B.

Tech II Semester (2020 Batch)

PROBABILITY AND STATISTICS (20BM1104)
(For CSE-3 & CSE-4)
Units – 2: Sampling Distribution

(Population and sample, sampling distribution of the mean (σ known), sampling distribution of the mean (σ
unknown), sampling distribution of the variance: Chi-square and F-distributions.)

Sampling Distribution: population and sample, sampling distribution of the mean (σ known), sampling
distribution of the mean (σ unknown), sampling distribution of the variance: Chi-square and F-
distributions.
Population: A collection of objects (collection of numbers, measurements, observations etc) is called a
Population
The number of objects in a population is called its Size and is denoted by N
If N is finite then the population is called Finite population
If N is infinite then the population is called Infinite population

Examples:
(1) Heights of the students in a University
(2) Marks obtained by the students of SSC in Mathematics
(3) Scores of candidates obtained in a competitive exam
(4) The set of outcomes when a coin tossed 1000 times
(5) Collection of all even numbers

Parameters: The statistical measures (Mean, Median, Variance etc.) about a population are called
Parameters. The Mean, Variance and Standard deviation of a population are respectively denoted by the
symbols  ,  2 and 

Sample: A finite sub collection from a population is called a Sample

The number of objects in a sample is called its Size and is denoted by n
If n  30 then the sample is called large sample
If n  30 then the sample is called small sample

Statistics: The statistical measures (Mean, Median, Variance etc.) about a sample are called statistics.
The Mean, Variance and Standard deviation of a sample are respectively denoted by the symbols
x , s 2 and s

Examples:
(1) For the population of ‘the Heights of the students in a University’, the heights of the students in class
of 40 is a sample
(2) For the population of ‘the Marks obtained by the students of SSC in Mathematics’, the marks of the
students of a particular school is a sample
(3) For the population of ‘the Scores of candidates obtained in a competitive exam’, the scores of the
candidates from a particular college is a sample
(4) For the population of ‘the set of outcomes when a coin tossed 1000 times’, a collection of 10 outcomes
is a sample
(5) For the population of ‘the Collection of all even numbers’, a collection of 20 even numbers is a sample
Random Sample: A sample of size n taken from a population is called a random sample if the probability
of any choice of n objects from the population is same.

Sampling: Collecting samples from a given population is called sampling

Large Sampling: Collecting large samples from a given population is called large sampling
Small Sampling: Collecting small samples from a given population is called small sampling
Sampling with replacement: Collecting samples in which the objects may repeat, from a given population
is called sampling with replacement
Sampling without replacement: Collecting samples in which the objects not repeat, from a given
population is called sampling without replacement

Number of Samples:
N
(i) The number of samples of size n without replacement, taken from a finite population of size N is C n
n
(ii) The number of samples of size n with replacement, taken from a finite population of size N is N
(iii) The number of samples of size n with or without replacement, taken from an infinite population is 

Examples:
(1) Consider the finite population 1, 2, 3, 4  of size N  4
The number of samples of size n  2 without replacement is given by Cn  C2  6
N 4

The samples are 1, 2, 1, 3 , 1, 4 ,  2, 3 ,  2, 4 , 3, 4 

The means of these 6 samples are respectively 1.5, 2, 2.5, 2.5, 3, 3.5
The frequency distribution of the these sample means is given as follows
Sample mean x 1.5 2 2.5 3 3.5
Frequency f 1 1 2 1 1
This frequency distribution is called the sampling distribution of the mean (SDM)

Note: Here drawing a sample of size n  2 without replacement from the above population is a random
experiment with the possible outcomes 1, 2, 1, 3 , 1, 4 ,  2, 3 ,  2, 4 , 3, 4 
That is, the sample space S  1, 2, 1, 3 , 1, 4 ,  2, 3 ,  2, 4 ,  3, 4 
If X is the random variable which gives the mean the sample, then the range of X is 1.5, 2, 2.5, 3, 3.5 

(2) Consider the finite population 1, 2, 3, 4  of size N  4

The number of samples of size n  2 with replacement is given by N  4  16
n 2

The samples are 1,1, 1, 2 , 1, 3 ,  1, 4 ,

 2,1,  2, 2 ,  2, 3 ,  2, 4 ,
3,1, 3, 2 , 3, 3 ,  3, 4 ,
 4,1,  4, 2 ,  4, 3 ,  4, 4 
The means of these 16 samples are respectively given follows
1, 1.5, 2, 2.5,
1.5, 2, 2.5, 3,
2, 2.5, 3, 3.5,
2.5, 3, 3.5, 4
The frequency distribution of the these sample means is given as follows
Sample mean x 1 1.5 2 2.5 3 3.5 4
Frequency f 1 2 3 4 3 2 1

This frequency distribution is the sampling distribution of the mean (SDM)

Sampling Distribution of the Mean (SDM):
The frequency distribution of the means of all random samples of fixed size, taken from a population is
called the Sampling Distribution of the Mean (SDM). It is denoted by X .

