Course Objective (CO2): The student will be Implications Of Sample Design

able to describe and choose appropriate A sample design has following implications:
sampling design and will also be able to 1. It gives a definite plan for obtaining a
estimate appropriate sample size. sample
2. It gives a technique that the researcher
Module 2: Sampling Design - Census and would adopt in selecting sample
Sample Survey, Implication of Sample Design, 3. It lay down the number (size) of sample
Steps in Sample Design, Sampling Procedure, required for the research
Different Types of Sampling Designs, How to
select a random sample, Random sample from Steps Of Sample Design
infinite universe, Sample size calculation. A sampling design is a term widely used in
research studies that provide an outline on the
Reference Book, Publisher, Edition, Page No. basis of which research is conducted. It tells
Business Research Methods – Research which category is to be surveyed, what should
Methodologies, Methods and Techniques –2nd be the sample size and how the respondents
Edition C R Kothari, New Age International (P) should be chosen out of the population.
Ltd Publisher, ISBN – 978-81-224-2488-1 A sampling design consists of following key
Chapter 4 elements:

Census & Sample Survey 1. Defining the universe - The universe

Basis Census Survey Sample Survey consists of all survey elements that
A complete A survey of qualify for inclusion in the research
survey of all some items study. The universe may be individuals,
groups of people, organizations, or even
the items in the taken from the objects. For example, research about
Meaning population/ population/ voting in an upcoming election would
universe is universe is have a universe comprising all the
called as called as voters of the region.
census survey. sample survey.
2. Selecting the sampling frame – A
Less but
sampling frame is the list of all the
acceptable elements (also referred to as units) from
Accuracy of Highest of the
accuracy as which the sample may be drawn for
survey accuracy.
compared to collecting the data. For example, the list
census survey. of all the voters of an upcoming election
Time, money would be treated as the sampling frame
for the research study. The sampling
and effort More Less
frame has to be selected from the
involved universe of the research study.
Survey of
Population costumers of a 3. Selecting the sampling unit – The
Examples sampling unit is the single element or
count particular
product the group of elements that will be used
to collect the data for the research study.
For example, for a survey research on

Compiled By Prof. Rohit Turani Page 1 of 7

Mail Id - [email protected]
 the upcoming elections every 25 th voter Sampling Procedure
mentioned in the sampling frame would
be considered for the purpose of data 1. Sampling procedure is the procedure of
collection. deciding upon the techniques to be used
in selecting the items for the sample.
4. Determining the sample size - Sample 2. Sampling procedure involves choosing
size refers to the number of participants one of the available sampling techniques
or observations included in a study. This
viz., probability sampling technique or
number is usually represented by n.
non-probability sampling technique
The basic formulae to calculate the 3. Sampling procedure involves calculating
sample size is: the sample size from the universe
available

Different Types Of Sampling Designs

Non-
Where, n = size of sample, N = size of Basis For Probability
Probability
population, z = standard variate at a Comparison Sampling
Sampling
given confidence level, 𝛔p = standard Probability Nonprobability
deviation of the population, e = sampling is a sampling is a
acceptable error (the precision). sampling method of
technique, in sampling
which the wherein, it is
subjects of the not known that
Meaning
population get which
an equal individual
opportunity to from the
be selected as a population will
representative be selected as a
sample. sample.
Alternately Random Non-random
known as sampling sampling
Basis of
selection of Randomly Deliberately
samples
5. Determine the type of sampling design – In non-
probability
In probability
sampling the
sampling the
researcher
Method of units are chosen
purposively
sample randomly
choose the
selection without any
particular units
purposeful
from the
effort.
available
population
Chance of All the units of All the units of

