INTRODUCTION TO SAMPLING

DISTRIBUTIONS
Unit 2D
SECTION K: Chapter 64 of John Bird textbook

PowerPoint slides from Lombard C, van der Merwe L, Kele T & Mouton A also used
1
 SAMPLING DISTRIBUTIONS

Up to this point you have learnt the following about

discrete, poisson, binomial and continuous probability
distributions

•Discrete probability distribution

•Binomial Distribution

•Poisson Distribution

•Continuous distribution (Normal distribution)

2
 SAMPLING DISTRIBUTIONS

Content

•Introduction

•Sampling distribution of the mean & Sampling error

•The central limit theorem

•Sampling from the normal population

3
 SAMPLING DISTRIBUTIONS
LEARNING OUTCOMES
At the end of the lecture you will be able to:

Apply prior knowledge of sampling distributions to

compute probabilities of the sampling distributions.

4
 SAMPLING DISTRIBUTIONS
Population parameters and associated sample statistics
Measure Sample statistic Population parameter
Mean x µ
Variance s2 2
Standard deviation s 
Proportion p 
•Sampling distribution – the distribution of a sample statistic. It is the
distribution of sample means obtained if samples, all of the same given size,
are selected from the same population;
•A sampling distribution shows the relationship between a sample statistic
and its corresponding parameter.
•To estimate a population mean (µ), we compute the sample mean ( ). x
•Although there is very little chance that the sample mean and the
population mean are identical, we expect them to be quite close. We need to
measure how close.
5
 SAMPLING DISTRIBUTIONS
Sampling distribution of the single mean
To obtain the sampling distribution of the sample mean:

i.draw k samples of the same sizes from a population;

ii.the sample means become random variables as they are
likely to vary from sample to sample;
iii.the mean of these k samples is then:
k

x1  x2  x3    xk x i
x   i 1
k k
6
 SAMPLING DISTRIBUTIONS
Sampling distribution of the single mean
• Unbiased estimator : x  
k

x i
• Mean of sampling distribution of xs ; x  i 1
k

• Standard error of the mean/stdev of all sample

means:  x  
n
• Sampling error – the difference between the calculated
measure of interest (statistic) from the sample and the same
calculated numerical measure of interest (parameter) from
the population.
• Sampling error of the mean = x 
7
 SAMPLING DISTRIBUTIONS
Sampling distribution of the single mean
Theorem 1:
If all possible samples of size n are drawn from a finite population, N,
without replacement, and the standard deviation of the mean values of the
sampling distribution of means is determined then:

x 
  N  n
n  N  1
where  x is the standard deviation of the sampling distribution of means or
standard error of the means and σ is the standard deviation of the population.

The equation is used for a finite population of size N and/or for sampling
without replacement.
  N  n
x 
n  N  1
n is the number in the sample.
8
 SAMPLING DISTRIBUTIONS
Sampling distribution of the single mean
Theorem 1:
If all possible samples of size n are drawn from a finite population, N,
without replacement, and the standard deviation of the mean values of the
sampling distribution of means is determined then:

x 
  N  n
n  N  1
x
where is the standard deviation of the sampling distribution of means or
standard error of the means and σ is the standard deviation of the population.

When N is very large compared with n or when the population is infinite or

when the sampling is done with replacement, the correction factor
approaches 1 . That is  N  n 
1
 N  1

And then  x 
n 9
 SAMPLING DISTRIBUTIONS
Sampling distribution of the single mean
Example: (page 679 of John Bird textbook)

X

N
1 1
X
N

i 1
X i  (5)  5
5
2
n
1 n

i1 N  i1 i  135   25
X  i
2
X 1 2
  5  2  1.4142
N 5
10
 SAMPLING DISTRIBUTIONS
Sampling distribution of the single mean
Example: (page 679 of John Bird textbook)

Sample Sample Sample mean Sample Sample Sample mean

number number
1 3;4 3.5 6 4;6 5
2 3;5 4 7 4;7 5.5
3 3;6 4.5 8 5;6 5.5
4 3;7 5 9 5;7 6
5 4;5 4.5 10 6;7 6.5