Sampling Distribution of the Mean with  (SDM with  ):

Theorem: If X is the random variable which gives the mean a random sample of size n, taken from a
population having mean  and variance  2 , then
(i) Mean of SDM X is given by  X   or E X    
(ii) For infinite population or sampling with replacement,
2 2
Variance of SDM X is given by  X  or V X  
2

n n
(iii) For finite population of size N and sampling without replacement,
2  N n
Variance of SDM X is given by  
2
 
X
n  N 1 

Note:
N n
(1) The value of is called the finite population correction factor
N 1
(2) The standard deviation of X is called Standard error of the mean; that is,

Standard error (SE), X 
n

(3) The value of 0.6745  is called the Probable error of the mean; that is,
n

Probable error (PE)  0.6745 
n

Chebyshev’s Theorem:
If X is the random variable which gives the mean a random sample of size n, taken from a population


having mean  and variance  , then P X    k  1 
2
 2
nk2
, where k is a positive constant.

Note:
 
1. P X    k  1 
2
nk2
or P   k  X    k   1 
2
nk2


2. P X    k   2
nk2

Central Limit Theorem:

If X is the random variable which gives the mean a random sample of size n, taken from a population
X 
having mean  and variance  2 , then Z  .
  
 
 n
Examples:
(1) Find the finite population correction factor for n  10 and N  1000
N  n 1000  10
Finite population correction factor,   0.991
N  1 1000  1
(2) Find the finite population correction factor for n  5 and N  100
N  n 100  5
Finite population correction factor,   0.9596
N  1 100  1
(3) What is the effect on standard error, if a sample is taken from an infinite population and its size is
increased from 400 to 900?
Here n1  400 and n2  900
 
Initially, the Standard Error SE1   X  
n1 400
 
After the sample size increasing, the Standard Error SE2   X  
n2 900
SE 2  400 2
Now consider   
SE1 900  3
2
Therefore, the Standard Error is decreased times of its original value
3
(4) What is the effect on standard error, if a sample is taken from an infinite population and its size is
decreased from 100 to 25?
Here n1  100 and n2  25
 
Initially, the Standard Error SE1   X  
n1 100
 
After the sample size decreasing, the Standard Error SE2   X  
n2 25
SE 2  100
Now consider   2
SE1 25 
Therefore, the Standard Error is increased 2 times of its original value
(5) *A population consists 4 numbers 1, 2, 3, 4 (i) Find the Mean and Variance of the population
(ii) Write all possible samples of size 2 without replacement (iii) Write the Sampling distribution of the
mean (SDM) (iv) Find Mean and Variance of the Sampling distribution of the mean (SDM) (v) Verify
the results in (iv) with suitable formula
Solution: (i) Mean and Variance of the population:
1  2  3  4 10
Mean,     2.5
4 4
1  4  9  16
 2.5  7.5  6.25  1.25
1 30
Variance,  2   x 2   2   2 
2

n 4 4
(ii) Samples of size 2 without replacement:
The number of samples of size n  2 without replacement is given by Cn  C2  6
N 4

The samples are 1, 2, 1, 3 , 1, 4 ,  2, 3 ,  2, 4 , 3, 4 

(iii) The Sampling distribution of the mean (SDM):
The means of the above 6 samples are respectively 1.5, 2, 2.5, 2.5, 3, 3.5
The frequency distribution of the these sample means is given as follows
 Sample mean x 1.5 2 2.5 3 3.5
Frequency f 1 1 2 1 1

Which is the sampling distribution of the mean (SDM)

(iv) Mean and Variance of the Sampling distribution of the mean (SDM)

1 1.5  2  2.5  2.5  3  3.5 15

Mean,  X 
6
 x
6

6
 2.5
1
Variance,  X2   x 2   X
2

6
2.25  4  6.25  6.25  9  12.25
 2.5
40
  X 
2 2

6 6
 6.6667  6.25  0.4167
(v) Verification:
We have  X  2.5  
 2  N  n  1.25  4  2 
     0.4167   X
2
and
n  N 1  2  4 1 
2  N n
Therefore,  X   and  X    are verified.
2

n  N 1 

(6) *A population consists 4 numbers 1, 2, 3, 4 (i) Find the Mean and Variance of the population
(ii) Write all possible samples of size 2 with replacement (iii) Write the Sampling distribution of the
mean (SDM) (iv) Find Mean and Variance of the Sampling distribution of the mean (SDM) (v) Verify
the results in (iv) with suitable formula
Solution: (i) Mean and Variance of the population:
1  2  3  4 10
Mean,     2.5
4 4
1 1  4  9  16
Variance,  2   x 2   2    2   2.52  7.5  6.25  1.25
30
n 4 4
(ii) Samples of size 2 with replacement:
The number of samples of size n  2 without replacement is given by N  4  16
n 2

The samples are 1,1, 1, 2 , 1, 3 ,  1, 4 ,

 2,1,  2, 2 ,  2, 3 ,  2, 4 ,
3,1, 3, 2 , 3, 3 ,  3, 4 ,
 4,1,  4, 2 ,  4, 3 ,  4, 4 
(iii) The Sampling distribution of the mean (SDM):
The means of the above 16 samples are respectively
1, 1.5, 2, 2.5,
1.5, 2, 2.5, 3,
2, 2.5, 3, 3.5,
2.5, 3, 3.5, 4
The frequency distribution of the these sample means is given as follows
Sample mean x 1 1.5 2 2.5 3 3.5 4
Frequency f 1 2 3 4 3 2 1

