Name : Ayush yadav
Roll No. : 241181134020
Experiment – 4
Object :- To calculate the sample mean (𝑦̅) For all
possible samples and also verify that -
(i) To verify that the sample mean is an unbiased estimator of
the population mean.
1 1
(ii) Also verify that the Var(𝑦̅) is equal to (𝑛 − 𝑁) 𝑆 2
Experiment :- In the simple random sampling a population of
size N = 6 the values of population characteristics are 8,3,1,11,4,7 ;
Draw the sample of size 2 without replacement.
Theory and Formula :-
❖ Simple random sampling (SRS) is the simplest method of
probability sampling where each unit of population
has an equal chance of being selected in the sample.
Properties of SRS
1. Equal probability of selection :-
[Link]........................
Name : Ayush yadav
Roll No. : 241181134020
𝟏
❖ Every unit has an equal chance to be chosen i.e. ( with
𝑵
replacement). Where N is the total number of units in the
population.
❖ In SRS without replacement (SRSWOR), the probability that a
specific unit is selected on the first draw
𝟏
is . For subsequent draws, selection probabilities depend on
𝑵
the draw order;
𝟏
hence a single expression like is not generally applicable.
𝑵−𝟏
2. Unbiased Selection :-
❖ No favouritism selection is purely random.
Merits of SRS :-
1. The simple random sampling is an unbiased representative
sample of the whole population.
2. Each unit of population has equal chance of being included
in the sample.
Demerits of SRS :-
1. The simple random sampling can give reliable results only
when the population is homogeneous.
2. SRS required large sample size than any other sampling
for a fixed level of precision.
Formula:-
[Link]........................
Name : Ayush yadav
Roll No. : 241181134020
1. Population mean :
N
1
̅ = ∑ Yi
𝐘
N
i=1
2. Sample mean:
n
1
̅ = ∑ yi
𝒚
n
i=1
3. Expectation of sample mean :
1
𝑛𝑜
̅𝒊 ) = ∑𝑖=1
𝑬(𝒚 𝑦̅𝑖
𝑛 𝑜
Where, 𝑛𝑜 ∶ 𝑇𝑜𝑡𝑎𝑙 𝑛𝑜. 𝑜𝑓 𝑠𝑎𝑚𝑝𝑙𝑒 𝑣𝑎𝑙𝑢𝑒𝑠
Note : 1. For , with replacement ;
𝑛𝑜 = 𝑁 𝑛
2. For , without replacement ;
𝑛 𝑜 = N Cn
N : Population size
n : Total sample size
4. Variance of sample mean :
1 𝑛𝑜
𝑽𝒂𝒓(𝒚
̅𝒊 ) = ∑𝑖=1 (𝑦̅𝑖 − 𝑌̅ )2
𝑛𝑜
[Link]........................
Name : Ayush yadav
Roll No. : 241181134020
5. Variance of sample mean for SRSWOR :
𝑁−𝑛 2
𝑽𝒂𝒓(𝒚
̅) = ( )𝑆
𝑁𝑛
6. Population mean square :
𝑁
1
𝑺𝟐 = ∑(𝑌𝑖 − 𝑌̅ )2
𝑁−1
𝑖=1
Calculation :-
Given , Population units :
Y = 8,3,1,11,4,7
N=6;n=2
(i) Population mean,
8+3+1+11+4+7 34
̅
Y= = ̅ = 𝟓. 𝟔𝟕 … … … (i)
=> 𝐘
6 6
No. of possible sample size without replacement,
𝑛𝑜 = 𝑁𝑐𝑛 => 𝑛𝑜 = 6𝑐2 = 𝟏𝟓
[Link]........................
Name : Ayush yadav
Roll No. : 241181134020
Table : 4.1
Population Sample Sample Sample mean ̅) (𝒚
̅𝒊 − 𝒀
(𝒚 ̅) (𝒀 − 𝒀
̅)𝟐 (𝒀 − 𝒀
̅𝒊 − 𝒀 ̅)𝟐
(Y) No. value (𝒚𝒊 ) ̅𝒊 )
(𝒚
8 1 8,3 5.5 -0.17 0.0278 2.333 5.444
3 2 8,1 4.5 -1.17 1.3611 -2.667 7.111
1 3 8,11 9.5 3.83 14.6944 -4.667 21.778
11 4 8,4 6 0.33 0.1111 5.333 28.444
4 5 8,7 7.5 1.83 3.3611 -1.667 2.778
7 6 3,1 2 -3.67 13.4444 1.333 1.778
7 3,11 7 1.33 1.7778
8 3,4 3.5 -2.17 4.6944
9 3,7 5 -0.67 0.4444
10 1,11 6 0.33 0.1111
11 1,4 2.5 -3.17 10.0278
12 1,7 4 -1.67 2.7778
13 11,4 7.5 1.83 3.3611
14 11,7 9 3.33 11.1111
15 4,7 5.5 -0.17 0.0278
34 Total 85 67.333 67.333
𝑛𝑜
1
̅𝒊 ) =
𝑬(𝒚 ∑ 𝑦̅𝑖
𝑛𝑜
𝑖=1
[Link]........................
Name : Ayush yadav
Roll No. : 241181134020
85
E(𝑦̅𝑖 ) = 15 = 𝟓. 𝟔𝟕 … … … … . (ii)
1 𝑛𝑜
(ii) 𝑽𝒂𝒓(𝒚
̅𝒊 ) = ∑𝑖=1 (𝑦̅𝑖 − 𝑌̅ )2
𝑛 𝑜
67.333
𝑉𝑎𝑟(𝑦̅𝑖 ) = { 𝑓𝑟𝑜𝑚 𝑡𝑎𝑏𝑙𝑒 4.1}
15
𝑉𝑎𝑟(𝑦̅𝑖 ) = 𝟒. 𝟒𝟖𝟗 … . … . (𝑖𝑖𝑖)
𝑁
1
𝑺𝟐 = ∑(𝑌𝑖 − 𝑌̅ )2
𝑁−1
𝑖=1
1
𝑆2 = ( ) × 67.333 = 𝟏𝟑. 𝟒𝟔𝟔𝟔 … … … (∗)
6−1
𝑵−𝒏 𝟐 6−2
( )𝑺 = ( ) × 13.4666 = 𝟒. 𝟒𝟖𝟗 … … . (𝒊𝒗)
𝑵𝒏 6×2
Conclusion :-
From eq.(i) & (ii) , we can say that ,For the population {8,3,1,11,4,7},
the sample mean based on samples of size 2 without replacement is an
unbiased estimator of the population mean i.e.
𝐄( 𝒚 ̅ = 𝟓. 𝟔𝟕
̅𝒊 ) = 𝐘
And from eq. (iii) & (iv) ,we can say that –
𝑁−𝑛 2
𝑽𝒂𝒓(𝒚̅𝒊 ) = ( ) 𝑆 = 𝟒. 𝟒𝟖𝟗
𝑁𝑛
[Link]........................
Name : Ayush yadav
Roll No. : 241181134020
[Link]........................