11
 SAMPLING DISTRIBUTIONS
Sampling distribution of the single mean
Example: (page 679 of John Bird textbook)

(a) k

x i
x  i 1
k
3.5  4  4.5  5  4.5  5  5.5  5.5  6  6.5

10
50

10
5
12
 SAMPLING DISTRIBUTIONS
Sampling distribution of the single mean
Example: (page 679 of John Bird textbook)

k
1 k 1 k
    257.5    50 
2 2
(c) x 
i
2
x
k 10 i 1 7.5
x  i 1 i 1
   0.866
k 10 10

x 
  N  n 
1.4142  5  2  0.866
n  N  1 2  5  1
N

(d) X i
3  4  5  6  7 25
 i 1
  5
N 5 5 13
 SAMPLING DISTRIBUTIONS
Sampling distribution of the single mean

1 2 3

2
n
1 n 
i1 X  N  i1 X i  14  1  6 2
N
1 1 2

X
N

i 1
X i 
3
(6)  2 
i

N
 3
3
 0.6667  0.8165

14
 SAMPLING DISTRIBUTIONS
Sampling distribution of the mean
Sample/Outcome Sample elements Sample mean
11
1 1;1 2
1

1 2
2 1;2  1.5
2

3 1;3 2
4 2;1 1.5
5 2;2 2
6 2;3 2.5
7 3;1 2
8 3;2 2.5
9 3;3 3
n is the number
k in the sample.
x i
18  0.8165
x  i 1
 2 x    0.5774
k 9 n 2
15
 SAMPLING DISTRIBUTIONS
Sampling distribution of the single mean
Example: (page 680 of John Bird textbook)
The heights of 3000 people are normally distributed with a mean of 175cm and a
standard deviation of 8cm. If random samples are taken of 40 people, predict the
standard deviation and the mean of the sampling distribution of means if sampling is
done
(a)with replacement;
(b)without replacement.

 8
(a)  x    175cm x    1.26cm
n 40

(b)  x  175cm

x 
  N  n 
8  3000  40   1.257cm
n  N  1 40  3000  1

16
 SAMPLING DISTRIBUTIONS
Sampling distribution of the single mean
Theorem 2:
If all possible samples of size n are drawn from a population of size N and
the mean value of the sampling distribution of means  x is determined, then
x  
where  x is the mean value of the population.
•In practice, all possible samples of size N are not drawn from the population;
•However, if the sample size is large (usually taken as 30 or more), then the relationship
between the mean of the sampling distribution of means and the mean of the population
is very near to that shown to this equation x  ;
• Similarly, the relationship between the standard error of the means and the standard

deviation of the population is very near to this equation x  ;
n
•When the sample size n is large, the sampling distribution of the mean is
approximately normally distributed with mean  x standard deviation  x .
and
 2 
X N  ; 
 n
17
 SAMPLING DISTRIBUTIONS
The central limit theorem
It states that if more and more observations are included in the sample (n becomes large), regardless of the
distribution of the population from which the sample was drawn, the sampling distribution of the mean ( )
is approximately normally distributed with mean of and standard deviation of

In short: As n increases the sampling distribution approaches normal distribution

x

It is recommended that if n ≥ 30 the sample is regarded large enough for normal approximation to be
applied x   x 
n
When applying central limit theorem we can write sampling distribution of the sample mean as:

•  2 
X N  ;  18
 n
 SAMPLING DISTRIBUTIONS
The central limit theorem

•
19
 SAMPLING DISTRIBUTIONS
Sampling from the normal population
Prior knowledge:

To calculate the area below a certain value of X, say x; P(X < x) we use

X 
X  N   ;  2  transformation is Z  n(0 ; 1)

 2  X 
X N   ;  transformation is Z= n(0 ; 1)
 n    
 
 n

•
20
 SAMPLING DISTRIBUTIONS
Sampling from the normal population
Examples:
A large batch of electric light bulbs have a mean time to failure of 800 hours and the standard deviation of 60 hours. A random sample of 64 light bulbs
is drawn from the batch. Determine the probability that
(a)
the mean time to failure will be less than 785 hours;
(b)
the mean time to failure will be between 790 hours and 810 hours.