Which is the sampling distribution of the mean (SDM)

 (iv) Mean and Variance of the Sampling distribution of the mean (SDM)

Mean,  X 
 x f  11  1.52  23  2.54  33  3.52  41  40  2.5
f 16 16

Variance,  2

x f 2
2
X
f X


1 1  1.5 2  2 3  2.5 4  3 3  3.5 2  4 1
2 2 2 2 2 2 2

2
X
16
 2.5
110

2
 6.875  6.25  0.625
16
(v) Verification:
2 1.25
We have  X  2.5   and   0.625   X2
n 2
 2

Therefore,  X   and  X2  are verified.

n
(7) A random sample of size 100 is taken from an infinite population having the mean 76 and the variance
256. What is the probability that the sample mean will be between 75 and 78
X 
Here n  100 ,   76,  2  256 and   16 , and by central limit theorem, Z 
  
 
 n
   
   
 75   X   78     75  76 78  76 
P75  X  78  P     P  16   Z   16  

            
 n   n   n    100    
         100 
 P 0.63  Z  1.25  F 1.25  F  0.63
 F 1.25  1  F 0.63  0.8944  1  0.7357  0.6301

(8) A normal population has mean 0.1 and standard deviation 2.1. Find the probability that the mean of a
sample of size 900 will be negative
X 
Here n  900,   0.1 and   2.1 and by central limit theorem, Z 
  
 
 n
   
   
 X  0   0  0.1 
P  X  0   P    P Z   2.1    PZ  1.43  F  1.43
       
  n   n      
      900  

 1  F 1.43  1  0.9236  0.0764

(9) A random sample of size 64 is taken from a normal population with mean 51.4 and standard deviation
6.8. Find the probability that the mean of the sample will (i) exceed 52.9 (ii) fall between 50.5 and
52.3 (iii) less than 50.6
X 
Here n  64,   51.4 and   6.8 , and by central limit theorem, Z 
  
 
 n
   
   
 X   52.9     52.9  51.4 
(i) Exceed 52.9, PX  52.9  P    P Z   6.8  
      
  n     
   
    n     64  
 PZ  1.76  1  PZ  1.76
 1  F 1.76  1  0.9608  0.0392
(ii) Fall between 50.5 and 52.3,
   
   
 50.5   X   52.3     50.5  51.4 52.3  51.4 
P50.5  X  52.3  P     P  6.8   Z   6.8  

         
        
    
  n   n   n     64   64  
 P 1.06  Z  1.06  F 1.06  F  1.06
 F 1.06  1  F 1.06  2F 1.06  1  20.8554 1  0.7108
   
   
 X   50.6     50.6  51.4 
(iii) Less than 50.6 PX  50.6  P   P Z 
        6.8  
         
 n   n     64  

 PZ  0.94  F  0.94

 1  F 0.94  1  0.8264  0.1736

(10) If a 1- gallon can of paint covers on an average 513 square feet with a standard deviation of 31.5
square feet, what is the probability that the mean area covered by a sample of 40 of these 1- gallon
cans will be anywhere from 510 to 520 square feet
X 
Here n  40,   513 and   31.5 and by central limit theorem, Z 
  
 
 n
   
   
 510   X   520     510  513 520  513 
P510  X  520  P    P Z
          31.5   31.5  
             
  n   n   n     40   40  
 P 0.6  Z  1.4  F 1.4  F  0.6  F 1.4  1  F 0.6
 0.9192 1  0.7258  0.645
(11) For a large sample of size n, verify that there is a 50-50 chances that the mean of a random sample
from an infinite population with standard deviation  differ from  by less than 0.6745  
n
  
We have prove that P X    0.6745    0.5
 n
 
 
    X  
Consider, P X    0.6745    P  0.6745   P Z  0.6745
 n     
  n  
 
 P 0.6745  Z  0.6745  F 0.6745  F  0.6745
 F 0.67  1  F 0.67  0.7486  1  0.7486  0.497

(12) If the mean of breaking strength of copper wire is 575 lbs, with a standard deviation of 8.3 lbs. How
large a sample must be used in order that there will be one chance in 100 that the mean breaking
strength of the sample is less than 572 lbs
X 
Here   575 and   8.3 By central limit theorem, Z 
  