Compiled By Prof. Rohit Turani Page 2 of 7

Mail Id - [email protected]
 sample to the population the population Convenience Sampling
get selected have an equal relies on the Advantages Disadvantages
chance to be a subjective Not very
part of the judgment of Simple and easy to
sophisticated or
sample the researcher execute
to get selected scientific method
as a sample Respondent may
Less time
Researcher Researcher’s not necessarily be
Supremacy consuming
cannot choose choice representative
of
the sample, concerning the Chances of
researcher in
since the selection of the
sample A large number of biasedness exist
samples are items remains
selection responses may since personal
taken randomly. supreme.
Personal Personal quickly be element has a
element has no element has a obtained chance of entering
Involvement
chance of great chance of into the sample.
of personal
entering into the entering into
element
selection of the the selection of 2. Judgment Sampling – Judgment or
sample the sample. purposive sampling is a non-probability
sampling technique in which an
Various types of Non-Probability & Probability
individual selects the sample based on
Sampling Techniques
his or her judgment about some
Types of non-probability sampling are as appropriate characteristic required of the
follows: sample members. For example, the
consumer price index or the inflation
1. Convenience sampling – Convenience rate is based on judgmental sampling.
sampling (also called as haphazard or Judgment Sampling
accidental sampling) refers to sampling Advantages Disadvantages
by obtaining units or people who are Entire
most conveniently available. For representation of
example, before elections, television Simple and easy to
population in the
channels often present person-on-the- execute
sample is not
street interviews that are presumed to possible
reflect public opinion. Respondent may
Less time
not necessarily be
Another example of convenient consuming
representative
sampling is when many internet surveys
A large number of
are conducted with volunteer The judgment of the
responses may
respondents who, either internationally researcher remains
quickly be
or unintentionally visit the supreme
obtained
organization’s website.
Researcher has
control over

Compiled By Prof. Rohit Turani Page 3 of 7

Mail Id - [email protected]
 selection the description of the sample needed. So
procedure of this referral technique goes on,
sample increasing the size of population like a
snowball. For example, a sports drink
3. Quota Sampling- When the responses manufacturer is considering the
are to be obtained from the sample marketing of an energy drink specially
existing the various subgroups in a formulated for athletes, it would be
population, the technique of quota easier to obtain the names/referrals of
sampling is to be used. For example, an the fellow athletes from the responding
interviewer in a particular city may be athlete.
assigned 100 interviews, 20 of which are Snowball Sampling
with small business owners, 18 with Advantages Disadvantages
professionals, 30 with managerial Reduced sample
employees, and rest with hourly Bias is likely to
size and cost is a
employees. enter into the study
clear advantage
Quota Sampling Two different
Advantages Disadvantages samples may give
Selections and highly similar
Entire
placing the Saves a lot of time responses, since the
representation of
samples in the individual may be
population in the
quotas may be well known to each
sample is possible
difficult other
Deciding the Easy technique to
Gives a systematic number of units in reach the targeted
framework to individual sample
obtain samples subgroups may
lead to error Types of probability sampling are as follows:
This sampling
technique ensures 1. Simple random sampling – Simple
random sampling assures that each
that various
element in the population has an equal
subgroups in the chance of being included in the sample.
population are For example, in the method the samples
represented are drawn randomly from a fish bowl.
Simple Random Sampling
4. Referral /Snowball Sampling - This Advantages Disadvantages
technique is used in the situations where Applicability of this
the population is completely unknown Easiest method type of technique is
and rare. Therefore we will take the help limited
from the first element which we select May be termed as a
for the population and ask him to Saves a lot of time
not so scientific
recommend other elements who will fit

Compiled By Prof. Rohit Turani Page 4 of 7

Mail Id - [email protected]
 way of sampling 3. Stratified Sampling - This technique
No chance for divides the elements of the population
personal element into small subgroups (strata) based on
the similarity in such a way that the
to entre in the
elements within the group are
sample homogeneous and heterogeneous among
All the units in the the other subgroups formed. And then
population have the elements are randomly selected from
equal chance of each of these strata. We need to have
being selected as a prior information about the population
to create subgroups.
sample
Stratified Sampling
2. Systematic Sampling – In this sampling Advantages Disadvantages
procedure an initial starting point is In this technique
selected by a random process, and then the chances of
every nth number on the list is selected. Applicability of this
sampling being
For example, to take a sample of type of technique is
able to accurately
consumers from a telephone directory limited
reflect the
every 23rd name might be selected as a
sample. population is high
Systematic Sampling Random sampling The process is time
Advantages Disadvantages error will be consuming as
Extremely simple Applicability of this reduced with the compared other
method of type of technique is use of stratified sampling
sampling limited sampling techniques
The problem of
4. Cluster Sampling - The population is
periodicity may
first split into groups. The overall
It results into occur if the list has
sample consists of every member from
perfectly section a systematic
some of the groups. The groups are
of samples pattern, for example
selected at random.
randomly taking the sales
figure as a sample For Example - An airline company
for every Monday. wants to survey its customers one day,
No chance for so they randomly select 5 flights that
personal element day and survey every passenger on those
to entre in the flights
sample
All the units in the Cluster sampling technique is used when
population have samples are available in form of a
equal chance of cluster or group sharing some common
being selected as a characteristics.
sample