(a)

  800;   60; n  64  60 2

X  N  800; 
 64 
   
   
X   785  800

P X  785  P  Z 
    
  PZ 
 60  
  P  Z  2   P  Z  2 
      
  n     64  
 1  P  Z  2 
 1  0.9772
 0.0228
•
21
 SAMPLING DISTRIBUTIONS
Sampling from the normal population
Examples:
A large batch of electric light bulbs have a mean time to failure of 800 hours and the standard deviation of 60 hours. A random sample of 64 light bulbs is drawn from the batch.
Determine the probability that
(b) the mean time to failure will be between 790 hours and 810 hours.

(b)
 60 2 
  800;   60; n  64 X  N  800; 
 64 

   
   
X   X   790  800 810  800
P  790  X  810   P  Z    P Z  
        60   60  
         
 n   n    64   64  
 P  1.33  Z  1.33
 P  Z  1.33  P  Z  133
 P  Z  1.33  P  Z  1.33 
 0.9082   1  0.9082 
 0.8164
•
22
 SAMPLING DISTRIBUTIONS
Sampling from the normal population
Examples:
 20 
X  N  100; 
1.  16 
2.
P  X  102 
3.

P  X  97 

P  97.8  X  102.5 

•
23
 SAMPLING DISTRIBUTIONS
Sampling from the normal population
Examples:

1.
 
 
X  
P  X  102   P  Z  
   
  
  n 
 
 
102  100
 PZ  
 20 
 
 16 
 P  Z  1.79 
 1  P  Z  1.79 
 1  0.9633
 0.0367
•
24
 SAMPLING DISTRIBUTIONS
Sampling from the normal population
Examples:

2.  
 
X 
P  X  97   P  Z  
    
   
  n  
 
 
97  100
 PZ  
 20 
 
 16 
 P  Z  2.68 
 P  Z  2.68 
 1  P  Z  2.68 
 1  0.9963
 0.0037
•
25
 SAMPLING DISTRIBUTIONS
Sampling from the normal population
Examples:

 
3.  
97.8  100 102.5  100
P  97.8  X  102.5  P  Z  
 20 20 
 
 16 16 
 P  1.97  Z  2.24 
 P  Z  2.24   P  Z  1.97 
 P  Z  2.24   P  Z  1.97 
 P  Z  2.24   1  P  Z  1.97  
 0.9875   1  0.9756
 0.9631
•
26
 SAMPLING DISTRIBUTIONS
Sampling distribution of the proportion
We have discussed the sampling distribution of the mean. What about categorical variable?

Population proportion :

Sample proportion :

Standard error of the proportion :

X
Transformation
 
N
x
p
n

  1  
 
Probability Rules (page 175 – 179) n
p  p 
Z  n(0 ; 1)
   1  
n
•

27
 SAMPLING DISTRIBUTIONS
The Normal distribution
Supplementary exercise 7 (page 210):

A random sample of 200 points was taken from a population with π= 0.30 representing the desired points. What is the probability that the proportion of successes in the sample is
less than or equal to 0.27?

Probability Rules (page 175 – 179)

•
28
 SAMPLING DISTRIBUTIONS
The Normal distribution
Supplementary exercise 7 (page 210):

n  200;   0.30
 
 
 p  
P p  0.27   P Z 
   1   
 
 n 
 
 
0.27  0.3
 PZ  
 0.3  1  0.3  
 
 200 
 P  Z  0.93 
 P  Z  0.93
 1  P  Z  0.93
 1  0.8238
 0.1762

•
29
 SAMPLING DISTRIBUTIONS
The Normal distribution
Supplementary exercise 7 (page 210):

A random sample of 200 points was taken from a population with π= 0.30 representing the desired points. What is the probability that the proportion of successes in the sample will be
between
0.25 and 0.34?

 
 
0.25  0.3 0.34  0.3
P  0.25  p  0.34   P  Z  

Probability
0.3 
Rules1(page
 175
0.3 – 179) 0.3  1  0.3  
 
 200 200 
 P ( 1.54  Z  1.23)
 0.9382  0.8907  1  0.8289

30
Lombard C, van der Merwe L, Kele T and Mouton S.
2012. Elementary Statistics for Business and
Economics.

John Bird. 2017. Higher Engineering Mathematics.

31