Unit IV: Sampling Distribution

• Objectives: The main purpose of this unit

is to disseminate knowledge about the
concept of parameter, statistic, standard
error and the sampling distribution in
general, and specially, the concept
computation of standard error of mean
and proportion, and application of
sampling distribution of mean and
proportion.
Parameter and Statistics:
Parameter:
• It is the characteristics of the population.
• It is constant related to the population as a whole.
• The value of parameter is assumed to be unknown.
• The value of parameter is estimated from the sample.
Statistics:
• It is the characteristics of the sample.
• It is variable value related to the sample only.
• The value of statistic is obtained after sampling.
• It is used to estimate the value of parameter.
• The sample values are different for different samples.
Example 1: If a researcher is wanted to investigate the average
time (in hrs) that the Facebook users spent per day, then if
average time spent by 55 users is found to be 2.5hrs. The
parameter is the average time spent by all users (unknown) and
its statistics is sample mean of 2.5hrs for the sample size 55.
Example 2: If the operation manager is wanted to find the
percentage of the defective items (say light bulbs) of a certain
electric company, for that he selects 150 items (bulbs) from lot
then found to be 0.04 (4%) defective items(bulbs), then the
proportion (in %) of all the defective items produced by the
company is a parameter (unknown) and the proportion of
defective items 0.04 (4%) obtained from the sample of 150 items
is statistics.
Notations:
Parameter Statistics
Population value = X Sample Value = x
Population size = N Sample size = n
Population mean = , Sample mean = 𝒙,
median, mode median, mode
Population proportion = P Sample proportion = p
Population sd = Sample sd = s or S
Population Variance = 2 Sample Variance = s2
Population correlation Sample correlation
coefficient = coefficient= r
Population regression Sample regression
coefficient =i coefficient = bi
Concept of sampling distribution:
Sampling : The process of selecting a sample from the
population.
Sampling distribution:
• It is basis for the inferential statistics.
• The probability distribution of the sample statistics from all
possible sample of fixed size from the fixed population.
• Since sample statistics is a variable (not constant value), so it
is possible to make the distribution of the statistics for all
samples.
The chances of the selection of a particular sample from the
fixed population with fixed size of sample are equal, so it is
related with the probability distribution
 Standard Error:
• The standard deviation of the sampling distribution
of the sample statistics is the standard error.
• It is used for testing the reliability of the estimate.
𝟏
Reliability (precision) 
𝑺𝑬
• It is used to estimate the population parameter.
• It is used to test the hypothesis about population
parameter
Example 1. For the given population data, draw a
sample of size 2 without replacement then
(a) Calculate the population mean and standard
deviation
(b) Write all possible samples.
(c) Find the sampling distribution of mean and
calculate standard deviation.
(b) Calculate the standard error of mean.
Data: 3,4, 5, 6, 7
Solution:
Calculation of Population Mean and Standard Deviation
SN X (X-5) (X-5)2
1 3 -2 4
2 4 -1 1
3 5 0 0
4 6 1 1
5 7 2 4
Total X=25 0 10
𝑋 25
Population Mean () = = =5
𝑁 5
2
Population standard deviation( )  
(X -  ) 10
  2  1.41
N 5
For sampling distribution:
5!
Here N= 5, n = 2, The number of possible sample = C(5,2) = 10 [C(5,2) = = 10]
2!3!
Possible samples[without replacement]: Possible samples = {(3,4), (3,5), (3,6), (3,7),
(4,5), (4,6), (4,7), (5,6), (5,7), (6,7)}
Sampling distribution of sample mean
S.N. Sample Mean(x) Probability P(x) xP(x)
1 (3, 4) 3.5 1/10=0.1 0.35
2 (3, 5) 4 1/10=0.1 0.40
3 (3, 6) 4.5 1/10=0.1 0.45
4 (3, 7) 5 1/10=0.1 0.50
5 (4, 5) 4.5 1/10=0.1 0.45
6 (4, 6) 5 1/10=0.1 0.50
7 (4,7) 5.5 1/10=0.1 0.55
8 (5, 6) 5.5 1/10=0.1 0.55
9 (5, 7) 6 1/10=0.1 0.60
10 (6, 7) 6.5 1/10=0.1 0.65
Total - 50 1 5.00
Mean of all sample means = Mean of all possible sample means   i   5
X 50
10 10
i.e. Mean of sample means = 5
That is, population mean is equal to the mean of sampling distribution.
 Calculation of Standard Deviation of Sampling Distribution of Mean
S.N. Sample Mean(𝑥𝑖 ) (𝑥𝑖 -5)2 Probability xP(x) (𝑥𝑖 -5)2P(x)
P(x)
1 (3, 4) 3.5 2.25 1/10=0.1 0.35 0.225
2 (3, 5) 4 1 1/10=0.1 0.40 0.100
3 (3, 6) 4.5 0.25 1/10=0.1 0.45 0.025
4 (3, 7) 5 0 1/10=0.1 0.50 0.000
5 (4, 5) 4.5 0.25 1/10=0.1 0.45 0.025
6 (4, 6) 5 0 1/10=0.1 0.50 0.000
7 (4,7) 5.5 0.25 1/10=0.1 0.55 0.025
8 (5, 6) 5.5 0.25 1/10=0.1 0.55 0.025
9 (5, 7) 6 1 1/10=0.1 0.60 0.100
10 (6, 7) 6.5 2.25 1/10=0.1 0.65 0.225
Total - 50 7.5 1 5.00 0.750
 ( x i - x) 2 7.5
Standard deviation of sample means    0.75  0.87
10 10
Standard Error of mean = S.E.(𝑥 ) =0.87
Standard deviation of sampling distribution is not equal to population standard
deviation.
Alternatevly
 N n 2 (5  2) 3
Standard Error of mean ( x )     0.75  0.87
n N 1 2 (5  1) 4
Example 2: In a population if the elements are 1, 3, 5, 7, then find
(a) Population mean and standard deviation.
(b) Mean and Standard deviation of sampling distribution with replacement
(c) Standard error of mean.
Solution: (a) Calculation of Mean and Standard Deviation of Population
S.n. X (X-  )2
1 1 9
2 3 1
3 5 1
4 7 9
Total 16 20
𝑿 𝟏𝟔
Population Mean () = = =𝟒
𝑵 𝟒

