Unit- 6

Sampling Design and Procedure

6.1. Some Fundamental Definitions
Population: is the theoretically specified aggregation of survey elements from which the survey sample is actually
selected.
Sampling Frame: is the list of elements from which the sample drawn
Sample: a subset or some part of a larger population.
Sample Design: is a definite plan for obtaining a sample frame.
Sampling: is the process of using a small number or part of a larger population to make conclusion about the whole
population.
Statistics: is a characteristic of a sample.

6.2. Steps Involved in Sample Planning (Sample procedure)

The first thing that the sample plan must include is a definition of the population to be investigated. This involves the
following procedure:

Defining Population

Census vs. Sample

Sample Design

Estimate Cost of Planning

Execute Sampling Procedure

I. Defining the Population

Defining the target population implies specifying the subject of the study. Specification of a population involves identifying
which elements (items) are included, as well as where and when. If the research problem is not properly defined then
defining population will be difficult.

1
II. Census Vs Sample
Once the population has been defined, the researcher must decide whether the survey is to be conducted among all
members of the population or only a subset of the population. That is, a choice must be made between census and sample.
Advantages of Census
 Reliability: data derived through census are highly reliable.
 Census data yield much information.
Limitation of census
 Expensiveness: Investigating each elements of the population is expensive to any individual researcher.
 Excessive time and energy: beside cost factor, census survey takes too long time and consumes too much
energy.
The Need for Sampling
Sampling is a critical issue in survey research; usually the time, money, and effort involved do not permit a researcher to
study all possible members of a population. Furthermore, it is generally not necessary to study all possible cases to
understand the phenomenon under consideration. Sampling comes to our aid by enabling us to study a portion of the
population rather than the entire population.
Since the purpose of drawing a sample from a population is to obtain information concerning that population, it is extremely
important that individuals included in a sample constitute a representative cross section of individuals in the population.
I.e. samples must be representation if one is to be able to generalize with confidence from the sample to the population.
So, randomness is basic to scientific observation and reference.
A type of sample that contains many unrepresentative characteristics would be termed as a biased sample. The findings of
a biased sample cannot legitimately be generalized to the population from which it is taken.
Limitations of sampling
 Less Accuracy: in comparison to census technique the conclusion derived from sample are more liable to error.
Therefore, sampling techniques is less accurate than the census technique.
 Misleading conclusion: if the sample is not carefully selected or if samples are arbitrarily selected, the
conclusion derived from them will become misleading if extended to all population.
 Need for specialized knowledge: the sample technique can be successful only if a competent and able scientist
makes the selection.
Sampling technique is used under the following conditions:
 Vast Data: When the number of units is very large, sampling technique must be used. Because it economize
money, time and effort.
 When utmost accuracy is not required: the sampling technique is very suitable in those situations where 100%
accuracy is not required, otherwise census technique is unavoidable.
 Infinite population: if the population is unlimited, sampling technique is imminent.
 When census is impossible: if we want to know the amount of mineral wealth in a country we cannot dig all
mines to discover and count. Rather we have to use the sampling technique.
 Homogeneity: if all units of the population are alike (similar) sampling technique is easy to use.

2
Essential of an ideal sample
An ideal sample should fulfill the following four basic characteristics
 Representativeness
 Independence
 Adequacy
 Homogeneity
III. Sample Design
Sample design should answer the following:
 What type of sample to use?
 What is the appropriate sample unit?
 What frame (List of Sampling Unit) is available for the population?
IV. Sample Size determination
A researcher is worried about sample size because of the fact that sample size (number of elements in sample)
and precision of the study are directly related. The larger the sample size the higher is the accuracy. The sample
size determination is purely statistical activity, which needs statistical knowledge. There are a number of sample
size determination methods.
Personal Judgment: the personal judgment and subjective decision of the researcher in some cases can be used as
a base to determine the size of the sample.
Traditional inferences: This is based on precision taste and confidence level. To estimate sample size using this
approach we need to have information about the estimated variance of the population, the magnitude of acceptable error
and the confidence interval.
 Variance or heterogeneity of the population: it refers to the standard deviation of the population parameter. The
sample size depends up on the variance of the population.
 If the population is similar (homogenous) small sample size can be enough.
 Magnitude of acceptable error: the magnitude of the error (range of possible error) indicates how precise the
study must be. It is acceptable error for that study. The researcher makes subjective judgment about the
desired magnitude of error.
I. V. Cost of Sampling
The sample plan must take into account the estimated cost of sampling. Such costs are two types, overhead costs and,
variable costs. In reality however, it may be difficult and even for some people not reasonable to separate sampling cost
from over all study cost.
VI. Execution of Sampling Process
The last step in sample planning is the execution of the sample process (procedure). In short the sample is actually chosen
at this stage. The actual requirements for sampling procedure are the following:
Sample must be representative: when it is a representative, a sample will be relatively small pieces of population that
mirror the various patterns and subclass of the population.
Sample must be adequate: a sample is adequate when it is of sufficient in size to provide confidence in the stability of its
characteristics.