Compiled By Prof. Rohit Turani Page 5 of 7

Mail Id - [email protected]
 Cluster Sampling 2670 7483 3408 2762 3563 1089 6913 7691
Advantages Disadvantages 560 5246 1112 6107 6008 8125 4233 8776

This is an 2754 9143 1405 9025 7002 6111 8816 6446

economical way of The problem may

It is important that the starting point in the table
sampling while rise if all the
of random numbers be selected in some random
retaining the elements in a
fashion so that every units has an equal chance
characteristics of cluster are closely
of being selected.
probability similar
One may question, and quite rightly, as to how
sampling
it is ensured that these digits are random. It may
This kind of
be pointed out that the digits in the table were
The tedious effort sampling technique
chosen haphazardly but the real guarantee of
of surveying all the is not useful if the
their randomness lies in practical tests. Tippett’s
clusters available is samples are spread
numbers have been subjected to numerous tests
eliminated out or not available
and used in many purposes. An example to
in clusters
illustrate how Tippett’s table of random
How To Select A Random Sample? numbers may be used is given below. Suppose
1. A random sample for a finite population we have to select 20 items out of 6,000. The
can be selected by writing each of the procedure is to number all the 6,000 items from
possible samples on a slip of a paper, 1 to 6,000. A page from Tippett’s table may
mixing these slips thoroughly in a then be consulted and the first twenty numbers
container and then draw as a lottery. up to 6,000 are noted down. Items bearing those
2. For example, if we want to select numbers will be included in the sample. Making
random 03 samples out of 06, the use of the portion of the table the required
probability of drawing one element in numbers are noted as:
the first draw is 3/6, the probability of 2952 3992 5911 3170 5624 4167
1545 1396 5356 1300 2693 2370
drawing one more element in second
3408 2762 3563 1089 560 5246
draw is 2/5, for third it will be 1/4. 1112 4233
Hence the joint probability will be the The items which bear the above numbers
product of their individual probabilities, constitute the sample.
i.e. 3/6 * 2/5 * 1/4 = 1/20.
3. Another method to select a random Random sample from infinite universe
sample is through using Tippett’s tables 1. If the sample is infinite, it wouldn’t be
of random number. possible to number the items, hence in
Trippett’s table of random numbers is such situation; the required number of
most popularly used in practice. We samples must be selected haphazardly.
give below the first forty sets from 2. For example, if we want to find out the
Tippett’s table as an illustration of the mean height of the trees in a forest, it
general appearance of random numbers: wouldn’t be possible to number the trees
2952 6641 3992 9792 7969 5911 3170 5624 and hence the required number of
4167 9524 1545 1396 7203 5356 1300 2693 sample must be selected haphazardly.

Compiled By Prof. Rohit Turani Page 6 of 7

Mail Id - [email protected]
 3. Suppose we consider the 20 throws of a 2. The formula to determine the sample
fair dice as a sample from the size for proportion or percentage of
hypothetically infinite population which population:
consists of the results of all possible n = pq/
throws of the dice. Where, n = size of the sample
4. If the probability of getting a particular Z = standardized value corresponding to a
number, say 1, is the same for each confidence level
throw for all the 20 throws, 20 times, p = proportion or percentage of success
then we say that the sample is random. q = p – 1 = proportion or percentage of failure
5. Similarly, it would be said to be E = acceptable magnitude of error
sampling from infinite population if we
sample with replacement from a finite
population and our sample would be
considered as a random sample if in
each draw all the elements of the
population have the same probability of
being selected and successive draws
happen to be independent.
6. In brief, once can say that the selection
of each item in a random sample from
an infinite population is controlled by
the same probabilities and that
successive selections are independent of
one another.
Sample size calculation
Basically there are 02 formulae to determine
the size of the sample:
1. The formula to determine the sample
size for questions involving mean is
given below:
n=
Where, n = size of the sample
Z = standardized value corresponding to a
confidence level
𝛔 = sample standard deviation or an estimated
standard deviation of the population
e = acceptable magnitude of error

Compiled By Prof. Rohit Turani Page 7 of 7

Mail Id - [email protected]