Chapter-8 (Book-2)

1
Sampling Methods

2
 Sampling Technique
The sample selection procedure basically
originated the concept of sampling
technique in statistics.

Usually, populations are so large that a

researcher cannot examine the entire group.

3
 Definition of Sampling
Technique

Sampling is the technique or process of selecting

a small portion (called sample) from the
population

4
 Some Statistical Terms
Sampling Frame: The list of all sampling units with proper
identification.
✓ Student’s ID registered in DU.
✓ List of bus drivers in Dhaka.
✓ List of fish processing workers in Khulna.

Sampling Unit: A particular unit from the sample is known

as a sampling unit.
✓ Students from each department of DU.
✓ Bus drivers from each bus station in Dhaka.
✓ Fish processing workers from each fish hatchery of
Khulna.

5
Types of Sampling
Technique

Sampling

Non-
Probability
Probability

6
 HW

Difference between probability and non-probability

sampling.
 Simple Random
Sampling
✓ Each element has an equal probability of being
selected.

✓ It is a basic type of sampling.

✓ It requires a complete sampling frame and a

homogeneous set of units. (Criteria)

✓ Tools like the lottery method or random number

generator are applied.
8
 Situation of using SRS
When population have a complete sampling frame, units
are homogeneous and it ensures to share equal chance to
get in sampling for each draw.

We can use SRS in that situation.

9
Simple Random Sampling (SRS)
For example,

You want to select a simple

random sample of 04
persons from a household.
You assign a number to
every household member
from 1 to 14 and use a
lottery method to select 04
numbers.
10
 Stratified Sampling
✓ A way where the population is divided into mutually
exclusive subgroups (called strata) and then the
selection of elements from each subgroup is
accomplished by using random sampling.

✓ The sample size from each stratum is determined by

allocating rules (e.g., arbitrary, equal, proportional).

11
 Stratified Sampling
✓ The process of creating subgroups is based on a
relevant characteristic (e.g., gender, age range, income
bracket, job role).

✓ The elements of strata must be homogenous internally

but heterogeneous externally.

✓ It requires a complete sampling frame and a

heterogeneous set of units (Criteria).

12
 Stratified Sampling
For example,

The shop has 12 female employees

and 9 male employees. You want
to ensure that the sample reflects
the gender balance of the shop, so
you sort the population into two
strata based on gender. Then you
use random sampling on each
group, selecting 04 [(7÷21)×12]
female and 3 male by proportional
allocation rule, which gives you a
representative sample of 07
people.
13
 Real Life Problem
✓ Wish to calculate the number of students in secondary
school in Dhaka city.

✓ The school size is differed in terms of students’ number

(e.g., small, middle, and large)

✓ Large school contains greater quantity of students.

✓ If SRS is applied here, it is very possible to create the

sample with too many of students from the large school.
14
 Real Life Problem

✓ The sample will not be representative of the

population (using SRS).

What is the solution????

15
 Situation of using Stratified
Sampling
✓ Divide (stratify) the schools (sampling units) before
the sampling based on the size of the school (small,
middle, and large).
✓ Select sample from each of the three groups by using
SRS.
✓ Calculate total number of students by combining
results of the three groups.
16
 Sample Size Determination

Sample size determination is choosing the number of

observations to make a representative sample to minimize
the sampling error.

17
 Formula to Find Sample Size
𝑧 2 𝑝𝑞
𝑛0 =
𝑑2
• When, population (N) is large (N>2000)
𝑛0
𝑛= 𝑛 −1 (for finite population)
1+ 0𝑁

• When, population (N) is large (N≤2000)

𝑁𝑧 2 𝑝𝑞
𝑛= (for finite population)
𝑁𝑑 2 +𝑧 2 𝑝𝑞

18
 Formula to Find Sample Size
Here,

p=0.5
q=0.5
z=1.96
d= 0.01 to 0.9
N= Population size

19
 Mathematical Problem (Sample
Size Determination)
Suppose treasurer Madam wants to distribute 5000 tk to
some of the students from section B of each. Suppose
section B has 150 students.

1. How many students will we choose?

2. Use appropriate sampling techniques to select the

sample from the population.

20
 Solution
1. How many students will we choose?

Answer: Here, the population size is 150. So we have to

use the following formula:

𝑁𝑧 2 𝑝𝑞
𝑛=
𝑁𝑑2 + 𝑧 2 𝑝𝑞

Given that N=150, z=1.96, p=0.5, q=0.5, and d=0.1

21
 Solution
150 × 1.962 × 0.5 × 0.5
𝑛=
(150 × 0.12 ) + (1.962 × 0.5 × 0.5)
=58.55

≅ 59

22
 Solution
2. Use appropriate sampling techniques to select the sample
from the population.
Answer: To choose 59 students, we can use simple random
sampling because every person has an equal probability of
being selected in the sample. We can construct the sample
by using any approach from the below:
• Using a random number table.
• Reminder method.
• Lottery method.
23
 Mathematical Problem (Sample
Size Determination)
Suppose treasurer Madam wants to distribute 5000 tk to some of
the students from section B of each. Suppose section B has 150
students. Consider out of 150 students, 60 students are from the
engineering department, 40 are from CBA, 30 are from BATHM,
and 20 are from other departments.
1. Select optimum sample size.
2. Use appropriate sampling techniques to select the sample
from the population.
24
 Solution
1. Select optimum sample size.