3
6.3. Sampling Techniques
They are basically of two types: non-probability sampling and probability sampling.
1. Non- probability sampling
The non-probability sampling is much less complicated and less expensive. It is very convenient in situations when the
sample to be selected is very small and the researcher wants to generalize some idea of the population characteristics or
of the problem using well-informed member in a short time. The non-probability sample may prove perfectly adequate if the
researcher has no desire to generalize his findings beyond the sample, or if the study is merely a trial run for a larger
study.
There are five types of non-probability sampling
1. Convenience sampling: Non-probability samples that are unrestricted are called convenience samples or
accidents sample. In this sample, the investigator chooses the closest live persons as respondents.
2. Quota Sampling: it is non-probability sampling, which is equivalent to stratified sampling with added
requirement that each stratum is generally represented in the sample in the same proportion as in the same
population. When stratified random sampling is impractical to use because of difficulty of getting complete list of
the units in the strata of the population, quota sampling becomes appropriate.
3. Dimensional sampling: it is basically a multi dimensional form of quota sampling. The idea is to specify all
dimensions/ variables of interest in the population to make sure that every combination of these dimensions is
represented by at least one case. This method is designed for studies in which only small sample is desired so
that each case drawn can be studied in more detail than is possible in a large-scale study. It is designed to
overcome the danger of missing important variables in selection of small samples.
4. Purposive Sampling: in purposive or judgmental sampling the investigator doesn’t necessary have a quota to
fill from within various strata as in quota sampling, but neither does he just pick the nearest bodies as in
convenience sampling. Rather the researcher uses his own judgment about which respondent to choose and pick
only those who best meet the purposes of the study.
5. Snow Balling Sampling: the term snowball comes from the analogy of snowball, begins small but becomes
bigger and bigger as it rolls down hill. This method of sampling is conducted in stages. In the first stage a few
persons having the required characteristics are identified and interviewed. The second stage involves
interviewing those persons who again lead to still more persons who can be interviewed in the third stage, and
so on.
II. Probability Sampling
It implies that the probability of selection of each respondent is known. Sampling is important in all research, but it is
particularly critical in surveys. The findings of survey are presented in numerical form, such as percentages, averages,
and measures of variability.
The different sampling methods of probability sampling include:
I. Simple Random Sampling: each member of the population understudy has an equal chance of being selected. The
method involves selecting at random from all list of the population (sample frame) the required number of subjects for the
sample. Because of probability chance, the sample should contain subjects with characteristics similar to the population
as a whole: some old, some young, some tall, some short, some rich, some poor etc.

4
2. Systematic sampling: this method is the modified form of simple random sampling. It involves selecting from
population lists in a systematic random fashion. For example, from a population of 2000 a sample of 200 is required and
then every 10th person can be selected. The starting point for the selection is chosen at random.
3. Stratified Sampling: it involves dividing the population in to homogonous groups, each group containing subjects with
similar characteristics. For example, group A might contain males and group B females. In order to obtain a sample
representative of the whole population in terms of sex, random sampling selection of subjects from group A and group B
must be taken. The needed, the exact proportion of males to females in the whole population can be reflected in the
sample.
4. Cluster sampling:
It is used when the population under study is large, when the distribution of the members is scattered, or when the
selection of individual members is not continent for several reasons. Cluster sampling is used in situations where the
population members are naturally grouped in unit that can be used conveniently as clusters. Example: A researcher is
interested to survey the Math’s achievement of 4th grade students in elementary schools found in Ethiopia. It is practically
impossible to test all 4th grade students in elementary school students in the country by a single researcher.
Since the 4th graders are naturally grouped by regions, Zones, Woredas, schools and classes, the researcher may take
zones or Woredas or schools as clusters. If schools are selected, it will be easier for him to take students as sample. Then,
all students in the selected schools are to be selected. Cluster sampling differs from stratified random sampling in that in
cluster sampling random selection occurs not with the individual members but with the clusters. The clusters for the
sample are randomly selected from the large population of cluster and once a cluster is selected for the sample, all the
population members in that clusters are included in the sample. This is in contrast to stratified sampling in which the
individual members within strata are randomly selected.
5. Stage (multi stage) Sampling
It is a further development of the principle of cluster sampling. It is used in large-scale surveys for a more comprehensive
investigation. The researcher may have to use two, three, or four stage sampling. For example, a researcher wants to
study the opinion of teachers towards self-centered class system. He wants to select a sample from all elementary school
teachers in Ethiopia. A simple random sampling would be impractical and so from the regional state a sample of five
regional states could be selected randomly from Northern, Eastern, southern, Western and central regions. From the five
states chosen, all zones could be listed and random sample of 15 zones selected. Form the 15 zones, 30 Woredas can be
selected randomly and from these Woredas all elementary schools could be listed and a random sample of 100 schools
selected. It wouldn’t be difficult to compile a list of all elementary school teachers working in 100 schools and random
sample of say 650 teachers selected. The successive random sampling of regional states, zones, Woredas schools and
finally teachers constitute a multistage /5stages/. Multi stage sampling is comparatively convenient less time consuming,
and less expensive method of sampling. However, an element sample bias gets introduced because of the unequal size of
some of the selected sub-samples.
6.4. Sampling and non Sampling Error
Sampling studies are subject to sampling and non-sampling errors, which are a random, or of a constant nature. The
errors created due to sampling and at which the average magnitude can be determined are sampling errors, while others
are called biases.
A. Sampling Error:

