SAMPLING

By,
Gaurav Gandhi
Siddharth Tewari
Dharamveer
Rashmi Rajan
The Concept

Sampling
A process of selecting a few (a sample) from a
bigger group (the sampling population) to
become the basis for estimating or predicting the
prevalence of an unknown piece of information,
situation or outcome regarding the bigger group.
Sample
A subgroup of the population we are interested
in.
Pros & Cons

 Advantage − Saves time as well as

financial and human resources.
 Disadvantage − Information about the
population’s characteristics of interest to
the researcher is not found/determined.
Instead, it is estimated or predicted.
Sampling Terminology
 Population or Study Population (N)
 Sample
 Sample Size (n)
 Sampling Design or Strategy
 Sampling Unit or Sampling Element
 Sampling Frame
 Sample Statistics
 Population Parameters
 Saturation Point
 Principles of Sampling
1. In a majority of cases of sampling there will be a
difference between the sample statistics and the true
population mean, which is attributable to the
selection of the units in the sample.
2. The greater the sample size, the greater will be the
estimate of the true population mean.
3. The greater the difference in the variable under
study in a population for a given sample size, the
greater will be the difference between the sample
statistics and the true population mean.
Factors Affecting Inferences Drawn
from a Sample
 Sample Size
 Larger is the sample size, more accurate will
be the findings
 Extent of Variation in the Sampling
Population
Higher the variation with respect to the
characteristics under study in the study
population, the greater will be the uncertainty for
a given sample
Aims in Selecting a Sample
 Achieve maximum precision in your estimates
within a given sample size
 Avoid bias in the selection of your sample
Bias in selection can occur if:
 Sampling is done by a non-random method
 Sampling Frame does not cover the sampling
population accurately & completely
 A section of a sampling population is impossible
to find or refuses to cooperate.
Types of Sampling
 Random/Probability Sampling Designs

 Non-Random/Non-Probability Sampling
Designs
 ‘Mixed’ Sampling Designs
Random/Probability Sampling
Designs
 When each element has equal and
independent chance of selection.
 ‘Equal’- when probability of selection of each
element in the population is the same.
 ‘Independent’- choice of one element is not
influenced by the choice of another.
Random Sampling-Methods
 Fishbowl Draw- Each element is numbered and selected
from a box/fishbowl. Useful when population is small.
 Computer Program- When sample is selected by using
computer programs.
Random Sampling-Methods
 Table of Random Numbers- It’s a table of
randomly generated numbers that is used to
select a sample.
 Procedure:
STEP 1: Identify the total number of elements in the
study population.
STEP 2: Number each element starting from one.
Random Sampling-Methods
 Procedure(continued):
STEP 3: If the table occupies more than a page
the starting page should be chosen
randomly.

STEP 4: Corresponding to the number of digits

to which the total population
runs, select the same
number,randomly, of columns or rows of
digits from the table.
Random Sampling-Methods
 Procedure (continued):
STEP 5: Decide the sample size.
STEP 6: Select the required number of elements
for your sample from the table.
Selecting a Random Sampling
 There are two ways to select a random
sample:

a) Sampling without replacement

b) Sampling with replacement

Random Sampling Designs
 Simple Random Sampling (SRS)- Most
common method and in sync with randomization
definition
 Procedure for SRS-
STEP 1: Identify by a number all elements or sampling
units in the population.
STEP 2: Decide on the sample size(n).
STEP 3: Select (n) by fishbowl draw, computer program
etc.
Random Sampling Designs
 Stratified Random Sampling – Aims at
reducing the heterogeneity in the population
so that the degree of accuracy in the estimate
is increased.
Hence the population is stratified in a way
that each stratum is homogeneous with
respect to the characteristic that is being
considered.
Random Sampling Designs
 Remember that this characteristic must be:
1. Easily Identifiable. E.g. gender rather than attitude.
2. Related to the main variable.
 After separating the population into non-
overlapping segments the required number of
elements can be chosen using simple random
sampling techniques.
Random Sampling Designs
Types of Stratified Sampling:
1. Proportionate Sampling- Here the
number of selected elements from each
stratum are in relation to its proportion in the
total population.
2. Disproportionate Sampling- No
consideration is given to the size of the
stratum.
Random Sampling Designs
 Cluster Sampling- Used when the total
population is large say a town or a state or an
entire country.
 Here the sampling population is first divided into
groups called clusters and then elements from
each cluster are selected using the simple
random sampling (SRS) technique.
Disproportionate stratified Proportionate stratified
sampling sampling
Step 7 Determine the number Step 7 Determine the proportion of each
number of elements to stratrum in the study population(p).
be selected from = elements(#) in each stratrum
stratrum. total population size
=Sample size(n) Step 8 Determine the number of elements to
no. of strata(k) be selected from each stratrum=(Sample
size) X (p)
Step 8 Select the required number of
elements from each stratrum with SRS
technique Step 9 Select the required number of
elements from each stratrum with SRS
technique

