Lecture 3

Methods of Selecting Simple Random Sample

To ensure the randomness in the selection and to make sample as a
representative part of the population, the method of selection must be independent
of human judgement as far as possible. We can do it by choosing sample by using
the following three methods:
(1) Lottery Method
(2) Random Number Table Method and
(3) Random Number Generator Software.
Lottery Method
➢ This is a very popular, simple and old method of taking a random sample.
Under this method, all items of the population are numbered or named on
separate sheet of papers of identical size, color and shape. The sheets are then
folded and mixed up in an urn. A blindfold selection is then made of the
number of sheets required to constitute the desired size of the sample. The
selection of items thus depends entirely on chance.
Use Random number Table
Several random number tables are available. Some of these are:
(i) Kendal and Smith,
(ii) Fisher and Yates,
(iii) Tippett,
(iv) Snedecor and Cochran,
(v) The Rand Corporation, etc.
The following steps should follow in drawing a simple random sample of n units
from a population of N units by using random number table.
1. Identify and assign N units in the population with numbers 1 to N.
2. Decide on the random number table to be used.
3. Choose an N-digit random number from any point in the random number
table.

1
 Rejection Method
4. If this random number is less than or equal to N, this is our first selected
unit.
5. Move on to the next random number not exceeding N, vertically,
horizontally, or in any other direction systematically and choose our
second unit.
6. If at any stage of our selection, the random number chosen exceeds N,
discard it and choose the next random number.
Remainder Method
In choosing random number following rejection method may involve a large
number of rejections since all values greater than N or equal to 0 appearing in the
random numbers are not considered for selection. To minimize this rejection rate,
we employ remainder method.
4. If this random number is less than or equal to N, this is our first selected unit.
5. Move on to the next random number vertically, horizontally, or in any other
7. Thedirection
process systematically
stops once we arrive at ourour
and choose desired sample
second unit. size.
8. 6.
In case
If at of
anysimple
stage random sample with
of our selection, thereplacement,
random number all random
chosennumbers
exceeds are
N,
accepted
divide even if repeated
the selected morenumber
random than once.
by In casethe
N and of simple
remainderrandom sampling
of division is
without replacement,
the selected randomdiscard the repeated numbers and draw more numbers.
number.

Random Number Generator Software.

Some statistical software like Excel, R, SAS etc. have inbuilt functions for drawing
a sample using simple random sample with/without replacement.
Advantages of Simple Random Sampling
1. Lack of Bias: It is a fair method of sampling and if applied appropriately it
helps to reduce any bias involved.
2. Simplicity: This is a very basic method of collecting data. No technical
knowledge is required to select simple random sampling.

2
Disadvantages of Simple Random Sampling
1. Costly and Time Consuming:
(a) It is a costly method of sampling as it requires a complete list of all
elements in the population (sampling frame).
(b) Sampled individuals may be so widely dispersed that visiting each
selected individual may be extremely expensive and time consuming.
2. Bias: Certain sub-groups in the population may be totally overlooked or may
be over-representative in the sample as a result of chance factor. In either case,
bias can occur.
Estimation of Population Parameters
Notions and Formulas

Let 𝑦1 , 𝑦2 , … … … … … 𝑦𝑁 , be the values of the population elements and

𝑦1 , 𝑦2 , … … … … 𝑦𝑛 be the values of the selected sample elements. Then

Population total: 𝑌 = ∑𝑁
𝑖=1 𝑦𝑖
∑𝑁
𝑖=1 𝑌𝑖
Population Mean: 𝑌̅ =
𝑁
𝑁 2
2 ∑𝑖=1(𝑦𝑖 −𝑌̅)
Population Variance: 𝜎 =
𝑁
∑𝑛
𝑖=1 𝑦𝑖
Sample Mean: 𝑦̅ =
𝑛

2 ∑𝑛 ̅)2
𝑖=1(𝑦𝑖 −𝑦
Sample Variance: 𝑠 =
𝑛−1
The estimators of population total and mean are given by

𝑁 𝑛
𝑌̂ = ∑ 𝑦𝑖 = 𝑁𝑦̅
𝑛 𝑖=1

3
and 𝑌̅̂ = ∑𝑛𝑖=1 𝑦𝑖 /𝑛 = 𝑦̅ respectively.

𝑁
The factor by which the sample is multiplied is sometimes called the expansion
𝑛
or raising or inflation factor.

In simple random sampling, wor, the sample mean 𝑦

̅ is an unbiased estimator of 𝑌̅
and its sample variance is given by

𝑛 𝑆2 𝑆2
𝑉 (𝑦̅) = (1 − ) = (1 − 𝑓)
𝑁 𝑛 𝑛

2 ∑𝑁 ̅ 2
𝑖=1(𝑦𝑖 −𝑌) 𝑁
where 𝑆 = = 𝜎2
𝑁−1 𝑁−1

𝑛
Here = 𝑓 is known as sampling fraction.
𝑁
𝑛
(1 − ) = (1 − 𝑓) is known as finite
𝑁
population correction (fpc).
̅ in random sampling, wor, is given by
An unbiased estimator of the variance of 𝒚

𝑠2
𝑣(𝑦̅) = (1 − 𝑓)
𝑛
Simple Random Sampling with Replacement (SRSWR)
When n units are selected with SRSWR, the total number of possible samples are
𝑁 𝑛.

4
The variance of 𝑦̅ under SRSWR is

𝑁−1 2
𝑣(𝑦̅) = 𝑆
𝑁𝑛
It is to be noted that if N is infinite (large enough), then

𝑆2
𝑣 (𝑦̅) =
𝑛