5
  Is the difference between the result of a sample and the result of census
 Is the difference between the sample estimation and the actual value of the population
 Is an error that is created because of chance only?
Although a sample is properly selected, there will be some difference between the estimate obtained from sample (sample
statistics) and the actual value of population (parameter). The mean of the sample might be different from the population
mean by chance alone. The standard deviation of the sample will also probably be different from the population standard
deviation. We can therefore, expect some difference between the sample statistics and the corresponding population
values known as parameters. This difference is known as the sampling error.
B. Systematic/Non Probability/Sampling Bias
It is a non-probability error, which can be created from errors in the sampling procedures and it cannot be reduced or
eliminated by increasing the samples size. Such errors occur because of human mistakes and not chance. The possible
factors that contribute to the creation of such bias include:
A. Inappropriate Sampling:
If the sample unit is misrepresentation of the universe, it will result in sampling bias. This could happen when a researcher
gathers data from a sample that was drawn from some favored locations. It occurs when there is a failure of all units in
the universe to have some probability of being selected for the sample.
B. Accessibility Bias: in a considerable number of research studies, researchers tend to select respondents who are the
most accessible to them (easily reached). But it should be noted that when all members of the population aren’t equally
accessible, the researcher must provide some mechanisms of controlling so as to ensure the absence of over and under
representation of some respondents.
C. Defective Measuring Devices: in some instances, questions in a questionnaire may not be phrased so that they are
fully understandable by respondents. Consequently, the answers obtained aren’t accurate. Furthermore, on any measuring
device, most individuals are likely to be mismeasured to some degree due to errors in procedures of observation,
interviewing, coding etc.
D. Non -Response Bias
This is an incomplete coverage of a sample or inability to get complete responses from all the individuals initially included
in the sample. This arises due to failure in locating some of the individuals of the population selected for sample or due to
their refusal.
Non- response errors are also due to the respondents not processing /having correct information or due to their giving
deliberately biased responses.
Note that non-sampling errors occur both in a sample survey and in a census; whereas the sampling errors occur only
when a sample survey is conducted. Preparing the survey questionnaire and handling the data carefully can minimize no-
sampling errors.
6.3. Sample Size Determination
The size of the sample is determined by a number of factors such as: the availability of resources (cost, time and
personnel), the nature of the population (degree of heterogeneity or homogeneity), and the degree of accuracy we seek or
magnitude of difference we wish to test, the confidence levels we wish to use, the appropriate size of the population from
which the sample is drawn, and the response rate. In any case, the size of the actual sample must be larger enough. In
deciding the sample size the researcher must note the following points:

6
 The size of the population to be sampled: Even if general rules are hard to make without knowledge of the
specific population, researchers recommend that the sample size of around 30 cases seems to be the minimum
for studies which require statistical data analysis and which don’t involve many sub-division of data.
 The actual number of cases from which the data are collected: in some instances, the data actually gathered
might be substantially fewer because of respondents who cannot be located, refuse to be interviewed or return
legible or not otherwise unusable questionnaires. Furthermore, the actual number of cases varies from question
of question depending on the knowledge of respondents who fail to answer particular question.
 The amount of heterogeneity of the variable within the population: the greater the heterogeneity among
members of the population a variable, the larger the sample required. However, for population in which there is
no variability or heterogeneity on a variable, a sample of any size will be enough regardless of the population size.
 The number of cells the sample may have to be sub-divided data analysis: Every researcher is required to
estimate the number of cells in which the sample may have to be shared during data analysis and ensure an
adequate sample size for each sub-division.
 The degree of accuracy that is needed for the study: if you seek a high degree of accuracy such as (o.o1)
your sample will need to be much larger than if you seek a relatively lower degree of accuracy such as (0.05)