 
 n
We have to find the value of n such that PX  572 
1
100
Now PX  572  0.01
 
 
 X   572   
 P    0.01

      
 n    
   n 
 
 
 572  575 
 P Z    0.01
  8 .3 
   
  n  
 3 n
 P Z     0.01

 8.3 
 3 n
 F     0.01

 8.3 
 3 n
 1  F    0.01

 8. 3 
 3 n
 F    0.99

 8.3 
 2.33  8.3 
2
3 n
  2.33  n     41.55  42
8.3  3 
Exercise:
(1) Define (i) Population (ii) Sample (iii) Large sample (iv) Small sample (v) Random sample
(2) Define (i) Sampling (ii) Sampling Distribution of the Mean (iii) finite population correction factor
(iv) Standard error of the mean (v) Probable error of the mean
(3) State (i) Chebyshev’s Theorem (ii) Central Limit Theorem
(4) Write all possible samples of size 3 without replacement from the population 1, 2, 3, 4 .
Also compute the means of all these samples
(5) Write all possible samples of size 3 without replacement from the population 1, 2, 3, 4 .
Also compute the means of all these samples
(6) Find the finite population correction factor for n  5 and N  1000
(7) Find the finite population correction factor for n  25 and N  1000
(8) What is the effect on standard error, if a sample is taken from an infinite population and its size is
decreased from 800 to 200?
(9) What is the effect on standard error, if a sample is taken from an infinite population and its size is
increased from 300 to 2700?
(10) A population consists 4 numbers 3, 7, 11, 15
(i) Find the Mean and Variance of the population
(ii) Write all possible samples of size 2 with replacement
(iii) Write the Sampling distribution of the mean (SDM)
(iv) Find Mean and Variance of the Sampling distribution of the mean (SDM)
(v) Verify the results in (iv) with suitable formula
(11) A population consists 4 numbers 3, 7, 11, 15
(i) Find the Mean and Variance of the population
(ii) Write all possible samples of size 2 without replacement
(iii) Write the Sampling distribution of the mean (SDM)
(iv) Find Mean and Variance of the Sampling distribution of the mean (SDM)
(v) Verify the results in (iv) with suitable formula
(12) If X is the mean of a random sample of size n, taken from the population 1, 2, 3,  N , find the
mean and variance of the Sampling distribution of the mean (SDM)
(13) Construct sampling distribution of means for the population 5, 10, 12, 18 by drawing samples of size
2 without replacement. Find (i) Mean of the population (ii) Standard deviation of the population (iii)
Mean of the sampling distribution of means (iv) Standard deviation of the sampling distribution of
means
(14) A random sample of size 100 is taken from a normal population with mean 76 and standard deviation
16. Find the probability that the mean of the sample will (i) exceed 77 (ii) fall between 75 and 78
(15) A random sample of size 81 is taken from a normal population with mean 65 and standard deviation
10. Find the probability that the mean of the sample will (i) exceed 67 (ii) fall between 66 and 68
(16) A random sample of size 36 is taken from a normal population with mean 155 and standard deviation
15. Find the probability that the mean of the sample will be (i) less than 157 (ii) in between 153 and
158 (iii) greater than 160
(17) A sample of size 400 is taken from an infinite population with Standard deviation 16. Find the
Standard error and Probable error of the mean
(18) If the mean of breaking strength of copper wire is 676 lbs, with a standard deviation of 12 lbs. How
large a sample must be used in order that there will be one chance in 100 that the mean breaking
strength of the sample is less than 672 lbs
(19) The mean of certain normal population is equal to the standard error of the mean of samples of size
64. Find the probability that the mean of the sample size 36 will be negative
Sampling Distribution of the Mean with unknown  (SDM with unknown  ):

X 
If we do not know the value of  , then we cannot use the Central limit theorem Z  .
  
 
 n
In this case we use sample standard deviation ' s ' in place of population standard deviation  so that we
X 
have a random variable different from Z . This new random variable is denoted by t ; that is t  .
 s 
 
 n
The probability distribution corresponding to this random variable t is called t -distribution with parameter
n  1 . This parameter is known as degrees of freedom, denoted by ; that is,   n  1.

Note: In the t -distribution, for the sample x1 , x2 , x3 ,  xn ,

1 n
(i) Sample mean x is given by x   xi
n i 1
1 n
(ii) Sample variance is given by s 2    x i  x 2
n  1 i 1

Properties of t - distribution:

(1) The curve given by t -distribution is called t -curve

(2) The t -curve is continuous and above the X-axis (or t - axis)
(3) The curve is symmetric about the Y-axis
(4) The t -curve is similar to Z -curve (standard normal curve)
(5) The area between t - axis and the curve from   to  is 1 unit
(6) The mean of t is 0 and the variance of t is greater than 1; that is,  t  0 and  t2  1
(7) As n  , variance of t tends to 1; that is,  t2  1 . In other words, as n  , t  Z
X 
(8) If n is large ( n  30 ), then we can write Z in place of t ; that is, Z 
 s 
 
 n
(9) P(t  a)  P(t  a)
1
(10) P(t  0)  P(t  0) 
2

t - Notation:
If   0 then t  is a point on t – axis such that P(t  t )   or P(t  t )  
That is, the area between t – axis and the curve from t to  is 
(or the area between t – axis and the curve from   to  t is  )
Note:
(1) t1   , t0   , t 1  0
2
(2) t  t1  0 or  t  t1 or t  t1
Testing a claim using t -distribution: To test a given claim using t -distribution, we follow the rule given
below.
(i) If t  t0.005 then the claim is accepted
(ii) If t  t 0.005 then the claim is rejected
t - Table:
In this table the values of t  are available for different values of  and 