 (X -  ) 2 20
Population standard deviation( )    5  2.24
N 4
For sampling distribution: N= 4, n = 2, Number of possible samples = 4x4 = 16
Possible sample ={(1,1), (1,3), (1,5), (1,7), (3,1), (3,3), (3,5), (3,7), (5,1), (5,3), (5,5), (5,7),
(7,1), (7,3), (7,5), (7,7)}
Calculation of Mean and SD of Sampling Distribution of Mean
Sample No Sample Mean(𝒙𝒊 ) (𝒙𝒊 -4)2
1 (1,1) 1 9
2 (1,3) 2 4
3 (1,5) 3 1
4 (1,7) 4 0
5 (3,1) 2 4
6 (3,3) 3 1
7 (3,5) 4 0
8 (3,7) 5 1
9 (5,1) 3 1
10 (5,3) 4 0
11 (5,5) 5 1
12 (5,7) 6 4
13 (7,1) 4 0
14 (7,3) 5 1
15 (7,5) 6 4
16 (7,7) 7 9
Total - 64 40

Mean of sample means ( x )   xi  64  4

16 16
 ( x i - x) 2 40 5
Standard deviation of sampling distribution(s )     1.58
16 16 2
(c) [Alternately]
 5
Stadard Error of mean  SE( x )    1.58
n 2
 Standard Error for statistics (Mean, Proportion, Correlation)
S.N. Standard Error Formula for SE Remarks
1. Standard error for mean 
SE( x ) 
with replacement n
2. Standard error for mean  N unknown or infinite
SE( x ) 
n
3. Standard error for mean  N n
without replacement SE( x )  .
n N 1
4. Standard error for PQ pq Where P + Q = 1, if P and
proportion with SE(p )   Q are unknown then
n n
replacement replace P=p and Q=q
5. Standard error for PQ N unknown or infinite
proportion SE(p ) 
n
6. Standard error for PQ N  n
proportion with SE(p ) 
n N 1
replacement
7. Standard error for 1 r2
correlation coefficient SE(r) 
n
Example 3: If a sample of size 24 is drawn from the population of size 100 having
standard deviation 12 find the standard error of the mean.
Solution: Here, n = 24, N = 100 ,  = 12, SE(𝒙) = ?
 N  n 12 100  24 12 76
Now, SE( x )  .    2.15
n N 1 24 100  1 24 99

Example 4: From a large consignment of calculators the sample of 250 are

randomly selected then the number of defective calculators found to be 18.
Calculate the standard error of proportions of defective calculators.
𝒙 𝟏𝟖
Solution: Here, n = 245, x= 17, p = = == 𝟎. 𝟎𝟕𝟐, q= 1- p = 1- 0.072 =0.928
𝒏 𝟐𝟓𝟎
PQ pq 0.072  0.928 0.0668
SE(p)      0.016
n n 250 250
 Central Limit Theorem:
[For any type of population(normal or non-normal]
Statement: The central limit theorem states that if you have a population
with mean μ and standard deviation σ and take sufficiently large random
samples from the population with replacement, then the distribution of
the sample means will be approximately normally distributed.
OR
If the random samples are drawn from any population (normal or non
normal) having mean µ and standard deviation  then the sampling
distribution of the mean is normally distributed with mean µ and

standard deviation n
as the sample size n is large.
Example 5. If a random sample of size 10 is drawn from a normal
population with mean 20 and standard deviation 12 find the probability
that sample mean is not more than 16.
Solution:
Here, n = 10,  = 20, = 12, 𝒙 = 𝟏𝟔
The standard normal variate (SNV) is
X 
Z

n
Then find the probability for P(x<16) from normal table.(Do Yourself)
 Exercise
1. If the population having elements 2, 4, 5, 6, 8 then find
(a) Population mean and standard deviation
(b) Prepare the sampling distribution of mean
(c) Calculate the mean and standard deviation of
sampling distribution of mean (i) without
replacement (ii) with replacement
(d) Also calculate the standard error directly.
2. In the survey of 32 customers who were visited last Sunday in
NIBL bank, the waiting mean time was found to be 10 minutes,
if the standard deviation of the waiting time of that bank is 3
minutes then find the standard error of the mean.
3. From a large consignment, 400 oranges are randomly selected
and 56 are found to be bad then find the standard error of the
proportion of bad oranges.
THANK YOU