As this method does not take te size of the As the sample is in proportion to the size of
stratrum into consideration in the selection of each stratrum in the population, this method
the sample, it is called disproportionate is called proportionate stratified sampling.
stratified sampling.
Non-Random/Non-Probability
Sampling Designs
 When the number of elements in a population
is either unknown or cannot be individually
identified.
 Selection of elements is dependent upon
other considerations.
Non-Random Designs
 Quota Sampling
 Accidental Sampling
 Judgemental or Purposive Sampling
 Snowball Sampling
Quota Sampling and Accidental
Sampling
Quota Sampling:
• Easily accessible location is selected to study the
Sample Population and for the convenience the
researcher is also guided by the characteristics, like
gender, race, education, religion etc.
Accidental Sampling
• Quota Sampling include people possessing visible
characteristics whereas accidental sampling makes
no such attempt.
Judgmental or Purposive Sampling
• When the research is based upon the
judgment of the researcher as who can
provide the best information to achieve the
objectives of the research.
Snowball Sampling
 When samples are selected using the help of
samples selected previously. People in the
sample are asked to identify other people
who have the required characteristics or
knowledge.
• Researcher have one or more predefined
groups they are seeking.
• The choice of entire sample rests upon the
choice of individuals at the first stage.
‘Mixed’ Sampling Designs
Systematic Sampling Design
Possesses characteristics of both random & non-random
sampling designs

Procedure for selecting a Systematic Sample:

Sample
1. Prepare a list of all the elements in the study population (N).
2. Decide on the sample size (n)
3. Determine the width of the interval (k) = total population
sample size
4. Using the Simple Random Sampling (SRS) technique, select
an element from the first interval (nth order)
5. Select the same order element from each subsequent interval .
Considerations…
 From the first interval, the choice of an element is on a
random basis; but the choice of element from the
subsequent intervals is dependent on the choice from the
first and hence, cannot be classified as random sample
 Although the above defined procedure fro selecting a
sample by systematic sampling technique is described as
above, one can deviate from it by selecting a different
element from each interval with the Simple Random
sampling (SRS) technique.
 To select a random sample one must have a sampling
frames. E.g. records of client in an agency,enrolment lists
of student , electoral list of people living in an area.
 Systematic Sampling
Sampling Frame

1 14 26 39 3
2 15 27 40
3 28 8
4 29
5
16
30
41 13
17 42
18 43 18
19 44
6 20 31 45 23
7 32
8 33 28
9 21 34 46
10 22 35 47 33
23 48
24 49 38
25 50
11 36 43
12 37
13 38 48
Calculation of Sample Size

• At what level of confidence do you want to test your

results, findings or hypotheses?
• With what degree of accuracy do you wish to estimate
the population parameters?
• What is the estimated level of variation (standard
deviation), with respect to the main variable you are
studying, in the study population?

 The size of the sample is important for testing a

hypothesis.
 General rule is the larger the sample size, the more
accurate will be your estimates.
Example
 Find out the average age of the students with in an accuracy
0.5 of a year find the average age within half a year of
accuracy at 95 per cent confidence level?
 The formula for determining the confidence limits is
y=x`  (t 0.05) 

Where
y=estimated value of the population mean
x’= average age calculated from the sample
t 0.05 = value of t at 95 per cent confidence level
 = standard error

σ = standard deviation
η =sample size
 Contd..
If we decide to tolerate an error of ½ years, that means:
y=x`  (t 0.05) 

x`  0.5
in other words (t 0.05)  = 0.5

=1.96 x 
0.5 Sam pli ng

By,
D hara m veer
Gaur avG andhi

Si ddhart hT ewar i
R ashm i R aj an

Let us assume that  is 1 year

=1.96 x 1 = 3.92
0.5
=15.37,say,16
95% accuracy & ½ year of error sample of 16 students is
necessary
If 99 per cent confident about the estimated age, tolerating an error
of ½ year
= 2.576 x 1 = 5.15
0.5
 =27

95 per cent accuracy = 16 students

99 percent accuracy = 27 students

Golden Rule:
“the greater the sample size, the more accurately your findings will
reflect the ‘true’ picture”.
THANK YOU