Example:
(i) For   7 or n  8 and   0.025 , t  t0.025  2.365
(ii) For   10 or n  11 and   0.05 , t  t0.05  1.812
(iii) For   15 or n  16 and   0.005 , t  t0.005  2.947

Problems:
(1) Find (i) t 0.05 when   12 (ii)  t 0.01 when   8 (iii) t 0.995 when   10
(i) When   12 , t0.05  1.782
(ii) When   8 ,  t0.01  2.896
(iii) When   10 , t0.995  t10.995  t0.005  3.169 ( t  t1 )

(2) (i) Find t 0.975 when   13 (ii) Find  t0.99 when   10 (iii) Find t 0.95 when   11
(i) When  13 , t0.975  t10.975  t0.025  2.160 ( t  t1 )
(ii) When  10 ,  t0.99  t10.99  t0.01  2.764 ( t  t1 )
(iii) When  11 , t0.95  t10.95  t0.05  1.796 ( t  t1 )
(3) Find (i) P(t  2.365) when   7 (ii) P(t  1.318) when   24
(iii) P(t  2.567) when   17 (iv) P(1.356  t  2.179) when   12

(i) When   7 , P(t  2.365)  P(t  t0.025)  1  P(t  t0.025)  1  0.025  0.975
(ii) When   24 , P(t  1.318)  P(t  t0.1 )  0.1
(iii) When   17 , P(t  2.567)  P(t  t0.01)  1  P(t  t0.01)  1  P(t  t0.01)  1  0.01  0.99
(iv) When   12 , P(1.356  t  2.179)  P(t0.1  t  t0.025)  1  (0.1  0.025)  0.875
P(t0.1  t  t0.025)  P(t  t0.1 )  P(t  t0.025)  P(t  t0.9 )  P(t  t0.025)  0.9  0.025  0.875

(i) P(t  t0.025) (ii) P(t  t0.1 ) (iii) P(t  t0.01) (iv) P(t0.1  t  t0.025)

(4) A random sample of size 25 from a normal population has mean x  47.5 and the standard deviation
s  8.4 . Does this information tend to support or refute the claim that the mean of the population is
  42.1

Here n  25  small sample  and   n  1  24

Also x  47.5 , s  8.4 ,   42.1 and t0.005  2.797
We know that given claim is accepted if and only if t  t0.005
X   47.5  42.1
Now t    3.2143 and t  3.2143
 s   8.4 
   
 n  25 
Since t  t 0.005 , reject the claim

(5) A process for making certain ball bearings is under control if the diameters of the bearings have a
mean of 0.5 cm. What can you say about the process if a random sample of 12 of these bearings has a
mean diameter of 0.515 cm and standard deviation of 0.017 cm

Here n  12  small sample  and   n 1  11

Also x  0.515 , s  0.017 ,   0.5 and t0.005  3.106
We know that given claim is accepted (the process is under control) if and only if t  t0.005
X   0.515  0.5
Now t    3.0566 and t  3.0566
 s   0.017 
   
 n  12 
Since t  t 0.005, accept the claim; that is, the process is under control

Note: In place of x  0.515 , if we take x  0.48 then t  4.0754 , t  t 0.005 and therefore we
need to reject the claim.
 1
(6) The t distribution with 1 degree of freedom is given by f (t )  (1  t 2 ) 1 for    t   . Verify

the value given for t 0.05 when   1 in the table.
From the tables, when   1, t0.05  6.314
 

Now P(t  6.314)   f (t )dt 

1

1
 6.314 1  t 2
1

dt  tan 1 t



6.314
6.314

1 

1

tan 1

()  tan 1 (6.314)  
 2
 7 11 
 1.4137    1.4137  0.05018  0.05
 22  7 

Exercise:

(1) (i) Find t 0.025 when   14 (ii) Find  t 0.01 when   10 (iii) Find t 0.995 when   7
(2) (i) Find t 0.99 when   6 (ii) Find t 0.975 when   24 (iii) Find  t 0.975 when   19
(3) Find (i) P(t  2.821) when   9 (ii) P(t  2.947) when   15
(iii) P(t  1.729) when   19 (iv) P(1.714  t  2.069) when   23
(4) Find k for a sample of size 24 from a normal population such that (i) P(2.069  t  k )  0.965
(ii) P(k  t  2.807)  0.095 (iii) P(k  t  k )  0.9
(5) A random sample of size 25 from a normal population has mean x  45.4 and the standard deviation
s  9.7 . Does this information tend to support or refute the claim that the mean of the population is
  40
(6) A process for making certain ball bearings is under control if the diameters of the bearings have a
mean of 0.5 cm. What can you say about the process if a random sample of 10 of these bearings has a
mean diameter of 0.506 cm and standard deviation of 0.004 cm. ( n  10 , t0.005  3.25 , t  4.7434
and reject the claim)
(7) The tensile strength (1,000 psi) of a new composite can be modeled as a normal distribution. A
random sample of size 25 specimens has mean x  45.3 and standard deviation s  7.9 . Does this
information tend to support or refute the claim that the mean of the population is   40.5 ?
(8) The process of making concrete in a mixer is under control if the rotations per minute of the mixer
have a mean of 22 rpm. What can we say about this process if a sample of 20 of these mixers has a
mean rpm of 22.75 rpm and a standard deviation of 3 rpm?
(9) A manufacturer of fuses claims that with 20% overload, the fuses will blow in 12.4 minutes on the
average. To test this claim, a sample of 20 of the fuses was subjected to a 20% overload, and the
times it took them to blow had a mean of 10.63 minutes and standard deviation of 2.48 minutes. If it
can be assumed that the data constitute a random sample from a normal population, do they tend to
support or refute the manufacturer’s claim
1
(10) The t distribution with 1 degree of freedom is given by f (t )  (1  t 2 ) 1 for    t   . Verify

