling maybe used
• when probability sampling is prohibitively expensive and when precise
representation is not necessary.
• It is used because of cost and time requirements
• It is used if there is no desire to generalize a population parameter

Purposive: Judgment sampling

It uses the judgment of experts in selecting cases or it selects cases with specific
purpose in mind.

When one draws a non-probability sample that But the researcher does not know whether the case selected represents the
confirms to a certain criteria, it is called purposive population:
sampling. • Ex: in the study of standard of living, the cost of electricity, refrigerator, video recorder, satellite
dish cannot be included for all people in Ethiopia.

Quota sampling:
It occurs when one picks sample members to conform
to some criteria. Quotas are assigned to different strata group.
The logic behind is that certain relevant characteristics describe the
dimensions of the population.
Purposive judgment
quota a researcher first identifies categories of people then decides how many to get in
sampling can sampling
be of two types; sampling. each category.
and It gives no assurance that the sample is a representative on the variable being
studied. As there is no element of randomization, the extent of sampling error
cannot be estimated.

4
 10/6/2024

Snowball sampling:
Also called network, chain referral or computational
sampling is a method for identifying and sampling or
selecting the case in a network.

It begins with one or a few people or cases and spreads out on

the basis of links to the initial case.

Snowball sampling is a judgment sample used to sample special

populations.
The researcher locates an initial set of respondents.

These respondents are used as informants to identify others

with desired characteristics.

SAMPLE SIZE AND ITS DETERMINATION

In sampling analysis, the most difficult question is: What
should be the size of the sample or how large or small should
be ‘n’?

If the sample size (‘n’) is too small, it may not serve to

achieve the objectives and if it is too large, we may incur
huge cost and waste resources.

As a general rule, one can say that the sample must be of an

optimum size i.e., it should neither be excessively large nor
too small.

Technically, the sample size should be large enough to give

a confidence interval of desired width and as such the size
of the sample must be chosen by some logical process
before sample is taken from the universe.

DETERMINATION OF SAMPLE SIZE THROUGH THE APPROACH BASED

Sample size determination: ON PRECISION RATE AND CONFIDENCE LEVEL
• There are two alternative approaches for determining the size • whenever a sample study is made, there arises some sampling
of the sample.
• The first approach is “to specify the precision of estimation desired
error which can be controlled by selecting a sample of adequate
and then to determine the sample size necessary to insure it” and size.
• the second approach “uses Bayesian statistics to weigh the cost of • Researcher will have to specify the precision that he wants in
additional information against the expected value of the additional
information.” respect of his estimates concerning the population parameters.
• The first approach is capable of giving a mathematical solution, • For instance,
and as such is a frequently used technique of determining ‘n’. • a researcher may like to estimate the mean of the universe within 3 of
• The limitation of this technique is that it does not analyze the the true mean with 95% confidence.
cost of gathering information vis-a-vis the expected value of • In this case we will say that the desired precision is 3 , i.e., if the sample
information. mean is 100, the true value of the mean will be no less than 97 and no
more than 103.
• The second approach is theoretically optimal, but it is seldom • In other words, all this means that the acceptable error, e, is equal to 3.
used because of the difficulty involved in measuring the value
of information. Hence, we shall mainly concentrate here on • Keeping this in view, we can now explain the determination of
the first approach sample size so that specified precision is ensured.

5
 10/6/2024

(a) Sample size when estimating a mean: • Suppose, we have p 4.8 for our purpose
• If the difference between µ and x̄ or the acceptable error
• The confidence interval for the universe mean,is gi ven by
is to be kept with in 3 of the sample mean with 95%
• x̄ ± ܼ(Gp/√݊) confidence, then we can express the acceptable error, ‘e’
• Where as equal to
•x̄ = sample mean
• z = the value of the standard variate at a given confidence
level (to be read from the table giving the areas under
normal curve as shown in appendix) and it is 1.96 for a
95% confidence level;
• n = size of the sample;
• Gp =standard deviation of the population (to be estimated
from past experience or on the basis of a trial sample).

• In a general way, if we want to estimate in a population with Illustration

standard deviation p with an error no greater than ‘e’ by • Determine the size of the sample for estimating the true weight
calculating a confidence interval with confidence corresponding of the cereal containers for the universe with N = 5000 on the
to z, the necessary sample size, n, equals as under: basis of the following information:
௭ మ ∗G௣ మ 1. the variance of weight = 4 ounces on the basis of past records.
• n= 2. estimate should be within 0.8 ounces of the true average weight with
௘మ
99% probability.
• All this is applicable when the population happens to be infinite.
• Will there be a change in the size of the sample if we assume
• in case of finite population, the above stated formula for infinite population in the given case?
determining sample size will become • If so, explain by how much?
௭ మ ∗N.Gpమ
• n=
ேିଵ ௘ మ ା௭ మ ∗Gpమ
• where
• N = size of population
• n = size of sample
• e = acceptable error (the precision)
• Gp = standard deviation of population
• z = standard variate at a given confidence level.

(b) Sample size when estimating a percentage or proportion:

• Since ṕ is actually what we are trying to estimate, then
• If we are to find the sample size for estimating a what value we should assign to it ? One method may be
proportion, our reasoning remains similar to what we
have said in the context of estimating the mean. to take the value of p = 0.5 in which case ‘n’ will be the
• First of all, we shall have to specify the precision and maximum and the sample will yield at least the desired
the confidence level and then we will work out the precision.
sample size as under: • This will be the most conservative sample size.
• Since the confidence interval for universe proportion ṕ
is given by • The other method may be to take an initial estimate of
• ‫ ݌‬± ‫√ݖ‬
௣∗௤ p which may either be based on personal judgement or
௡ may be the result of a pilot study.
• where
• p = sample proportion, q = 1 – p; • In this context it has been suggested that a pilot study
• z = the value of the standard variate at a given confidence of something like 225 or more items may result in a
level and to be worked out from table showing area under reasonable approximation of p value.
Normal Curve;
• n = size of sample.

6
 10/6/2024

• Then with the given precision rate, the acceptable • The formula gives the size of the sample in case of
error, ‘e’, can be expressed as under: infinite population when we are to estimate the
proportion in the universe.
• But in case of finite population the above stated
formula will be changed as under:

Illustration Illustration
• What should be the size of the sample if a simple random sample • Suppose a certain hotel management is interested in
from a population of 4000 items is to be drawn to estimate the determining the percentage of the hotel’s guests who stay
percent defective within 2 per cent of the true value with 95.5 per for more than 3 days. The reservation manager wants to
cent probability? What would be the size of the sample if the
population is assumed to be infinite in the given case? be 95 per cent confident that the percentage has been
estimated to be within 3% of the true value. What is the
• As we have not been given the p value being the proportion of
defectives in the universe, let us assume it to be p = .02 most conservative sample size needed for this problem?
• Solution: We have been given the following: Population is
infinite;
௭ మ .௣.௤
n= ௘మ

ଵ.ଽ଺మ .(.ହ).(ଵି.ହ) .ଽ଺଴ସ

• n= ଴ଷమ
= .଴଴ଽ
= 1067.11 ≈ 1067