Answer: Here, the population size is 150. So we have to

use the following formula:

𝑁𝑧 2 𝑝𝑞
𝑛=
𝑁𝑑2 + 𝑧 2 𝑝𝑞

Given that N=150, z=1.96, p=0.5, q=0.5, and d=0.1

25
 Solution
150 × 1.962 × 0.5 × 0.5
𝑛=
(150 × 0.12 ) + (1.962 × 0.5 × 0.5)
=58.55

≅ 59

26
 Solution
2. Use appropriate sampling techniques to select the
sample from the population.

Answer: Here, all students do not have an equal

probability of being selected in the sample. Therefore,
to choose 59 students, we must use stratified random
sampling.

27
 Solution
For engineering,
𝑁1 = 60, 𝑁 = 150, 𝑛 = 59

𝑁1
𝑛1 = ×𝑛
𝑁
60
= × 59 = 23.6 ≅ 23
150
For CBA,
𝑁1 = 40, 𝑁 = 150, 𝑛 = 59

𝑁1
𝑛1 = ×𝑛
𝑁
40
= × 59 = 15.73 ≅ 16
150
28
 Solution
For BATHM,
𝑁1 = 30, 𝑁 = 150, 𝑛 = 59

𝑁1
𝑛1 = ×𝑛
𝑁
30
= × 59 = 11.8 ≅ 12
150
For others,
𝑁1 = 20, 𝑁 = 150, 𝑛 = 59

𝑁1
𝑛1 = ×𝑛
𝑁
20
= × 59 = 7.86 ≅ 8
150
29
 Solution
Engineeri CBA BATHM Others Total
ng
𝑁1 =60 𝑁2 =40 𝑁3 =30 𝑁4 =20 𝑁=150
𝑛1 =23 𝑛2 =16 𝑛3 =12 𝑛4 =8 𝑛=59
Now, we can use simple random sampling to select persons from each
group. In a simple random sample, there are three techniques:
1. Using a random number table.
2. Reminder method.
3. Lottery method.
Here, the whole or total process is known as stratified random
sampling.
30
 Mathematical Problem (Sample
Size Determination)
Consider the following scenario: VC sir wishes to award a
scholarship of 5000 tk to a few students from section B.
Consider that section B has 5000 students. Suppose, 3000
students are Muslim, 1500 students are Hindu, and 500 are
other religions.
1. Determine the optimum number of samples.
2. Use appropriate sampling techniques to select the
sample from the population.
31
 Solution
1. Determine the optimum number of samples.

Answer: Here, the population size is 5000. So we have

to use the following formula:

𝑧 2 𝑝𝑞
𝑛0 =
𝑑2
Given that N=5000, z=1.96, p=0.5, q=0.5, and d=0.03

32
 Solution
1.962 × 0.5 × 0.5
𝑛=
0.032
=1067.11

≅ 1067

1067
Now, 𝑛 = 1067−1 = 879.42 ≅ 880
1+ 5000

33
 Solution
2. Use appropriate sampling techniques to select the
sample from the population.

Answer: Here, all students do not have an equal

probability of being selected in the sample. Therefore,
to choose 880 students, we must use stratified random
sampling.

34
 Solution
For Muslim,
𝑁1 = 3000, 𝑁 = 5000, 𝑛 = 880

𝑁1
𝑛1 = ×𝑛
𝑁
3000
= × 880 = 528
5000
For Hindu,
𝑁1 = 1500, 𝑁 = 5000, 𝑛 = 880

𝑁1
𝑛1 = ×𝑛
𝑁
1500
= × 880 = 264
5000

35
 Solution
For others Religions,
𝑁1 = 500, 𝑁 = 5000, 𝑛 = 880

𝑁1
𝑛1 = ×𝑛
𝑁
500
= × 880 = 88
5000

36
 Solution
Muslim Hindu Other Total
religions
𝑁1 =3000 𝑁2 =1500 𝑁3 =500 𝑁=5000
𝑛1 =528 𝑛2 =264 𝑛3 =88 𝑛=880
Now, we can use simple random sampling to select persons from each
group. In a simple random sample, there are three techniques:
1. Using a random number table.
2. Reminder method.
3. Lottery method.
Here, the whole or total process is known as stratified random
sampling.
37