the value given for t 0.1 when   1 in the table.

Sampling Distribution of the Variance (SDV):

The frequency distribution of the variances of all random samples of fixed size, taken from a population is
called the Sampling Distribution of the Variance (SDV). It is denoted by S 2 .
Examples:

(1) Consider the finite population 1, 2, 3, 4  of size N  4

The number of samples of size n  2 without replacement is given by Cn  C2  6
N 4

The samples are 1, 2, 1, 3 , 1, 4 ,  2, 3 ,  2, 4 , 3, 4 

The means of these 6 samples are respectively 1.5, 2, 2.5, 2.5, 3, 3.5
1 n 1 n 2
The variance   xi  x  or  xi  ( x ) 2
2

n i 1 n i 1
The variances of these 6 samples are respectively given as below.

1 2
2
1  2 2   1.5  2.5  2.25  0.25
2


1 2
2
1  32   2  5  4  1
2


1 2
2
1  4 2   2.5  8.5  6.25  2.25
2


1 2
2
2  32   2.5  6.5  6.25  0.25
2


1 2
2
2  4 2   3  10  9  1
2


1 2
2
3  4 2   3.5  12.5  12.25  0.25
2

The frequency distribution of the these sample variances is given as follows

Sample variance s 2 0.25 1 2.25

Frequency f 3 2 1
Which is the sampling distribution of the Variance (SDV)

Theorem: If S 2 is the random variable which gives the variance a random sample of size n, taken from a

 X  X
n
2

(n  1) S 2 i
normal population having variance  2 , then  2   i 1
is a random variable having
2 2
the Chi-square distribution with the parameter   n  1.

 2  Distribution: A continuous random variable having the probability density function given by
 x
1 1 
f ( x)  for x  0 and   0 , is called the Chi-square (  ) random variable with
2 2 2

x e
 
2 2  
2
parameter  known as degrees of freedom. The probability distribution is called  2  Distribution.

Note:
 

(i) (n)   e x dx   e  x x 2 n1dx

x n 1 2

0 0

(ii) (1)  1 and (n)  (n  1) ! for n  2, 3, 4, 

x x
1  1 
(iii) If   2 , then f ( x)  e 2 e 2 for x  0
21 2
 x x
1  1 
(iv) If   4 , then f ( x)  xe 2  xe 2 for x  0
2 2
2
4
x x
1  1 2 2
(v) If   6 , then f ( x)  x 2e 2  x e for x  0
2 3
3
16

Properties of  2  Distribution:
(i) The curve given by the probability density function is called the  2  curve.
(ii) The  2  curve is lies in the 1st quadrant
(iii) It is not symmetric about any axis and has the following shape.
(iv) It depends on the value of 
(v)  
P  2  0  1 ; That is, the area between  2 – axis and
the curve from 0 to  is 1

2  Notation:
If   0 then  2 is a point on  2 – axis such that P(  2  2 )  
That is, the area between  2 – axis and the curve from 2 to  is 

2 - Table:
In this table the values of  2 are available for different values of  and 
Example:
(i) For   6 or n  7 and   0.025 , 2   02.025  14.449
(ii) For   10 or n  11 and   0.05 , 2  02.05  18.307
(iii) For   15 or n  16 and   0.99 , 2  02.99  5.229

esting a claim using  2 -distribution: To test a given claim using  2 -distribution, we follow the rule
given below.
(i) If  2   02.005 then the claim is accepted
(ii) If  2   02.005 then the claim is rejected

Problems:
(1) Find (i)  02.025 when   12 (ii)  02.05 when   8 (iii)  02.99 when   16
Solution:
(i) When   12 ,  02.025  23.337
(ii) When   8 ,  02.05  15.507
(iii) When   16 ,  02.99  5.812

(2) Find (i) P(  2  12.833) when   5 (ii) P(  2  3.325) when   9 (iii) P(  2  13.12) when
  25
Solution:
(i) When   5 , P(  2  12.833)  P(  2   02.025)  1  P(  2   02.025)  1  0.025  0.975
(ii) When   9 , P(  2  3.325)  P(  2   02.95 )  0.95
(iii) When   25 , P(  2  13.12)  P(  2   02.975)  1  P(  2   02.975)  1  0.975  0.025
(3)   
Find (i) P 2.088   2  16.919 when   9 (ii) P 7.564   2  35.718 when   17 
Solution:
 
(i) When   9 , P 2.088   2  16.919  P  02.99   2   02.05 
   
 P  2   02.99  P  2   02.05  0.99  0.05  0.94
 
(ii) When   17 , P 7.564   2  35.718  P  02.975   2   02.005 
   
 P  2   02.975  P  2   02.005  0.975  0.005  0.97

(4) The claim that the variance of a normal population  2  21.3 is rejected if the variance of a random
sample of size 15 exceeds 39.74. What is the probability that the claim will be rejected even though
 2  21.3 .
Solution: Here n  15,   n  1  14 and  2  21.3
Given that the claim is rejected if S 2  39.74
The probability that the claim will be rejected is given by
 (n  1) S 2 (n  1)  39.74  14 
P( S  39.74)  P 
2
 39.74   P   2  
     21.3 
2 2

  
 P  2  26.120  P  2   02.025  0.025 
(5) The claim that the variance of a normal population  2  18.5 is rejected if the standard deviation of a
random sample of size 25 exceeds 5.7559. What is the probability that the claim will be rejected even
though  2  18.5 .
Solution: Here n  25,   n  1  24 and  2  18.5
Given that the claim is rejected if S  5.7559
The probability that the claim will be rejected is given by
 (n  1) S 2 (n  1)   33.1304  24 
P(S  5.7559)  P(S 2  33.1304)  P   33.1304   P   2  
     
2 2
18.5
 
 P  2  42.9799  P  2   02.01   0.01
(6) A random sample of 10 observations is taken from a normal population having the variance
 2  42.5. Find the approximate probability of obtaining a sample standard deviation between 3.14
and 8.94.
(n  1) S 2
Solution: Here n  10 and   n 1  9 . Also  2  42.5 , we know that  2 
2
The probability of obtaining a sample standard deviation between 3.14 and 8.94 is given by
P(3.14  S  8.94)  P(9.8596  S 2  79.9236)
 (n  1) (n  1) S 2 (n  1) 
 P  9.8596   79.9236 
  2
 2
 2 
 9.8596  9 79.9236  9 
P  2  
 42.5 42.5 

 P 2.0879    16.9249
2

 P  02.99   2   02.05   P  2   02.99   P  2   02.05   0.99  0.05  0.94
  1  2x
(7) The Chi-square distribution with 4 degrees of freedom is given by f ( x)   4 xe x0

 0 x0
Find the probability that the variance of a random sample with   12 will exceed 180.
Solution: Here   4 and   12
The probability that the variance of a random sample will exceed 180 is given by
 (n  1) S 2 (n  1)  180  4 
P( S 2  180)  P 

 180

  P   2  
 P  5
2

 144 
2 2
 

  x 
 
       
x 
 e 2 
x
1 2 1  e 2 
 f ( x)dx   xe dx  x   1  2 
4 4   1  1 
5 5
         
  
  2    2   5
1   
5 5 5
 7 2
 0    10e  4e   e
 2 2 
4    2
Exercise:
(1) Find (i)  02.05 when   6 (ii)  02.99 when   10 (iii)  02.025 when   14 (iv)  02.01 when   21
(2) Determine the probabilities: (i) P(  2  27.688) when   13 (ii) P(  2  18.475) when   7
(iii) P(  2  14.256) when   29
(3)   
Find (i) P 12.461   2  48.278 when   28 (ii) P 28.685   2  29.141 when   14
(4) A random sample of 12 observations is taken from a normal population having the variance
 2  42.5. Find the approximate probability of obtaining a sample variance between 10.057 and
84.691. (Hint: P(10.057  s 2  84.691)  P 2.6029   2  21.92  P  02.995   2   02.025   0.97 )
(5) The claim that the variance of a normal population  2  4 is rejected if the variance of a random
sample of size 9 exceeds 7.7535. What is the probability that the claim will be rejected even though
 2  4.
(6) A random sample of 15 observations is taken from a normal population having variance  2  90.25.
Find the approximate probability of obtaining a sample standard deviation between 7.25 and 10.75.
(7) A manufacturer claims that any of his lists of items cannot have variance more than 1. A sample of
25 items has a variance of 1.2 . Test whether the claim of the manufacturer is correct.
 1  2x
 x0
(8) The Chi-square distribution with 2 degrees of freedom is given by f ( x)   2 e

 0 x0
Find the probability that the standard deviation of a random sample with   10 will exceed 10 2 .

 
(Hint: P( S  10 2 )  P( S  200)  P   4   f ( x)dx )
2 2

4
F  Distribution: A continuous random variable having the probability density function given by
1
      2
 1 2   1  1
1
 2 
 2 x2
f ( x)  for x  0,  1  0 and  2  0 , is called the F  random variable
 1   2   1  2
      1  2
 2   2  1  x 
 2 
with parameters  1 and  2 known as degrees of freedoms. The probability distribution is called F 
Distribution.

Note:
 
(n)   e x dx   e  x x 2 n1dx
x n 1 2
(i)
0 0
(ii) (1)  1 and (n)  (n  1) ! for n  2, 3, 4, 
1
(iii) If  1  2 and  2  2 , then f ( x)  for x  0
(1  x) 2
6x
(iv) If  1  4 and  2  4 , then f ( x)  for x  0
(1  x) 4

Properties of F  Distribution:
(i) The curve given by the probability density function is called the F  curve.
(ii) The F  curve is lies in the 1st quadrant
(iii) It is not symmetric about any axis and has the following shape.
(iv) It depends on the values of  1 and  2
(v) PF  0  1 ; That is, the area between F – axis and
the curve from 0 to  is 1

F  Notation:
If   0 then F is a point on F – axis such that P( F  F )  
That is, the area between F – axis and the curve from F to  is  .
The value of F corresponding to  1 and  2 is denoted by F ( 1 , 2 )
1
Note: F ( 1 , 2 ) 
F1 ( 2 , 1 )

Theorem: If S12 and S 22 are the variances of independent random samples of sizes n1 and n2 , taken from
S12
two normal populations having the same variance, then F  is a random variable having the
S 22
F  distribution with the parameters  1  n1 1 and  2  n2 1 .
S 22
Note: If F  , then the parameters are in the order of  2  n2  1 and  1  n1  1
S12
F0.05  Tables: In this table the values of F0.05 ( 1 , 2 ) are available for different values of  1 and  2
F0.01  Tables: In this table the values of F0.01 ( 1 , 2 ) are available for different values of  1 and  2
Example:

(i) For  1  10,  2  6 and   0.05 , F0.05 ( 1 , 2 )  F0.05 (10, 6)  4.06

1 1 1
(ii) For  1  6,  2  10 and   0.95 , F0.95 ( 1 , 2 )  F0.95 (6,10)   
F10.95 (10, 6) F0.05 (10, 6) 4.06
(iii) For  1  9,  2  13 and   0.01 , F0.01( 1 , 2 )  F0.01(9,13)  4.19
1 1 1
(iv) For  1  13,  2  9 and   0.99 , F0.99 ( 1 , 2 )  F0.99 (13, 9)   
F10.99 (9,13) F0.01(9, 13) 4.19

Problems:
(1) For an F  distribution find
(i) F0.05 with  1  7 and  2  15
(ii) F0.01 with  1  25 and  2  19
(iii) F0.95 with  1  19 and  2  25
(iv) F0.99 with  1  22 and  2  12

Solution:
(i) F0.05 (7,15)  2.71
(ii) F0.01(25,19)  2.91
1 1 1
(iii) F0.95 (19, 25)     0.4739
F10.95 (25,19) F0.05 (25,19) 2.11
1 1 1
(iv) F0.99 (22,12)     0.3205
F10.99 (12, 22) F0.01 (12, 22) 3.12

(2) If two independent random samples of size n1  7 and n2  13 are taken from a normal population,
what is the probability that the variance of the first sample will be at least three times as large as that
of the second sample?
Solution:
Here n1  7 and n2  13 , so  1  6 and  2  12
Therefore, the probability that the variance of the first sample will be at least three times as large as
that of the second sample is given by
 S12 
P( S1  3S 2 )  P 2  3   P( F  3)  P( F  F0.05 )  0.05
2 2
 F0.05 (6,12)  3
 S2 

(3) If two independent random samples of size n1  8 and n2  8 are taken from a normal population,
what is the probability that the variance of the first sample will be at least seven times as large as that
of the second sample?
 Solution:
Here n1  8 and n2  8 , so  1  7 and  2  7
Therefore, the probability that the variance of the first sample will be at least seven times as large as
that of the second sample is given by
 S12 
P( S1  7 S 2 )  P 2  7   P( F  7)  P( F  F0.01 )  0.01
2 2
 F0.01(7, 7)  6.99  7
 2
S 
(4) If two independent random samples of size n1  13 and n2  26 are taken from a normal population,
what is the probability that the variance of the second sample will be at least 2.5 times as large as that
of the first sample?
Solution:
Here n1  13 and n2  26 , so  1  12 and  2  25
Therefore, the probability that the variance of the second sample will be at least 2.5 times as large as
that of the first sample is given by
 S 22 
P( S 2  2.5 S1 )  P 2  2.5   P( F  2.5)  P( F  F0.05 )  0.05
2 2
 F0.05 (25,12)  2.5
 S1 
Problems:

(1) For an F  distribution find

(i) F0.05 with  1  20 and  2  10
(ii) F0.01 with  1  20 and  2  5
(iii) F0.95 with  1  15 and  2  12
(iv) F0.95 with  1  12 and  2  15
(v) F0.99 with  1  5 and  2  20
(vi) F0.99 with  1  20 and  2  5

(2) If two independent random samples of size n1  26 and n2  8 are taken from a normal population,
what is the probability that the variance of the second sample will be at least 2.4 times as large as that
of the first sample?
(3) If two independent random samples of size n1  9 and n2  16 are taken from a normal population,
what is the probability that the variance of the first sample will be at least four times as large as that of
the second sample?
(4) If two independent random samples of size n1  13 and n2  7 are taken from a normal population,
what is the probability that the variance of the first sample will be at least four times as large as that of
the second sample?